Daniel B. McLaughlinJohn R. Olson
HealthcareOperationsManagementT h i r d E d i T i o n
AUPHA/HAP Editorial Board for Graduate Studies
Nir Menachemi, PhD, ChairmanIndiana University
LTC Lee W. Bewley, PhD, FACHEUniversity of Louisville
Jan Clement, PhDVirginia Commonwealth University
Michael Counte, PhDSt. Louis University
Joseph F. Crosby Jr., PhDArmstrong Atlantic State University
Mark L. Diana, PhDTulane University
Peter D. Jacobson, JDUniversity of Michigan
Brian J. Nickerson, PhDIcahn School of Medicine at Mount Sinai
Mark A. Norrell, FACHEIndiana University
Maia Platt, PhDUniversity of Detroit Mercy
Debra Scammon, PhDUniversity of Utah
Tina SmithUniversity of Toronto
Carla Stebbins, PhDDes Moines University
Cynda M. Tipple, FACHEMarymount University
Health Administration Press, Chicago, Illinois
Association of University Programs in Health Administration, Washington, DC
Your board, staff, or clients may also benefit from this book’s insight. For more information on quantity discounts, contact the Health Administration Press Marketing Manager at (312) 424-9450.
This publication is intended to provide accurate and authoritative information in regard to the subject matter covered. It is sold, or otherwise provided, with the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought.
The statements and opinions contained in this book are strictly those of the authors and do not represent the official positions of the American College of Healthcare Executives, the Foundation of the American College of Healthcare Executives, or the Association of University Programs in Health Administration.
Copyright © 2017 by the Foundation of the American College of Healthcare Executives. Printed in the United States of America. All rights reserved. This book or parts thereof may not be reproduced in any form without written permission of the publisher.
21 20 19 18 17 5 4 3 2 1
Library of Congress Cataloging-in-Publication Data Names: McLaughlin, Daniel B., 1945– author. | Olson, John R. (Professor), author.Title: Healthcare operations management / Daniel B. McLaughlin and John R. Olson.Description: Third edition. | Chicago, Illinois : Health Administration Press; Washington, DC : Association of University Programs in Health Administration, [2017] | Includes bibliographical references and index.Identifiers: LCCN 2016046001 (print) | LCCN 2016046925 (ebook) | ISBN 9781567938517 (alk. paper) | ISBN 9781567938524 (ebook) | ISBN 9781567938531 (xml) | ISBN 9781567938548 (epub) | ISBN 9781567938555 (mobi)Subjects: LCSH: Medical care—Quality control. | Health services administration—Quality control. | Organizational effectiveness. | Total quality management.Classification: LCC RA399.A1 M374 2017 (print) | LCC RA399.A1 (ebook) | DDC 362.1068— dc23LC record available at https://lccn.loc.gov/2016046001
The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984. ∞ ™
Acquisitions editor: Janet Davis; Project manager: Joyce Dunne; Cover designer: James Slate; Layout: Cepheus Edmondson
Found an error or a typo? We want to know! Please e-mail it to [email protected], mentioning the book’s title and putting “Book Error” in the subject line.
For photocopying and copyright information, please contact Copyright Clearance Center at www.copyright.com or at (978) 750-8400.
Health Administration Press Association of University ProgramsA division of the Foundation of the American in Health Administration College of Healthcare Executives 1730 M Street, NWOne North Franklin Street, Suite 1700 Suite 407Chicago, IL 60606-3529 Washington, DC 20036(312) 424-2800 (202) 763-7283
To my wife, Sharon, and daughters, Kelly and Katie, for their love and support throughout my career.
—Dan McLaughlin
To my father, Adolph Olson, who passed away in 2011. Your strength as you battled cancer inspired me to change and educate others about our healthcare system.
—John Olson
The first edition of this book was coauthored by Julie Hays. During the final stages of the completion of the book, Julie unexpectedly died. As Dr. Christopher Puto, dean of the Opus College of Business at the University of St. Thomas, said, “Julie cared deeply about students and their learning experience, and she was an accomplished scholar who was well respected by her peers.” This book is a final tribute to Julie’s accomplished career and is dedicated to her legacy.
—Dan McLaughlin and John Olson
vii
BRIEF CONTENTS
Preface …………………………………………………………………………………………xv
Part I Introduction to Healthcare Operations
Chapter 1. The Challenge and the Opportunity …………………………….3
Chapter 2. History of Performance Improvement ………………………..17
Chapter 3. Evidence-Based Medicine and Value-Based Purchasing ….45
Part II Setting Goals and Executing Strategy
Chapter 4. Strategy and the Balanced Scorecard …………………………..71
Chapter 5. Project Management ……………………………………………….97
Part III Performance Improvement Tools, Techniques, and Programs
Chapter 6. Tools for Problem Solving and Decision Making ………..135
Chapter 7. Statistical Thinking and Statistical Problem Solving ……..167
Chapter 8. Healthcare Analytics ……………………………………………..203
Chapter 9. Quality Management: Focus on Six Sigma …………………221
Chapter 10. The Lean Enterprise ………………………………………………255
Part IV Applications to Contemporary Healthcare Operations Issues
Chapter 11. Process Improvement and Patient Flow …………………….281
Chapter 12. Scheduling and Capacity Management ………………………323
Chapter 13. Supply Chain Management ……………………………………..345
Chapter 14. Improving Financial Performance with Operations Management ………………………………………………………..369
viii Brief Contents
Part V Putting It All Together for Operational Excellence
Chapter 15. Holding the Gains …………………………………………………391
Glossary …………………………………………………………………………………….411Index ………………………………………………………………………………………..419About the Authors ………………………………………………………………………..437
ix
DETAILED CONTENTS
Preface …………………………………………………………………………………………xv
Part I Introduction to Healthcare Operations
Chapter 1. The Challenge and the Opportunity …………………………….3Overview ………………………………………………………………..3The Purpose of This Book ………………………………………….3The Challenge ………………………………………………………….4The Opportunity ……………………………………………………..6A Systems Look at Healthcare …………………………………….8An Integrating Framework for Operations Management
in Healthcare ……………………………………………………..12Conclusion …………………………………………………………….15Discussion Questions ………………………………………………15References ……………………………………………………………..15
Chapter 2. History of Performance Improvement ………………………..17Operations Management in Action …………………………….17Overview ………………………………………………………………17Background……………………………………………………………18Knowledge-Based Management …………………………………20History of Scientific Management ………………………………22Project Management ……………………………………………….26Introduction to Quality ……………………………………………27Philosophies of Performance Improvement ………………….34Supply Chain Management ……………………………………….38Big Data and Analytics …………………………………………….40Conclusion …………………………………………………………….41Discussion Questions ………………………………………………41References ……………………………………………………………..42
Chapter 3. Evidence-Based Medicine and Value-Based Purchasing ….45Operations Management in Action …………………………….45
x Detai led Contents
Overview ………………………………………………………………45Evidence-Based Medicine …………………………………………46Tools to Expand the Use of Evidence-Based Medicine …..54Clinical Decision Support …………………………………………59The Future of Evidence-Based Medicine and Value
Purchasing …………………………………………………………62Vincent Valley Hospital and Health System and Pay for
Performance ………………………………………………………63Conclusion …………………………………………………………….64Discussion Questions ………………………………………………64Note …………………………………………………………………….64References ……………………………………………………………..65
Part II Setting Goals and Executing Strategy
Chapter 4. Strategy and the Balanced Scorecard …………………………..71Operations Management in Action …………………………….71Overview ………………………………………………………………71Moving Strategy to Execution …………………………………..72The Balanced Scorecard in Healthcare ……………………….75The Balanced Scorecard as Part of a Strategic
Management System ……………………………………………76Elements of the Balanced Scorecard System …………………76Conclusion …………………………………………………………….93Discussion Questions ………………………………………………93Exercises ……………………………………………………………….94References ……………………………………………………………..94Further Reading ……………………………………………………..95
Chapter 5. Project Management ……………………………………………….97Operations Management in Action ……………………………97Overview ………………………………………………………………97Definition of a Project ……………………………………………..99Project Selection and Chartering ……………………………..100Project Scope and Work Breakdown …………………………107Scheduling …………………………………………………………..113Project Control …………………………………………………….117Quality Management, Procurement, the Project
Management Office, and Project Closure ………………120Agile Project Management ……………………………………..124Innovation Centers ………………………………………………..125
xiDetai led Contents
The Project Manager and Project Team …………………….126Conclusion …………………………………………………………..129Discussion Questions …………………………………………….129Exercises ……………………………………………………………..129References ……………………………………………………………130Further Reading ……………………………………………………130
Part III Performance Improvement Tools, Techniques, and Programs
Chapter 6. Tools for Problem Solving and Decision Making ………..135Operations Management in Action …………………………..135Overview …………………………………………………………….135Decision-Making Framework …………………………………..136Mapping Techniques ……………………………………………..138Problem Identification Tools …………………………………..143Analytical Tools …………………………………………………….153Implementation: Force Field Analysis ……………………….162Conclusion …………………………………………………………..163Discussion Questions …………………………………………….163Exercises ……………………………………………………………..164References ……………………………………………………………165
Chapter 7. Statistical Thinking and Statistical Problem Solving ……..167Operations Management in Action …………………………..167Overview: Statistical Thinking in Healthcare ………………167Foundations of Data Analysis …………………………………..169Graphic Tools ……………………………………………………….169Mathematical Descriptions ……………………………………..174Probability …………………………………………………………..178Confidence Intervals and Hypothesis Testing ……………..185Simple Linear Regression………………………………………..192Conclusion …………………………………………………………..198Discussion Questions …………………………………………….199Exercises ……………………………………………………………..199References ……………………………………………………………201
Chapter 8. Healthcare Analytics ………………………………………………203Operations Management in Action …………………………..203Overview …………………………………………………………….203What Is Analytics in Healthcare? ………………………………203Introduction to Data Analytics ………………………………..205
xii Detai led Contents
Data Visualization …………………………………………………209Data Mining for Discovery ……………………………………..214Conclusion …………………………………………………………..217Discussion Questions …………………………………………….218Note …………………………………………………………………..218References …………………………………………………………..219
Chapter 9. Quality Management—Focus on Six Sigma ……………….221Operations Management in Action …………………………..221Overview …………………………………………………………….221Defining Quality …………………………………………………..222Cost of Quality ……………………………………………………..223The Six Sigma Quality Program ……………………………….225Additional Quality Tools ………………………………………..240Riverview Clinic Six Sigma Generic Drug Project ……….245Conclusion …………………………………………………………..250Discussion Questions …………………………………………….250Exercises ……………………………………………………………..250References ……………………………………………………………253
Chapter 10. The Lean Enterprise ………………………………………………255Operations Management in Action …………………………..255Overview …………………………………………………………….255What Is Lean? ………………………………………………………256Types of Waste ……………………………………………………..257Kaizen …………………………………………………………………259Value Stream Mapping …………………………………………..259Additional Measures and Tools ………………………………..261The Merging of Lean and Six Sigma Programs …………..274Conclusion …………………………………………………………..276Discussion Questions …………………………………………….276Exercises ……………………………………………………………..277References ……………………………………………………………277
Part IV Applications to Contemporary Healthcare Operations Issues
Chapter 11. Process Improvement and Patient Flow …………………….281Operations Management in Action …………………………..281Overview …………………………………………………………….281Problem Types ……………………………………………………..282Patient Flow …………………………………………………………283
xiiiDetai led Contents
Process Improvement Approaches ……………………………284The Science of Lines: Queuing Theory …………………….292Process Improvement in Practice ……………………………..304Conclusion …………………………………………………………..318Discussion Questions …………………………………………….319Exercises ……………………………………………………………..319References ……………………………………………………………320Further Reading ……………………………………………………321
Chapter 12. Scheduling and Capacity Management ………………………323Operations Management in Action …………………………..323Overview …………………………………………………………….323Hospital Census and Rough-Cut Capacity Planning ……324Staff Scheduling ……………………………………………………326Job and Operation Scheduling and Sequencing Rules ….330Patient Appointment Scheduling Models …………………..334Advanced-Access Patient Scheduling …………………………337Conclusion …………………………………………………………..341Discussion Questions …………………………………………….341Exercises ……………………………………………………………..341References ……………………………………………………………342
Chapter 13. Supply Chain Management ……………………………………..345Operations Management in Action …………………………..345Overview …………………………………………………………….345Supply Chain Management ……………………………………..346Tracking and Managing Inventory ……………………………347Demand Forecasting ……………………………………………..349Order Amount and Timing …………………………………….354Inventory Systems …………………………………………………362Procurement and Vendor Relationship Management ……364Strategic View ………………………………………………………364Conclusion …………………………………………………………..365Discussion Questions …………………………………………….366Exercises ……………………………………………………………..366References ……………………………………………………………368
Chapter 14. Improving Financial Performance with Operations Management ………………………………………………………..369Operations Management in Action …………………………..369Overview: The Financial Pressure for Change …………….369
xiv Detai led Contents
Making Ends Meet on Medicare and the Pressure of Narrow Networks ……………………………………………..370
Conclusion …………………………………………………………..386Discussion Questions …………………………………………….386Exercises ……………………………………………………………..387Note …………………………………………………………………..387References ……………………………………………………………387
Part V Putting It All Together for Operational Excellence
Chapter 15. Holding the Gains …………………………………………………391Overview …………………………………………………………….391Approaches to Holding Gains ………………………………….391Which Tools to Use: A General Algorithm …………………397Data and Statistics …………………………………………………404Operational Excellence …………………………………………..405The Healthcare Organization of the Future ……………….407Conclusion …………………………………………………………..408Discussion Questions …………………………………………….408Case Study …………………………………………………………..409References ……………………………………………………………410
Glossary …………………………………………………………………………………….411Index ………………………………………………………………………………………..419About the Authors ………………………………………………………………………..437
xv
PREFACE
This book is intended to help healthcare professionals meet the challenges and take advantage of the opportunities found in healthcare today. We believe that the answers to many of the dilemmas faced by the US healthcare system, such as increasing costs, inadequate access, and uneven quality, lie in organizational operations—the nuts and bolts of healthcare delivery. The healthcare arena is filled with opportunities for significant operational improvements. We hope that this book encourages healthcare management students and working profession-als to find ways to improve the management and delivery of healthcare, thereby increasing the effectiveness and efficiency of tomorrow’s healthcare system.
Many industries outside healthcare have successfully used the programs, techniques, and tools of operations improvement for decades. Leading health-care organizations have now begun to employ the same tools. Although numer-ous other operations management texts are available, few focus on healthcare operations, and none takes an integrated approach. Students interested in healthcare process improvement have difficulty seeing the applicability of the science of operations management when most texts focus on widgets and production lines rather than on patients and providers.
This book covers the basics of operations improvement and provides an overview of the significant trends in the healthcare industry. We focus on the strategic implementation of process improvement programs, techniques, and tools in the healthcare environment, with its complex web of reimburse-ment systems, physician relations, workforce challenges, and governmental regulations. This integrated approach helps healthcare professionals gain an understanding of strategic operations management and, more important, its applicability to the healthcare field.
How This Book Is Organized
We have organized this book into five parts:
1. Introduction to Healthcare Operations2. Setting Goals and Executing Strategy3. Performance Improvement Tools, Techniques, and Programs
xvi Preface
4. Applications to Contemporary Healthcare Operations Issues5. Putting It All Together for Operational Excellence
Although this structure is helpful for most readers, each chapter also stands alone, and the chapters can be covered or read in any order that makes sense for a particular course or student.
The first part of the book, Introduction to Healthcare Operations, begins with an overview of the challenges and opportunities found in today’s healthcare environment (chapter 1). We follow with a history of the field of management science and operations improvement (chapter 2). Next, we discuss two of the most influential environmental changes facing healthcare today: evidence-based medicine and value-based purchasing, or simply value purchasing (chapter 3).
In part II, Setting Goals and Executing Strategy, chapter 4 highlights the importance of tying the strategic direction of the organization to operational initiatives. This chapter outlines the use of the balanced scorecard technique to execute and monitor these initiatives toward achieving organizational objec-tives. Typically, strategic initiatives are large in scope, and the tools of project management (chapter 5) are needed to successfully manage them. Indeed, the use of project management tools can help to ensure the success of any size project. Strategic focus and project management provide the organizational foundation for the remainder of this book.
The next part of the book, Performance Improvement Tools, Tech-niques, and Programs, provides an introduction to basic decision-making and problem-solving processes and describes some of the associated tools (chapter 6). Most performance improvement initiatives (e.g., Six Sigma, Lean) follow these same processes and make use of some or all of the tools discussed in chapter 6.
Good decisions and effective solutions are based on facts, not intuition. Chapter 7 provides an overview of data collection processes and analysis tech-niques to enable fact-based decision making. Chapter 8 builds on the statistical approaches of chapter 7 by presenting the new tools of advanced analytics and big data.
Six Sigma, Lean, simulation, and supply chain management are specific philosophies or techniques that can be used to improve processes and systems. The Six Sigma methodology (chapter 9) is the latest manifestation of the use of quality improvement tools to reduce variation and errors in a process. The Lean methodology (chapter 10) is focused on eliminating waste in a system or process.
The fourth section of the book, Applications to Contemporary Health-care Operations Issues, begins with an integrated approach to applying the various tools and techniques for process improvement in the healthcare environ-ment (chapter 11). We then focus on a special and important case of process improvement: patient scheduling in the ambulatory setting (chapter 12).
xviiPreface
Supply chain management extends the boundaries of the hospital or healthcare system to include both upstream suppliers and downstream custom-ers, and this is the focus of chapter 13. The need to “bend” the healthcare cost inflation curve downward is one of the most pressing issues in healthcare today, and the use of operations management tools to achieve this goal is addressed in chapter 14.
Part V, Putting It All Together for Operational Excellence, concludes the book with a discussion of strategies for implementing and maintaining the focus on continuous improvement in healthcare organizations (chapter 15).
Many features in this book should enhance student understanding and learning. Most chapters begin with a vignette, called Operations Management in Action, that offers a real-world example related to the content of that chapter. Throughout the book, we use a fictitious but realistic organization, Vincent Valley Hospital and Health System, to illustrate the various tools, techniques, and programs discussed. Each chapter concludes with questions for discussion, and parts II through IV include exercises to be solved.
We include abundant examples throughout the text of the use of various contemporary software tools essential for effective operations management. Readers will see notes appended to some of the exhibits, for example, that indicate what software was used to create charts, graphs, and so on from the data provided. Healthcare leaders and managers must be experts in the appli-cation of these tools and stay current with the latest versions. Just as we ask healthcare providers to stay up-to-date with the latest clinical advances, so too must healthcare managers stay current with basic software tools.
Acknowledgments
A number of people contributed to this work. Dan McLaughlin would like to thank his many colleagues at the University of St. Thomas Opus College of Business. Specifically, Dr. Ernest Owens provided guidance on the project man-agement chapter, and Dr. Michael Sheppeck assisted on the human resources implications of operations improvement. Dean Stefanie Lenway and Associate Dean Michael Garrison encouraged and supported this work and helped create our new Center for Innovation in the Business of Healthcare.
Dan would also like to thank the outstanding professionals at Hennepin County Medical Center in Minneapolis, Minnesota, who provided many of the practical and realistic examples in this book. They continue to be invaluable healthcare resources for all of the residents of Minnesota.
John Olson would like to thank his many colleagues at the University of St. Thomas Opus College of Business. In addition, he would like to thank the Minnesota Hospital Association (MHA). Attributing much of his under-standing of healthcare analytics to working with the highly professional staff
xviii Preface
of the MHA, he wishes to acknowledge Rahul Korrane, Tanya Daniels, Mark Sonneborn, and Julie Apold (now with Optum) as true agents for change in the US healthcare system.
The dedicated employees of the Veterans Administration have helped John embrace the challenges that confront healthcare today—in particular Christine Wolohan, Lori Fox, Susan Chattin, Eric James, Denise Lingen, and Carl (Marty) Young of the continuous improvement group, who are helping to create an organization of excellence. John acknowledges their dedication to serving US veterans and the amazing, high-quality service they deliver.
John and Dan also want to thank the skilled professionals of Health Administration Press for their support, especially Janet Davis, acquisitions edi-tor, and Joyce Dunne, who edited this third edition.
Finally, this book still contains many passages that were written by Julie Hays and are a tribute to her skill and dedication to the field of operations management.
Instructor Resources
This book’s Instructor Resources include PowerPoint slides; an updated test bank; teaching notes for the end-of-chapter exercises; Excel files and cases for selected chapters; and new case studies, for most chapters, with accompanying teaching notes. Each of the new case studies is one to three pages long and is suitable for one class session or an online learning module.
For the most up-to-date information about this book and its Instructor Resources, visit ache.org/HAP and browse for the book’s title or author names.
This book’s Instructor Resources are available to instructors who adopt this book for use in their course. For access information, please e-mail [email protected].
Student Resources
Case studies, exercises, tools, and web links to resources are available at ache.org/books/OpsManagement3.
PART
INTRODUCTION TO HEALTHCARE OPERATIONS
I
CHAPTER
3
THE CHALLENGE AND THE OPPORTUNITY
The Purpose of This Book
Excellence in healthcare derives from four major areas of expertise: clinical care, population health, leadership, and operations. Although clinical expertise, the health of a population, and leadership are critical to an orga-nization’s success, this book focuses on operations—how to deliver high-quality health services in a consistent, efficient manner.
Many books cover opera-tional improvement tools, and some focus on using these tools in health-care environments. So why have we devoted a book to the broad topic of healthcare operations? Because we see a need for organizations to adopt an integrated approach to operations improvement that puts all the tools in a logical context and provides a road map for their use. An integrated approach uses a clinical analogy: First, find and diagnose an operations issue. Second, apply the appropriate treat-ment tool to solve the problem.
The field of operations research and management science is too deep to cover in one book. In Healthcare Operations Management, only those tools and techniques currently being deployed in leading healthcare organi-zations are covered, in part so that we may describe them in enough detail
1OVE RVI EW
The challenges and opportunities in today’s complex healthcare
delivery systems demand that leaders take charge of their opera-
tions. A strong operations focus can reduce costs, increase safety—for
patients, visitors, and staff alike—improve clinical outcomes, and allow
an organization to compete effectively in an aggressive marketplace.
In the recent past, success for many organizations in the US
healthcare system has been achieved by executing a few critical strate-
gies: First, attract and retain talented clinicians. Next, add new technol-
ogy and specialty care services. Finally, find new methods to maximize
the organization’s reimbursement for these services. In most organiza-
tions, new services, not ongoing operations, were the key to success.
However, that era is ending. Payer resistance to cost
increases and a surge in public reporting on the quality of health-
care are forces driving a major change in strategy. The passage of
the Affordable Care Act (ACA) in 2010 represented a culmination
of these forces. Although portions of this law may be repealed or
changed, the general direction of health policy in the United States
has been set. To succeed in this new environment, a healthcare
enterprise must focus on making significant improvements in its
core operations.
This book is about improvement and how to get things done.
It offers an integrated, systematic approach and set of contemporary
operations improvement tools that can be used to make significant
gains in any organization. These tools have been successfully deployed
in much of the global business community for more than 40 years and
now are being used by leading healthcare delivery organizations.
This chapter outlines the purpose of the book, identifies
challenges that healthcare systems currently face, presents a systems
view of healthcare, and provides a comprehensive framework for the
use of operations tools and methods in healthcare. Finally, Vincent
Valley Hospital and Health System (VVH), the fictional healthcare
delivery system used in examples throughout the book, is described.
Healthcare Operat ions Management4
to enable students and practitioners to use them in their work. Each chap-ter provides many references for further reading and deeper study. We also
include additional resources, case studies, exercises, and tools on the companion website that accompanies this book.
This book is organized so that each chapter builds on the previous one and is cross-referenced. However, each chapter also stands alone, so a reader interested in Six Sigma can start in chapter 9 and then move to the other chapters in any order he wishes.
This book does not specifically explore quality in healthcare as defined by the many agencies that have as their mission to ensure healthcare quality, such as The Joint Commission, the National Committee for Quality Assurance, the National Quality Forum, and some federally funded quality improvement organizations. In particular, The Healthcare Quality Book: Vision, Strategy, and Tools (Joshi et al. 2014) delves into this perspective in depth and may be considered a useful companion to this book. However, the systems, tools, and techniques discussed here are essential to completing the operational improve-ments needed to meet the expectations of these quality assurance organizations.
The Challenge
Health spending is projected to grow 1.3 percent faster per year than the gross domestic product (GDP) between 2015 and 2025. As a result, the health share of GDP is expected to rise from 17.5 percent in 2014 to 20.1 percent by 2025 (CMS 2015). In addition, healthcare spending is placing increasing pressure on the federal budget. In its expenditure report summary, the Centers for Medicare & Medicaid Services (CMS 2015) notes that “federal, state and local governments are projected to finance 47 percent of national health spending by 2024 (from 45 percent in 2014).”
Despite the high cost, the value delivered by the system has been ques-tioned by many policymakers. For example, unexplained quality variations in healthcare were estimated in 1999 to result in 44,000 to 98,000 preventable deaths every year (IOM 1999). And those problems persist. A 2010 study of hospitals in North Carolina showed a high rate of adverse events, unchanged over time even though hospitals had sought to improve the safety of inpatient care (Landrigan et al. 2010).
Clearly, the pace of quality improvement is slow. “National Healthcare Quality Report, 2009,” published by the Agency for Healthcare Research and Quality (AHRQ), reported: “Quality is improving at a slow pace. Of the 33 core measures, two-thirds improved, 14 (42%) with a rate between 1% and 5% per year and 8 (24%) with a rate greater than 5% per year. . . . The
Agency for Healthcare Research and Quality (AHRQ)A federal agency that is part of the Department of Health and Human Services. It provides leadership and funding to identify and communicate the most effective methods to deliver high-quality healthcare in the United States.
On the web at ache.org/books/OpsManagement3
Chapter 1 : The Chal lenge and the Oppor tunity 5
median rate of change was 2% per year. Across all 169 measures, results were similar, although the median rate of change was slightly higher at 2.3% per year” (AHRQ 2010).
These problems were studied in the landmark work of the Institute of Medicine (IOM), Crossing the Quality Chasm: A New Health System for the 21st Century. The IOM (2001) panel concluded that the knowledge to improve patient care is available, but a gap—a chasm—separates that knowledge from everyday practice. The panel summarized the goals of a new health system in terms of six aims, as described in exhibit 1.1.
Although this seminal work was published more than a decade ago, its goals still guide much of the quality improvement effort today.
Many healthcare leaders are addressing these issues by capitalizing on proven tools employed by other industries to ensure high performance and quality outcomes. For major change to occur in the US health system, however, these strategies must be adopted by a broad spectrum of healthcare providers and implemented consistently throughout the continuum of care—in ambula-tory, inpatient, acute, and long-term care settings—to undergird population health initiatives.
The payers for healthcare must engage with the delivery system to find new ways to partner for improvement. In addition, patients need to assume strong financial and self-care roles in this new system. The ACA and subsequent health policy initiatives provide many new policies to support the achievement of these goals.
Although not all of the IOM goals can be accomplished through opera-tional improvements, this book provides methods and tools to actively change the system toward accomplishing several aspects of these aims.
Institute of Medicine (IOM)The healthcare arm of the National Academy of Sciences; an independent, nonprofit organization providing unbiased and authoritative advice to decision makers and the public.
1. Safe, avoiding injuries to patients from the care that is intended to help them
2. Effective, providing services based on scientific knowledge to all who could benefit, and refraining from providing services to those not likely to benefit (avoiding underuse and overuse, respectively);
3. Patient centered, providing care that is respectful of and responsive to individual patient preferences, needs, and values, and ensuring that patient values guide all clinical decisions;
4. Timely, reducing wait times and harmful delays for both those who receive and those who give care;
5. Efficient, avoiding waste of equipment, supplies, ideas, and energy; and6. Equitable, providing care that does not vary in quality because of per-
sonal characteristics such as gender, ethnicity, geographic location, and socioeconomic status.
EXHIBIT 1.1Six Aims for the US Health System
Source: Information from IOM (2001).
Healthcare Operat ions Management6
The Opportunity
While the current US health system presents numerous challenges, opportuni-ties for improvement are emerging as well. A number of major trends provide hope that significant change is possible. The following trends represent this groundswell:
• Informatics systems are maturing, and big data and analytics tools are becoming ever more powerful.
• Automation, robots, and the Internet of Things will begin to replace human labor in healthcare.
• Supply chains and the relationships among health plans, healthcare systems, and individual providers are changing through mergers, partnerships, and acquisitions.
• Primary care is being redesigned with new provider models and new tools, such as telemedicine and mobile applications.
• Medicine itself is undergoing rapid change with the adoption of precision medicine tools, such as pharmacogenomics, to individualize patient treatments.
• A new emphasis on population health accountability and management will lead to healthier environments and lifestyles.
Evidence-Based MedicineThe use of evidence-based medicine (EBM) for the delivery of healthcare in the United States is the result of 40 years of work by some of the most progres-sive and thoughtful practitioners in the nation. The movement has produced an array of care guidelines, care patterns, and shared decision-making tools for caregivers and patients.
The impact of EBM on care delivery can be powerful. Rotter and col-leagues (2010) reviewed 27 studies worldwide including 11,938 patients and assessed the use of clinical pathways. They found that the cost of care for patients whose treatment was delivered using the pathways was $4,919 per admission less than for those who did not receive pathway-centered care.
Comprehensive resources are available to healthcare organizations that wish to emphasize EBM. For example, the National Guideline Clearinghouse (NGC 2016) is a comprehensive database of more than 4,000 evidence-based clinical practice guidelines and related documents. NGC is an initiative of AHRQ, which itself is a division of the US Department of Health and Human Services. NGC was originally created in partnership with the American Medical Association and American Association of Health Plans, now America’s Health Insurance Plans.
Evidence-based medicine (EBM)The conscientious and judicious use of the best current evidence in making decisions about the care of individual patients.
Chapter 1 : The Chal lenge and the Oppor tunity 7
Big Data and AnalyticsHealthcare delivery has been slow to adopt information technologies, but many organizations have now implemented electronic health record (EHR) systems and other automated tools. Although implementation of these systems
Evidence-Based Medicine (EBM)The Institute of Medicine has been a leading advocate for comparative effec-tiveness research, the National Academy of Sciences’ concomitant deploy-ment of EBM. The IOM Roundtable on Value and Science-Driven Healthcare has set a “goal that by the year 2020, 90 percent of clinical decisions will be supported by accurate, timely, and up-to-date clinical information and will reflect the best available evidence” (IOM 2011, 4; emphasis in original).
To achieve this end, the IOM Roundtable recommends a sophisticated set of processes and infrastructure, which it describes as follows (IOM 2011, 10).
Infrastructure Required for Comparative Effectiveness Research: Common
Themes
• Care that is effective and efficient stems from the integrity of the
infrastructure for learning.
• Coordinating work and ensuring standards are key components of the
evidence infrastructure.
• Learning about effectiveness must continue beyond the transition from
testing to practice.
• Timely and dynamic evidence of clinical effectiveness requires bridging
research and practice.
• Current infrastructure planning must build to future needs and
opportunities.
• Keeping pace with technological innovation compels more than a head-
to-head and time-to-time focus.
• Real-time learning depends on health information technology
investment.
• Developing and applying tools that foster real-time data analysis is an
important element.
• A trained workforce is a vital link in the chain of evidence stewardship.
• Approaches are needed that draw effectively on both public and private
capacities.
• Efficiency and effectiveness compel globalizing evidence and localizing
decisions.
In short, EBM is the conscientious and judicious use of the best cur-rent evidence in making decisions about the care of individual patients.
Healthcare Operat ions Management8
has sometimes been organizationally painful, EHRs are now becoming mature enough to have a substantial positive impact on operations.
In addition, data science computer engineering has evolved to provide significant new tools in the following areas:
• Big data storage and retrieval—high volume, high velocity, and high variety of data types
• New analytical tools for reporting and prediction• Portable and wearable devices• Interoperabilty of devices and databases
Chapter 8 describes a set of analytical tools to fully utilize these new resources.
Active and Engaged ConsumersConsumers are assuming new roles in their own care through the use of health education and information and by partnering effectively with their healthcare providers. Personal maintenance of wellness though a healthy lifestyle is one essential component. Understanding one’s disease and treatment options and having an awareness of the cost of care are also important responsibilities of the consumer.
Patients are becoming good consumers of healthcare by finding and considering price information when selecting providers and treatments. Many employers now offer high-deductible health plans with accompanying health savings accounts (HSAs). This type of consumer-directed healthcare is likely to grow and increase pressure on providers to deliver cost-effective, customer-sensitive, high-quality care. In addition, the ACA provides new tools for employ-ers to motivate their employees financially to engage in healthy lifestyles.
The healthcare delivery system of the future will support and empower active, informed consumers.
A Systems Look at HealthcareThe Clinical SystemTo participate in the improvement of healthcare operations, healthcare leaders must understand the series of interconnected systems that influence the delivery of clinical care (exhibit 1.2).
In the patient care microsystem, the healthcare professional provides hands-on care to the patient. Elements of the clinical microsystem include
• the team of health professionals who provide clinical care to the patient,• the tools that the team has at its disposal to diagnose and treat the
patient (e.g., imaging capabilities, laboratory tests, drugs), and
Health savings account (HSA)A personal monetary account that can only be used for healthcare expenses. The funds are not taxed, and the balance can be rolled over from year to year. HSAs are normally used with high-deductible health insurance plans.
Consumer-directed healthcareIn general, the consumer (patient) is well informed about healthcare prices and quality and makes personal buying decisions on the basis of this information. The health savings account is frequently included as a key component of consumer-directed healthcare.
Patient care microsystemThe level of healthcare delivery that includes providers, technology, and treatment processes.
Chapter 1 : The Chal lenge and the Oppor tunity 9
• the logic for determining the appropriate treatments and the processes to deliver that care.
Because common conditions (e.g., hypertension) affect a large number of patients, clinical research has been conducted to determine the most effec-tive ways to treat these patients. Therefore, in many cases, the organization and functioning of the microsystem can be optimized. Process improvements can be made at this level to ensure that the most effective, least costly care is delivered. In addition, the use of EBM guidelines can help ensure that the patient receives the correct treatment at the correct time.
The organizational infrastructure also influences the effective delivery of care to the patient. Ensuring that providers have the correct tools and skills is an important element of infrastructure.
The EHR is one of the most important advances in the clinical micro-system for both process improvement and the wider adoption of EBM.
Another key component of infrastructure is the leadership displayed by senior staff. Without leadership, progress and change do not occur.
Finally, the environment strongly influences the delivery of care. Key environmental factors include market competition, government regulation, demographics, and payer policies. An organization’s strategy is frequently influ-enced by such factors (e.g., a new regulation from Medicare, a new competitor).
Many of the systems concepts regarding healthcare delivery were ini-tially developed by Avedis Donabedian. These fundamental contributions are discussed in depth in chapter 2.
OrganizationLevel C
MicrosystemLevel B
PatientLevel A
EnvironmentLevel D
EXHIBIT 1.2A Systems View of Healthcare
Source: Ransom, Joshi, and Nash (2005). Based on Ferlie, E., and S. M. Shortell. 2001. “Improving the Quality of Healthcare in the United Kingdom and the United States: A Framework for Change.” Milbank Quarterly 79 (2): 281–316.
Healthcare Operat ions Management10
System Stability and ChangeElements in each layer of this system interact. Peter Senge (1990) provides a useful theory for understanding the interaction of elements in a complex system such as healthcare. In his model, the structure of a system is the primary mecha-nism for producing an outcome. For example, the presence of an organized structure of facilities, trained professionals, supplies, equipment, and EBM care guidelines leads to a high probability of producing an expected clinical outcome.
No system is ever completely stable. Each system’s performance is modi-fied and controlled by feedback (exhibit 1.3). Senge (1990, 75) defines feedback as “any reciprocal flow of influence. In systems thinking it is an axiom that every influence is both cause and effect.” As shown in exhibit 1.3, increased salaries provide an incentive for employees to achieve improvement in performance level. This improved performance leads to enhanced financial performance and profitability for the organization, and increased profits provide additional funds for higher salaries, and the cycle continues. Another frequent example in healthcare delivery is patient lab results that directly influence the medication
+
+
+
–
–
Employeemotivation
Salaries
Financialperformance,profit
Add orreduce staff
Actualstaffinglevel
Compare actual toneeded staff basedon patient demand
EXHIBIT 1.3Systems with
Reinforcing and Balancing
Feedback
Chapter 1 : The Chal lenge and the Oppor tunity 11
ordered by a physician. A third example is a financial report that shows an over-expenditure in one category that prompts a manager to reduce spending to meet budget goals.
A more complete definition of a feedback-driven operational system includes an operational process, a sensor that monitors process output, a feed-back loop, and a control that modifies how the process operates.
Feedback can be either reinforcing or balancing. Reinforcing feedback prompts change that builds on itself and amplifies the outcome of a process, taking the process further and further from its starting point. The effect of rein-forcing feedback can be either positive or negative. For example, a reinforcing change of positive financial results for an organization could lead to increases in salaries, which would then lead to even better financial performance because the employees are highly motivated. In contrast, a poor supervisor could cause employee turnover, possibly resulting in short staffing and even more turnover.
Balancing feedback prompts change that seeks stability. A balancing feedback loop attempts to return the system to its starting point. The human body provides a good example of a complex system that has many balancing feedback mechanisms. For example, an overheated body prompts perspiration until the body is cooled through evaporation. The clinical term for this type of balance is homeostasis. A treatment process that controls drug dosing via real-time monitoring of the patient’s physiological responses is an example of balancing feedback. Inpatient unit staffing levels that determine where in a hospital patients are admitted is another. All of these feedback mechanisms are designed to maintain balance in the system.
A confounding problem with feedback is delay. Delays occur when interruptions arise between actions and consequences. In the midst of delays, systems tend to “overshoot” and thus perform poorly. For example, an emer-gency department might experience a surge in patients and call in additional staff. When the surge subsides, the added staff stay on shift but are no longer needed, and unnecessary expense is incurred.
As healthcare leaders focus on improving their operations, they must understand the systems in which change resides. Every change will be resisted and reinforced by feedback mechanisms, many of which are not clearly visible. Taking a broad systems view can improve the effectiveness of change.
Many subsystems in the total healthcare system are interconnected. These connections have feedback mechanisms that either reinforce or balance the subsystem’s performance. Exhibit 1.4 shows a simple connection that origi-nates in the environmental segment of the total health system. Each process has both reinforcing and balancing feedback.
This general systems model can be converted to a more quantitative system dynamics model, which is useful as part of a predictive analytics system. This concept is addressed in more depth in chapter 8.
Healthcare Operat ions Management12
An Integrating Framework for Operations Management in Healthcare
The five-part framework of this book (illustrated in exhibit 1.5) reflects our view that effective operations management in healthcare consists of highly focused strategy execution and organizational change accompanied by the disciplined use of analytical tools, techniques, and programs. An organization needs to understand the environment, develop a strategy, and implement a system to effectively deploy this strategy. At the same time, the organization must become adept at using all the tools of operations improvement contained in this book. These improvement tools can then be combined to attack the fundamental challenges of operating a complex healthcare delivery organization.
Introduction to Healthcare OperationsThe introductory chapters provide an overview of the significant environmental trends healthcare delivery organizations face. Annual updates to industrywide trends can be found in Futurescan: Healthcare Trends and Implications 2016–2021 (SHSMD and ACHE 2016). Progressive organizations tend to review these publications care-fully, as they can use this information in response to external forces by identifying either new strategies or current operating problems that must be addressed.
Business has aggressively used operations improvement tools for the past 40 years, but the field of operations science actually began many centuries ago. Chapter 2 provides a brief history.
Healthcare operations are increasingly driven by the effects of EBM and pay for performance; chapter 3 offers an overview of these trends and how organizations can effect change to meet current challenges and opportunities.
Setting Goals and Executing StrategyA key component of effective operations is the ability to move strategy to action. Chapter 4 shows how the use of the balanced scorecard and strategy maps can help accomplish this aim. Change in all organizations is challenging, and the formal methods of project management (chapter 5) can deliver effec-tive, lasting improvements in an organization’s operations.
Payers wantto reducecosts for chemotherapy
New payment method for chemotherapy is created
Environment Organization Clinical microsystem Patient
Changes are made incare processes andsupport systems tomaintain qualitywhile reducing costs
Chemotherapy treatment needs to be more efficient to meet payment levels
EXHIBIT 1.4Linkages Within
the Healthcare System:
Chemotherapy
Chapter 1 : The Chal lenge and the Oppor tunity 13
Performance Improvement Tools, Techniques, and ProgramsOnce an organization has its strategy implementation and change management processes in place, it needs to select the correct tools, techniques, and programs to analyze current operations and develop effective adjustments.
Chapter 6 outlines the basic steps of problem solving, which begins by framing the question or problem and continues through data collection and analyses to enable effective decision making. Chapter 7 introduces the building blocks for many of the advanced tools used later in the book. (This chapter may serve as a review or reference for readers who already have good statistical skills.)
Closely related to statistical thinking is the emerging science of analyt-ics. With powerful new software tools and big data repositories, the ability to understand and predict organizational performance is significantly enhanced. Chapter 8 is new to this edition and presents several tools that have become available to healthcare analysts and leaders since publication of the second edition.
Some projects require a focus on process improvement. Six Sigma tools (chapter 9) can be used to reduce variability in the outcome of a process. Lean tools (chapter 10) help eliminate waste and increase speed.
Applications to Contemporary Healthcare Operations IssuesThis part of the book demonstrates how these concepts can be applied to some of today’s fundamental healthcare challenges. Process improvement techniques are now widely deployed in many organizations to significantly improve performance; chapter 11 reviews the tools of process improvement and demonstrates their use in improving patient flow.
Scheduling and capacity management continue to be major concerns for most healthcare delivery organizations, particularly with the advent of advanced-access scheduling, a concept promoted by the Institute for Healthcare Improve-ment and discussed in chapter 12. Specifically, the chapter demonstrates how
Setting goals and executing strategy
Performanceimprovementtools,techniques, and programs
Fundamentalhealthcareoperationsissues
High performance
EXHIBIT 1.5Framework for Effective Operations Management in Healthcare
Healthcare Operat ions Management14
simulation can be used to optimize scheduling. Chapter 13 explores the optimal methods for acquiring supplies and maintaining appropriate inventory levels. Chapter 14 outlines a systems approach to improving financial results, with a special emphasis on cost reduction—one of today’s most important challenges.
Putting It All Together for Operational ExcellenceIn the end, any operations improvement will fail unless steps are taken to maintain the gains; chapter 15 contains the necessary tools to do so. The chapter also provides a detailed algorithm that helps practitioners select the appropriate tools, methods, and techniques to effect significant operational improvements. It demonstrates how our fictionalized case study healthcare system, Vincent Valley Hospital and Health System (VVH), uses all the tools presented in the book to achieve operational excellence. In this way, a future is envisioned in which many of the tools and methods contained in the book are widely deployed in the US healthcare system.
Vincent Valley Hospital and Health SystemWoven throughout the chapters are examples featuring VVH, a fictitious but realistic health system. The companion website contains an expansive descrip-tion of VVH; here we provide some essential details.
VVH is located in a midwestern city with a population of 1.5 million. The health system employs 5,000 staff members, oper-ates 350 inpatient beds, and has a medical staff of 450 physicians. It operates nine clinics staffed by physicians who are employees of the system. VVH competes with
two major hospitals and an independent ambulatory surgery center that was formed by several surgeons from all three hospitals.
The VVH brand includes an accountable care organization to reflect the increased emphasis it has placed on population health in its community. The organization also is working to create a Medicare Advantage plan. It has significantly restructured its primary care delivery segment and has contracted with a variety of retail clinics to supplement the traditional office-based primary care physicians with whom it is affiliated. It recently added an online diagnosis and treatment service, with 24-hour telehealth now available.
Three major health plans provide most of the private payment to VVH, which, along with the state Medicaid system, have recently begun a pay-for-performance reimbursement initiative. VVH has a strong balance sheet and a profit margin of approximately 2 percent, but its senior leaders feel the orga-nization is financially challenged.
The board of VVH includes many local industry leaders, who have asked the chief executive to focus on using the operational techniques that have led them to succeed in their own businesses.
On the web at ache.org/books/OpsManagement3
Chapter 1 : The Chal lenge and the Oppor tunity 15
Conclusion
This book is an overview of operations management approaches and tools. The reader is expected to understand all the concepts in the book (and in current use in the field) and be able to apply, at the basic level, most of the tools, techniques, and programs presented. The reader is not expected to execute at the more advanced (e.g., Six Sigma black belt, project management professional) level. However, this book prepares readers to work effectively with knowledgeable professionals and, most important, enables them to direct the work of those professionals.
Final Note About the Third EditionPrior editions of this book included a chapter on simulation. Although simula-tion is a valuable tool in many industries, it is not used widely in healthcare, so the chapter was eliminated, with some of the principles of simulation moved to chapter 11. We hope the industry embraces this tool in the future—and then we will bring this chapter back.
Discussion Questions
1. Provide three examples of system improvements at the boundaries of the healthcare subsystems (patient, microsystem, organization, and environment).
2. Identify three systems in a healthcare organization (at any level) that have reinforcing feedback.
3. Identify three systems in a healthcare organization (at any level) that have balancing feedback.
4. Identify three systems in a healthcare organization (at any level) in which feedback delays affect the performance of the system.
References
Agency for Healthcare Research and Quality (AHRQ). 2010. “National Healthcare Quality Report, 2009: Key Themes and Highlights from the National Healthcare Qual-ity Report.” Last reviewed March. http://archive.ahrq.gov/research/findings/nhqrdr/nhqr09/Key.html.
Centers for Medicare & Medicaid Services (CMS). 2015. “National Health Expenditure Projections 2014-2025 Forecast Summary.” Published July 14. www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/National HealthExpendData/Downloads/Proj2015.pdf.
Healthcare Operat ions Management16
Institute of Medicine (IOM). 2011. Learning What Works: Infrastructure Required for Comparative Effectiveness Research. Workshop Summary. Accessed August 8, 2016. www.nap.edu/catalog/12214/learning-what-works-infrastructure-required-for-comparative-effectiveness-research-workshop.
———. 2001. Crossing the Quality Chasm: A New Health System for the 21st Century. Wash-ington, DC: National Academies Press.
———. 1999. To Err Is Human: Building a Safer Health System. Washington, DC: National Academies Press.
Joshi, M. S., E. R. Ransom, D. B. Nash, and S. B. Ransom. 2014. The Healthcare Quality Book: Vision, Strategy and Tools, 3rd edition. Chicago: Health Administration Press.
Landrigan, C. P., G. J. Parry, C. B. Bones, A. D. Hackbarth, D. A. Goldmann, and P. J. Sharek. 2010. “Temporal Trends in Rates of Patient Harm Resulting from Medical Care.” New England Journal of Medicine 363 (22): 2124–34.
National Guideline Clearinghouse (NGC). 2016. Home page. Accessed August 8. https://guideline.gov/.
Ransom, S. B., M. S. Joshi, and D. B. Nash (eds.). 2005. The Healthcare Quality Book: Vision, Strategy, and Tools. Chicago: Health Administration Press.
Rotter, T., L. Kinsman, E. L. James, A. Machotta, H. Gothe, J. Willis, P. Snow, and J. Kugler. 2010. “Clinical Pathways: Effects on Professional Practice, Patient Outcomes, Length of Stay and Hospital Costs.” Cochrane Database of Systematic Reviews 3: CD006632.
Senge, P. M. 1990. The Fifth Discipline: The Art and Practice of the Learning Organization. New York: Doubleday.
Society for Healthcare Strategy and Market Development (SHSMD) and American Col-lege of Healthcare Executives (ACHE). 2016. Futurescan: Healthcare Trends and Implications 2016–2021. Chicago: SHSMD and Health Administration Press.
CHAPTER
17
2HISTORY OF PERFORMANCE IMPROVEMENT
Operations Management in Action
During the Crimean War, a conflict that waged from October 1853 to February 1856 pitting Russia against Britain, France, and Ottoman Turkey, reports of ter-rible conditions in military hospitals began to emerge that alarmed British citizens. In response to the out-cry, the British government commissioned Florence Nightingale, now widely recognized as a pioneer in nursing practice, to oversee the introduction of nurses to military hospitals and to improve conditions in the hospitals. When Nightingale arrived in Scutari, Turkey, she found the military hospital there overcrowded and filthy. She instituted many changes to improve the sanitary conditions in the hospital, and many lives were saved as a result of these reforms.
Nightingale was among the first healthcare professionals to collect, tabulate, interpret, and graph-ically display data related to the impact of process changes on care outcomes—what is known today as evidence-based medicine. To quantify the overcrowd-ing problem, she compared the average amount of space per patient in London hospitals—1,600 square feet—to the space in Scutari—about 400 square feet. She developed a standardized document, the Model Hospital Statistical Form, to enable the collection of consistent data for analysis and comparison. In Feb-ruary 1855, the patient mortality rate at the military hospital in Scutari was 42 percent. As a result of Night-ingale’s changes, by June of that year the mortality rate had decreased to 2.2 percent.
To present these data in a persuasive manner, she developed a new type of graphic display, the polar area diagram. The diagram was a pie chart with a monthly slice for mortality numbers and their causes displayed in a different color. A quick glance at the diagram “showed that except for the bloodiest month in the siege of Sevastopol, battle deaths take up a very small portion of each slice,” notes Lienhard
OVE RVI EW
This chapter provides the background and historical
context for performance improvement—which is not
a new concept. Several of the tools, techniques, and
philosophies outlined in this text are based in past
efforts. Although the terminology has changed, many
of the core concepts remain the same.
The major topics in this chapter include the
following:
• Background for understanding operations
management
• Systems thinking and knowledge-based
management
• Scientific management
• Project management
• Introduction to quality, and quality experts of
note
• Philosophies of performance improvement,
including Six Sigma, Lean, and others
• Introduction to supply chain management
• Introduction to big data and analytics
Although these tools and techniques have been
adapted for contemporary healthcare, their roots
are in the past, and an understanding of this history
(exhibit 2.1) can enable organizations to move success-
fully into the future.
Healthcare Operat ions Management18
(2016). It revealed that “The Russians were a minor enemy. The real enemies were cholera, typhus, and dysentery. Once the military looked at that eloquent graph, the modern army hospital system was inevitable” (Lienhard 2016).
After the war, Nightingale used the data she had collected to demonstrate that the mortality rate in Scutari following her reforms was significantly lower than in other British military hospitals. Although the British military hierarchy was resis-tant to her changes, the data were convincing and resulted in reforms to military hospitals and the establishment of the Royal Commission on the Health of the Army.
Were she alive today, Nightingale would recognize many of the philosophies, tools, and techniques outlined in this text as essentially the same as those she employed to achieve lasting reform in hospitals throughout the world.
Sources: Information from Cohen (1984), Lienhard (2016), Neuhauser (2003), and Nightingale (1858).
Background
The healthcare industry faces many challenges. The costs of care and level of services delivered are increasing; even as the population ages, we are able to pro-long lives to an ever greater extent as technology advances and expertise grows. The expectation of quality care with zero defects, or failures in care, is being pursued by government and other stakeholders, driving the need for healthcare providers to produce more of a high-quality product or service at a reduced cost. This need can only be met through improved utilization of resources.
Specifically, providers must offer their services more effectively and effi-ciently than at any time in the past by optimizing their use of limited financial assets, employees and staff, machines and facilities, and time.
Enter operations management.Operations management is the design, implementation, and improve-
ment of the processes and systems that create and deliver the organization’s products and services. Operations managers plan and control delivery processes and systems within the organization.
Forward-thinking healthcare leaders and professionals have realized that the theories, tools, and techniques of operations management, if properly applied, can enable their organizations to become efficient and effective care delivery environments. However, for many of the aims identified by the US healthcare system to be achieved, essentially all healthcare providers must adopt these tools and techniques, many of which have enabled other service indus-tries and manufacturing sectors to improve efficiency and effectiveness. The operations management information presented in this book should similarly enable hospitals and other healthcare organizations to design systems, processes, products, and services that meet the needs of their stakeholders. Importantly, it should also allow continuous improvement in these systems and services to keep pace with the quickly changing healthcare landscape.
Chapter 2: History of Performance Improvement 19
CPM
met
hod
PERT
met
hod
Proj
ect M
anag
emen
tIn
stit
ute
Hen
ry G
antt
,G
antt
cha
rts
1925
1300
1900
1950
1975
2000
100K
Live
s
SCM
IOM
repo
rtTo
Err
Is H
uman
,B
aldr
ige
Awar
din
Hea
lthc
are,
AHRQ
Six
Sigm
a
Bal
drig
eAw
ard,
ISO
9000
TQM
,JI
T,Av
edis
Don
abed
ian
Inst
itut
e fo
r Hea
lthc
are
Impr
ovem
ent;
Jam
es W
omac
k,Th
e M
achi
ne th
atCh
ange
d th
e W
orld
;Ro
bert
S. K
apla
n,ba
lanc
ed s
core
card
W. E
dwar
ds D
emin
g(U
S)
Har
lan
Clev
elan
d,kn
owle
dge
hier
arch
y
Kaor
u ls
hika
wa,
TQM
, fis
hbon
e
If Ja
pan
Can
. . .
Why
Can
’t W
e?
Shig
eo S
hing
o,po
ka-y
oke
and
SMED
Eliy
ahu
M. G
oldr
att,
TOC
Rus
sell
L. A
ckof
f,sy
stem
s th
inki
ng
Gen
ichi
Tag
uchi
,co
st o
f var
iati
on
W. E
dwar
ds D
emin
g,fa
ther
of
qual
ity
mov
emen
t(J
apan
)
Wal
ter A
. She
wha
rt,
gran
dfat
her o
f qu
alit
y m
ovem
ent
Hen
ry F
ord,
mas
s pr
oduc
tion
Jose
ph M
. Jur
an,
qual
ity
trilo
gy
TPS
Fran
k an
d Li
llian
Gilb
reth
,ti
me
and
mot
ion
Fred
eric
k W
insl
ow T
aylo
r,fa
ther
of
scie
ntifi
c m
anag
emen
t A. E
rlan
g,qu
euei
ng
Veni
ce A
rsen
al,
first
mov
ing
asse
mbl
y lin
e
Adam
Sm
ith,
spec
ializ
atio
n of
labo
rFl
oren
ceN
ight
inga
le
EXH
IBIT
2.1
Impo
rtan
t Eve
nts
in P
erfo
rman
ce Im
prov
emen
t
Healthcare Operat ions Management20
To improve systems and processes, however, one must first know the system or process and its desired inputs and outputs.
Knowledge-Based Management
This book takes a systems view of service provision and delivery, as illustrated in exhibit 2.2, and focuses on knowledge-based management (KBM)—using data and information toward basing management decisions on facts rather than on feelings or intuition—to frame that view. The improvement in computer systems and new analytical approaches support the increased use of KBM, especially in terms of building a knowledge hierarchy.
The knowledge hierarchy relates to the learning that ultimately under-pins KBM. As illustrated in exhibit 2.3, the knowledge hierarchy consists of the following five categories (Zeleny 1987):
Knowledge hierarchyThe foundation of knowledge-based management, composed of five categories of learning: data, information, knowledge, understanding, and wisdom.
Feedback
Transformationprocess
LaborMaterialMachinesManagementCapital
Goods orservices
TUPTUOTUPNI
EXHIBIT 2.2Systems View
of the Provision of Services for
Purposes of This Book
Impo
rtan
ce
Understanding
Wisdom
morals
principles
patterns
relationships
Knowledge
Learning
Information
Data
EXHIBIT 2.3Knowledge
Hierarchy
Chapter 2: History of Performance Improvement 21
1. Data. Symbols or raw numbers that simply exist; they have no structure or organization. Entities collect data with their computer systems; individuals collect data through their experiences. At this stage of the hierarchy, one can presume to know nothing because raw data alone are not adequate for decision making.
2. Information. Data that are organized or processed to have meaning. Information can be useful, but it is not necessarily useful. It can answer such questions as who, what, where, and when—in other words, know what.
3. Knowledge. Information that is deliberately useful. Knowledge enables decision making—know how.
4. Understanding. A mental frame that allows use of what is known and enables the development of new knowledge. Understanding represents the difference between learning and memorizing—know why.
5. Wisdom. A high-level stage that adds moral and ethical views to understanding. Wisdom answers questions to which there is no known correct answer and, in some cases, to which there will never be a known correct answer—know right.
A simple example may help explain this hierarchy. Say your height is 67 inches and your weight is 175 pounds (data). You have a body mass index (BMI) of 26.7 (information). A healthy BMI is 18.5 to 25.5 (knowledge). Your BMI is high, and to be healthy you should lower it (understanding). You begin a diet and exercise program and lower your BMI (wisdom).
Finnie (1997, 24) summarizes the relationships in the hierarchy and notes our tendency to focus on its less important levels:
We talk about the accumulation of information, but we fail to distinguish between
data, information, knowledge, understanding, and wisdom. An ounce of information
is worth a pound of data, an ounce of knowledge is worth a pound of information,
an ounce of understanding is worth a pound of knowledge, an ounce of wisdom is
worth a pound of understanding. In the past, our focus has been inversely related to
importance. We have focused mainly on data and information, a little bit on knowl-
edge, nothing on understanding, and virtually less than nothing on wisdom.
Knowledge Through the AgesThe roots of the knowledge hierarchy can be traced to eighteenth-century philosopher Immanuel Kant, much of whose work attempted to address the questions of what and how we can know.
The two major philosophical movements that significantly influenced Kant were empiricism and rationalism (McCormick 2006). The empiricists, most notably John Locke, argued that human knowledge originates in one’s
Healthcare Operat ions Management22
experiences. According to Locke, the mind is a blank slate that fills with ideas through its interaction with the world. The rationalists, including Descartes and Galileo, argued that the world is knowable through an analysis of ideas and logical reasoning. Both the empiricists and the rationalists viewed the mind as passive, either by receiving ideas onto a blank slate or because it possesses innate ideas that can be logically analyzed.
Kant joined these philosophical ideologies by arguing that experience leads to knowing only if the mind provides a structure for those experiences. Although the idea that the rational mind plays a role in defining reality is now common, in Kant’s time this was a major insight into what and how we know. Knowledge does not flow from our experiences alone, nor only from our ability to reason; rather, knowledge flows from our ability to apply reasoning to our experiences.
Relating Kant’s philosophy to the knowledge hierarchy, data are our experiences, information is obtained through logical reasoning, and knowledge is obtained when we apply structured reasoning to data to acquire knowledge (Ressler and Ahrens 2006).
The intent of this text is to enable readers to gain knowledge. We discuss tools and techniques that allow the application of logical reasoning to data toward obtaining knowledge and using it to make decisions. This knowledge and understanding should help the reader provide healthcare in an efficient and effective manner.
History of Scientific Management
Frederick Taylor (whose work is covered in more detail later in the chapter) originated the term scientific management in The Principles of Scientific Man-agement (Taylor 1911). Scientific management methods called for eliminating the old rule-of-thumb, individual way of performing work and, through study and optimization of the work, replacing the varied methods with the one “best” way of performing the work to improve productivity and efficiency. Today, the term scientific management has been replaced with operations management, but the concept is similar: Study the process or system and determine ways to optimize it to achieve improved efficiency and effectiveness.
Mass ProductionThe Industrial Revolution and mass production set the stage for much of Tay-lor’s work. Prior to the Industrial Revolution, individual craftsmen performed all tasks necessary to produce a good using their own tools and procedures. In the eighteenth century, Adam Smith advocated for the division of labor—increasing work efficiency through specialization. To support a division of labor, a large number of workers are brought together, and each performs a specific task related to the production of a good. Thus, the factory system of
Scientific managementA disciplined approach to studying a system or process and then using data to optimize it to achieve improved efficiency and effectiveness.
Chapter 2: History of Performance Improvement 23
mass production was born, and Henry Ford’s assembly line eventually emerged, making industrial conditions ripe for Taylor to introduce scientific management.
Mass production allows for significant economies of scale, as predicted by Smith. Before Ford set up his moving assembly line, each car chassis was assembled by a single worker and took about 12½ hours to produce. After the introduction of the assembly line, this time was reduced to 93 minutes (Bellis 2006). The standardization of products and work ushered in by the assembly line not only led to a reduction in the time needed to produce cars but also significantly reduced the costs of production. The selling price of the Model T fell from $1,000 to $360 between 1908 and 1916 (Simkin 2005), allowing Ford to capture a large portion of the market.
Although Ford is commonly credited with introducing the moving assembly line and mass production in modern times, both processes were in practice several hundred years earlier. The Venetian Arsenal of the 1500s employed 16,000 people and produced nearly one ship every day (NationMas-ter.com 2004). Ships were mass produced using premanufactured, standardized parts on a floating assembly line (Schmenner 2001).
One of the first examples of mass production in the healthcare industry is Shouldice Hospital (Heskett 2003). Much like Ford, who is commonly cited as saying people could have the Model T in any color, “so long as it’s black,” Shouldice, founded in 1945 in Toronto, performs just one type of surgery—routine hernia operations—and it continues to thrive with its unique approach (Heskett 2003).
Furthermore, evidence is growing in healthcare that level of experience in treating specific illnesses and conditions affects the outcome of that care. Higher volumes of cases often result in better outcomes (Halm, Lee, and Chassin 2002). Specifically, the additional practice associated with higher volume results in bet-ter outcomes. The idea of “practice makes perfect,” or learning-curve effects, has led organizations such as the Leapfrog Group (made up of organizations that provide healthcare benefits) to list patient volume among its criteria for quality (Halm, Lee, and Chassin 2002). The Agency for Healthcare Research and Quality (AHRQ) report Localizing Care to High-Volume Centers devotes an entire chapter to this issue and its impact on medical practice (Auerbach 2001).
Frederick TaylorTaylor began his work when mass production and the factory system were in their infancy. He believed that US industry was “wasting” human effort and that, as a result, national efficiency (now called productivity) was significantly lower than it could be. The introduction to The Principles of Scientific Manage-ment (Taylor 1911) illustrates his intent:
[O]ur larger wastes of human effort, which go on every day through such of our acts
as are blundering, ill-directed, or inefficient, and which Mr. [Theodore] Roosevelt
Healthcare Operat ions Management24
refers to as a lack of “national efficiency,” are less visible, less tangible, and are but
vaguely appreciated. . . . This paper has been written:
First. To point out, through a series of simple illustrations, the great loss which the
whole country is suffering through inefficiency in almost all of our daily acts.
Second. To try to convince the reader that the remedy for this inefficiency lies in
systematic management, rather than in searching for some unusual or extraordinary
man [referring to the so-called great man theory prevalent at the time].
Third. To prove that the best management is a true science, resting upon clearly
defined laws, rules, and principles, as a foundation. And further to show that the
fundamental principles of scientific management are applicable to all kinds of human
activities, from our simplest individual acts to the work of our great corporations,
which call for the most elaborate cooperation. And, briefly, through a series of illus-
trations, to convince the reader that whenever these principles are correctly applied,
results must follow which are truly astounding.
Note that Taylor specifically mentions systems management as opposed to the individual; this is a common theme that we revisit throughout this book. Rather than focusing on individuals as the cause of problems and the source of solutions, emphasis is placed on systems and their optimization.
Taylor believed that much waste was the result of what he called “sol-diering,” which today might be thought of as slacking. Further, he believed that the underlying causes of soldiering were as follows (Taylor 1911):
First. The fallacy, which has from time immemorial been almost universal among
workmen, that a material increase in the output of each man or each machine in
the trade would result in the end in throwing a large number of men out of work.
Second. The defective systems of management which are in common use, and which
make it necessary for each workman to soldier, or work slowly, in order that he may
protect his own best interests.
Third. The inefficient rule-of-thumb methods, which are still almost universal in all
trades, and in practicing which our workmen waste a large part of their effort.
To eliminate soldiering, Taylor proposed instituting incentive schemes. While at Midvale Steel Company, he used time studies to set daily production quotas. Incentives were paid to those workers who reached their daily goals, and those who did not reach their goals were paid significantly less. Productiv-ity at Midvale doubled. Not surprisingly, Taylor’s ideas produced considerable backlash. The resistance to increasingly popular pay-for-performance programs in healthcare today is analogous to that experienced by Taylor.
Taylor believed that “one best way” existed to perform any task and that careful study and analysis would lead to the discovery of that way. For
Chapter 2: History of Performance Improvement 25
example, while at Bethlehem Steel Corporation, he studied the shoveling of coal. Using time studies and a careful analysis of how the work was performed, he determined that the optimal amount of coal per shovel load was 21 pounds. Taylor then developed shovels that would hold exactly 21 pounds for each type of coal; workers had previously supplied their own shovels (NetMBA.com 2005). He also determined the ideal work rate and rest periods to ensure that workers could shovel all day without fatigue. As a result of Taylor’s improved methods, Bethlehem Steel was able to reduce the number of workers shoveling coal from 500 to 140 (Nelson 1980).
Taylor’s four principles of scientific management are to
1. develop and standardize work methods on the basis of scientific study, and use these to replace individual rule-of-thumb methods;
2. select, train, and develop workers rather than allowing them to choose their own tasks and train themselves;
3. develop a spirit of cooperation between management and workers to ensure that the scientifically developed work methods are both sustainable and implemented on a continuing basis; and
4. divide work between management and workers so that each has an equal share, where management plans the work and workers perform the work.
Although some would be problematic today—particularly the notion that workers are “machinelike” and motivated solely by money—many of Taylor’s ideas can be seen in the foundations of newer initiatives such as Six Sigma and Lean, two important quality improvement approaches discussed in depth later in the book.
Frank and Lillian GilbrethThe Gilbreths were contemporaries of Frederick Taylor. Frank, who worked in the construction industry, noticed that no two bricklayers performed their tasks the same way. He believed that bricklaying could be standardized and the one best way determined. He studied the work of bricklaying and analyzed the workers’ motions, finding much unnecessary stooping, walking, and reaching. He eliminated these motions by developing an adjustable scaffold designed to hold both bricks and mortar (Taylor 1911). As a result of this and other improvements, Frank Gilbreth reduced the number of motions in bricklaying from 18 to 5 (International Work Simplification Institute 1968) and raised out-put from 1,000 to 2,700 bricks a day (Perkins 1997). He applied what he had learned from his bricklaying experiments to other industries and types of work.
In his study of surgical operations, Frank Gilbreth found that doctors spent more time searching for instruments than performing the surgery. He
Healthcare Operat ions Management26
developed a technique still seen in operating rooms today: When the doctor needs an instrument, he extends his hand, palm up, and asks for the instru-ment, which is then placed in his hand. This technique eliminates searching for the instrument and allows the doctor to stay focused on the surgical area, thus reducing surgical time (Perkins 1997).
Frank and Lillian Gilbreth may be more familiarly known as the parents in the book Cheaper by the Dozen (Gilbreth and Carey 1948) (which was made into a movie by the same title in 1950 and remade in 2003). The Gilbreths incorporated many of their time-saving ideas in their family as well. For example, they bought just one type of sock for all 12 of their children, thus eliminating time-consuming sorting.
Scientific Management TodayScientific management fell out of favor during the Depression, partly because of the sense that it dehumanized employees, but mainly because of a general belief in society that productivity improvements resulted in downsizing and increased unemployment. Not until World War II did scientific management, renamed operations research, see a resurgence of interest.
In healthcare today, standardized methods and procedures are used to reduce costs and increase the quality of outcomes. Specialized equipment has been developed to speed procedures and reduce labor costs. In a sense, we are still searching for the one best way. However, we must heed the lessons of the past. If the tools of operations management are perceived to be dehumanizing or to result in downsizing by healthcare organizations, their implementation will meet significant resistance.
Project Management
The discipline of project management began with the development of the Gantt chart in the early twentieth century. Henry Gantt worked closely with Frederick Taylor at Midvale Steel and in Navy ship construction during World War I. From this work, he developed bar graphs to illustrate the duration of project tasks and display scheduled and actual progress. These Gantt charts were used to help manage large projects, including construction of the Hoover Dam, and proved to be such a powerful tool that they are commonly used today.
Although Gantt charts were originally adopted to track large projects, they are not ideal for very large, complicated projects because they do not explicitly show precedence relationships, that is, what tasks need to be completed before other tasks can start. In the 1950s, two mathematic project scheduling techniques were developed: the program evaluation and review technique (PERT) and the critical path method (CPM). Both techniques begin by developing a project network showing the precedence relationships among tasks and task duration.
Program evaluation and review technique (PERT)A graphic technique to link and analyze all tasks within a project; the resulting graph helps optimize the project’s schedule.
Critical path method (CPM)The critical path is the longest course through a graph of linked tasks in a project. The critical path method is used to reduce the total time of a project by decreasing the duration of tasks on the critical path.
Chapter 2: History of Performance Improvement 27
PERT was developed by the US Navy to address the desire to acceler-ate the Polaris missile program. This “need for speed” was precipitated by the Soviet launch of Sputnik, the first space satellite. PERT uses a probability distribution (the beta distribution), rather than a point estimate, for the dura-tion of each project task. The probability of completing the entire project in a given amount of time can then be determined. This technique is most useful for estimating project completion time when task times are uncertain and for evaluating risks to project completion prior to the start of a project.
The CPM technique was developed at the same time as PERT by the DuPont and Remington Rand corporations to manage plant maintenance projects. CPM uses the project network and point estimates of task duration times to determine the critical path through the network, or the sequence of activities that will take the longest to complete. If any one of the activities on the critical path is delayed, the entire project is delayed. This technique is most useful when task times can be estimated with certainty and is typically used in project management and control.
Although both of these techniques are powerful analytical tools for planning, implementing, controlling, and evaluating a project plan, perform-ing the required calculations by hand is tedious, and use of the techniques was not initially widespread. With the advent of commercially available project management software for personal computers in the late 1960s, use of PERT and CPM increased considerably. Today, numerous project management soft-ware packages are commercially available. Microsoft Project, for instance, can perform network analysis on the basis of either PERT or CPM; the default is CPM, making it the more commonly used technique.
Projects are an integral part of many of the process improvement ini-tiatives found in the healthcare industry. Project management and its tools are needed to ensure that projects related to quality, Lean, and supply chain management are completed in the most effective and timely manner possible.
Introduction to Quality
Any discussion of quality in industry—including healthcare—should begin with those recognized as originators in quality improvement methodology. Here we introduce the individuals credited with developing various quality approaches, and later in the section we discuss some prevailing quality improve-ment processes. This introductory discussion establishes the background for the in-depth treatment of the concepts throughout the book.
Walter ShewhartIf W. Edwards Deming and Joseph Juran (profiled in later subsections) are considered the fathers of the quality movement, Walter Shewhart may be seen
Healthcare Operat ions Management28
as its grandfather. Both Deming and Juran studied under Shewhart, and much of their work was influenced by his ideas.
Shewhart believed that managers need certain information to enable them to make scientific, efficient, and economical decisions. He developed statistical process control (SPC) charts to supply that information (Shewhart 1931). He also believed that management and production practices need to be continu-ously evaluated, and then adopted or rejected on the basis of this evaluation, if an organization hopes to evolve and survive. Deming’s cycle of improvement, known as plan-do-check-act (PDCA) (sometimes rendered as plan-do-study-act), was adapted from Shewhart’s work (Shewhart and Deming 1939).
W. Edwards DemingDeming was an employee of the US government in the 1930s and 1940s, work-ing with statistical sampling techniques. He became a supporter and student of Shewhart, believing Shewhart’s techniques could be useful in nonmanufactur-ing environments. Deming applied SPC methods to his work at the National Bureau of the Census to improve clerical operations in preparation for the 1940 population census. As a result, in some cases productivity improved by a factor of six (Kansal and Rao 2006).
Deming taught seminars to bring his and Shewhart’s work to US and Canadian organizations, where major reductions in scrap and rework resulted. However, after World War II, Deming’s ideas lost popularity in the United States, mainly because demand for all products was so great that quality became unimportant; any product, regardless of how well it was made, was snapped up by hungry consumers.
After the war, Deming traveled to Japan as an adviser for that country’s census. While he was there, the Union of Japanese Scientists and Engineers invited him to lecture on quality control techniques, and Deming brought his message to Japanese executives: Improving quality reduces expenses while increasing productivity and market share. During the 1950s and 1960s, Deming’s ideas were widely known and implemented in Japan, but not in the United States.
The energy crisis of the 1970s was the turning point. In part as a result of oil shortages, the small, well-built Japanese automobiles increased in popular-ity, and the US auto industry saw declines in demand, setting the stage for the return of Deming’s ideas. The 1980 television documentary If Japan Can . . . Why Can’t We?, investigating the increasing competition that numerous US industries faced from Japan, made Deming and his quality ideas known to a broad audience. Much like the Institute of Medicine report To Err Is Human (1999) increased awareness of the need for quality in healthcare, this documen-tary drove US industry’s attention to the need for quality in manufacturing.
Deming’s quality ideas reflected his statistical background, but his expe-rience in their implementation prompted him to expand his approach. He instructed managers in the two types of variation—special cause, resulting from
Statistical process control (SPC)A scientific approach to controlling the performance of a process by measuring the process outputs and then using statistical tools to determine whether this process is meeting expected performance.
Plan-do-check-act (PDCA)A core process improvement tool with four elements: Plan a change to a process, enact the change, check to make sure it is working as expected, and act to make sure the change is sustainable. PDCA functions as a continuous cycle and, as such, is sometimes referred to as the Deming wheel.
Chapter 2: History of Performance Improvement 29
a change in the system that can be identified or assigned and the problem fixed, and common cause, deriving from the natural differences in the system that cannot be eliminated without changing the system. Although identifying the common causes of variation is possible, these causes cannot be fixed without the authority and ability to improve the system, for which management is typically responsible.
Moving far beyond SPC, Deming’s quality methods include a systematic approach to problem solving and continuous process improvement with his PDCA cycle. He also believed that management is ultimately responsible for quality and must actively support and encourage quality “transformations” in organizations. In the preface to Out of the Crisis, Deming (1986) writes:
Drastic changes are required. The first step in the transformation is to learn how to
change. . . . Long term commitment to new learning and new philosophy is required
of any management that seeks transformation. The timid and the faint-hearted, and
people that expect quick results are doomed to disappointment. Whilst the intro-
duction of statistical problem solving and quality techniques and computerization
and robotization have a part to play, this is not the solution: Solving problems, big
problems and little problems, will not halt the decline of American industry, nor will
expansion in use of computers, gadgets, and robotic machinery.
Benefits from massive expansion of new machinery also constitute a vain
hope. Massive immediate expansion in the teaching of statistical methods to pro-
duction workers is not the answer either, nor wholesale flashes of quality control
circles. All these activities make their contribution, but they only prolong the life of
the patient, they cannot halt the decline. Only transformation of management and
of Government’s relations with industry can halt the decline.
Out of the Crisis contains Deming’s famous 14 points for management. Although not as well known, he also included an adaptation of the 14 points for medical services (exhibit 2.4), which he attributed to Drs. Paul B. Batalden and Loren Vorlicky of the Health Services Research Center in Minneapolis (Deming 1986).
1. Establish constancy of purpose toward service.a. Define in operational terms what you mean by “service to patients.”b. Specify standards of service for a year hence and for five years hence.c. Define the patients whom you are seeking to serve.d. Constancy of purpose brings innovation.e. Innovate for better service.f. Put resources into maintenance and new aids to production.
g. Decide whom the administrators are responsible to and the means by which they will be held responsible.
EXHIBIT 2.4Deming’s Adaptation of the 14 Points for Medical Service
(continued)
Healthcare Operat ions Management30
h. Translate this constancy of purpose to service to patients and the community.
i. The board of directors must hold onto the purpose.
2. Adopt the new philosophy. We are in a new economic age. We can no lon-ger live with commonly accepted levels of mistakes, materials not suited to the job, people on the job who do not know what the job is and are afraid to ask, failure of management to understand their job, antiquated methods of training on the job, and inadequate and ineffective supervi-sion. The board must put resources into this new philosophy, with com-mitment to in-service training.
3. a. Require statistical evidence of quality of incoming materials, such as pharmaceuticals. Inspection is not the answer. Inspection is too late and is unreliable. Inspection does not produce quality. The quality is already built in and paid for. Require corrective action, where needed, for all tasks that are performed in the hospital.
b. Institute a rigid program of feedback from patients in regard to their satisfaction with services.
c. Look for evidence of rework or defects and the cost that may accrue.
4. Deal with vendors that can furnish statistical evidence of control. We must take a clear stand that price of services has no meaning without adequate measure of quality. Without such a stand for rigorous mea-sures of quality, business drifts to the lowest bidder, low quality and high cost being the inevitable result.
Requirement of suitable measures of quality will, in all likelihood, require us to reduce the number of vendors. We must work with vendors so that we understand the procedures that they use to achieve reduced numbers of defects.
5. Improve constantly and forever the system of production and service.
6. Restructure training.a. Develop the concept of tutors.b. Develop increased in-service education.c. Teach employees methods of statistical control on the job.d. Provide operational definitions of all jobs.e. Provide training until the learner’s work reaches the state of statisti-
cal control.
7. Improve supervision. Supervision is the responsibility of the management.a. Supervisors need time to help people on the job.b. Supervisors need to find ways to translate the constancy of purpose
to the individual employee.c. Supervisors must be trained in simple statistical methods with the
aim to detect and eliminate special causes of mistakes and rework.d. Focus supervisory time on people who are out of statistical control
and not those who are low performers. If the members of a group are
EXHIBIT 2.4Deming’s
Adaptation of the 14 Points for Medical Service (continued from previous page)
(continued)
Chapter 2: History of Performance Improvement 31
The New Economics for Industry, Government, Education (Deming 1994) outlines the Deming System of Profound Knowledge. Deming believed that to transform organizations, the individuals in those organizations need to understand the four parts of this system.
1. Appreciation for a system: Everything is related to everything else, and those inside the system need to understand the relationships in it.
2. Knowledge about variation: This part of the system refers to what can and cannot be done to decrease either of the two types of variation.
in fact in statistical control, there will be some low performers and some high performers.
e. Teach supervisors how to use the results of surveys of patients.
8. Drive out fear. We must break down the class distinctions between types of workers within the organization—physicians, nonphysicians, clinical providers versus nonclinical providers, physician to physician. Discon-tinue gossip. Cease to blame employees for problems of the system. Man-agement should be held responsible for faults of the system. People need to feel secure to make suggestions. Management must follow through on suggestions. People on the job cannot work effectively if they dare not offer suggestions for simplification and improvement of the system.
9. Break down barriers between departments. One way would be to encour-age switches of personnel in related departments.
10. Eliminate numerical goals, slogans, and posters imploring people to do better. Instead, display accomplishments of the management in respect to helping employees improve their performance.
11. Eliminate work standards that set quotas. Work standards must produce quality, not mere quantity. It is better to take aim at rework, error, and defects.
12. Institute a massive training program in statistical techniques. Bring statisti-cal techniques down to the level of the individual employee’s job, and help him to gather information about the nature of his job in a systematic way.
13. Institute a vigorous program for retraining people in new skills. People must be secure about their jobs in the future and must know that acquir-ing new skills will facilitate security.
14. Create a structure in top management that will push every day on the previous 13 points. Top management may organize a task force with the authority and obligation to act. This task force will require guidance from an experienced consultant, but the consultant cannot take on obligations that only the management can carry out.
EXHIBIT 2.4Deming’s Adaptation of the 14 Points for Medical Service (continued from previous page)
Source: Full credit and proper copyright notice must be given for material used. Please credit as follows: Deming, W. Edwards, Out of the Crisis, pp. 199–203, © 2000 Massachusetts Institute of Technology, by permission of The MIT Press.
Healthcare Operat ions Management32
3. Theory of knowledge: The theory highlights the need for understanding and knowledge rather than information.
4. Knowledge of psychology: People are intrinsically motivated and different from one another, and attempts to use generic extrinsic motivators can result in unwanted outcomes.
Deming’s 14 points and System of Profound Knowledge still provide a road map for organizational transformation.
Joseph M. JuranJuran was a contemporary of Deming and a student of Shewhart. He began his career at the Western Electric Hawthorne Works plant, the site of the famous Hawthorne studies (Mayo 1933) related to worker motivation. Western Electric had close ties to Bell Telephone, Shewhart’s employer, because the company was the sole supplier of telephone equipment to Bell.
During World War II, Juran served as assistant administrator for the Lend-Lease Administration. Juran’s quality improvement techniques made him instrumental in improving the efficiency of processes by eliminating unnecessary paperwork and ensuring the timely arrival of supplies to US allies.
Juran’s Quality Handbook (Juran and Godfrey 1998) was first published in 1951 and remains a standard reference for quality. Juran was among the first quality experts to define quality from the customer perspective as “fitness for use.”
His contributions to quality include the adaptation of the Pareto prin-ciple to the quality arena (see chapter 9 for its application in quality improve-ment). According to this principle, 80 percent of defects are caused by 20 percent of problems, and quality improvement should therefore focus on the “vital few” to gain the most benefit. The roots of Six Sigma programs can be seen in Juran’s (1986) quality trilogy, shown in exhibit 2.5.
Avedis DonabedianAvedis Donabedian was born in 1919 in Beirut, Lebanon, and received a medical degree from the American University of Beirut. In 1955, he earned a master’s degree in public health from Harvard University. While a student at Harvard, Donabedian wrote a paper on quality assessment that brought his work to the attention of various experts in the field of public health. He taught for a short period at New York Medical College before becoming a faculty member at the School of Public Health of the University of Michigan, where he stayed for the remainder of his career.
Shortly after Donabedian joined the University of Michigan faculty, the US Public Health Service began a project looking at the entire field of health services research, for which Donabedian was asked to review and evalu-ate the literature on quality assessment. This work culminated in his famous
Pareto principleDeveloped by Italian economist Vilfredo Pareto in 1906 on the basis of his observation that 80 percent of the wealth in Italy was owned by 20 percent of the population.
Chapter 2: History of Performance Improvement 33
article, “Evaluating the Quality of Medical Care” (Donabedian 1966), followed by a three-volume book series, titled Exploration in Quality Assessment and Monitoring (Donabedian 1980, 1982, 1985). Over the course of his career, Donabedian wrote 16 books and more than 100 articles on quality assessment and improvement in the healthcare sector on such topics as the definition of quality in healthcare, the relationship between outcomes and process, the impact of clinical decisions on quality, the effectiveness of quality programs, and the relationship between quality and cost (Sunol 2000).
Donabedian (1980) defined healthcare quality in terms of efficacy, effi-ciency, optimality, adaptability, legitimacy, equality, and cost. He was among the first quality researchers to view healthcare as a system composed of structure, process, and outcome, providing a framework for health services research still used today (Donabedian 1966). He also highlighted many of the issues that arise when attempting to measure structures, processes, and outcomes.
Basic Quality Processes
Quality Planning • Identify the customers, both external and internal.• Determine customer needs.• Develop product features that respond to customer.• Establish quality goals that meet the needs of custom-
ers and suppliers alike, and do so at a minimum com-bined cost.
• Develop a process that can produce the needed product features.
• Prove the process capability—prove that the process can meet quality goals under operating conditions.
Control • Choose control subjects—what to control.• Choose units of measurement.• Establish measurement.• Establish standards of performance.• Measure actual performance.• Interpret the difference (actual versus standard).• Take action on the difference.
Improvement • Prove the need for improvement.• Identify specific projects for improvement.• Organize to guide the projects.• Organize for diagnosis—for discovery of causes.• Diagnose to find the causes.• Provide remedies.• Prove that the remedies are effective under operating
conditions.• Provide for control to hold the gains.
Source: Juran, J. M. 1986. “The Quality Trilogy.” Quality Progress 19 (8): 19–24. Reprinted with per-mission from Juran Institute, Inc.
EXHIBIT 2.5Juran’s Quality Trilogy
Healthcare Operat ions Management34
Outcomes were viewed by Donabedian in terms of recovery, restoration of function, and survival, but he also included less easily measured outcome areas such as patient satisfaction (Donabedian 1966). He noted that process of care consists of the methods by which care is delivered, including gathering appropri-ate and necessary information, developing competence in diagnosis and therapy, and providing preventive care. Finally, he established the principle that structure is related to the environment in which care takes place, including facilities and equipment, medical staff qualifications, administrative structure, and programs. Donabedian (1966, 188) believed that quality of care is related not only to each of these elements individually but also to the relationships among them:
Clearly, the relationships between process and outcome, and between structure
and both process and outcome, are not fully understood. With regard to this, the
requirements of validation are best expressed by the concept . . . of a chain of events
in which each event is an end to the one that comes before it and a necessary condi-
tion to the one that follows.
Similar to Deming and Juran, Donabedian advocated the continuous improvement of healthcare quality through a cycle of structure and process changes supported by outcome assessment.
The influence of Donabedian’s seminal work in healthcare can still be seen. Pay-for-performance programs (structure) reward providers for deliv-ering care that meets evidence-based goals (assessed in terms of process or outcomes). The 5 Million Lives Campaign, and its predecessor, the 100,000 Lives Campaign (IHI 2006), are programs (structure) designed to decrease mortality (outcome) through the use of evidence-based practices and procedures (process). Not only are assessments of process, structure, and outcome being developed, implemented, and reported in healthcare, but the quality focus is shifting toward the systematic view of healthcare advocated by Donabedian.
Philosophies of Performance ImprovementTQM and CQI, Leading to Six SigmaThe US Navy is credited with coining the term total quality management (TQM) in the 1980s to describe its approach, informed by Japanese models, to quality management and improvement (Hefkin 1993). TQM has come to refer to a management philosophy or program aimed at ensuring quality—defined as customer satisfaction—by focusing on it throughout the organization and for each product or service life cycle. All stakeholders in the organization par-ticipate in a continuous improvement cycle.
TQM, referred to in healthcare as continuous quality improvement (CQI), is defined differently by different organizations and individuals, but in general it has come to encompass the theory and ideas of such quality experts
Total quality management (TQM)A management philosophy or program aimed at ensuring quality—defined as customer satisfaction—by focusing on it throughout the organization and for each product or service life cycle.
Continuous quality improvement (CQI)A comprehensive quality improvement and management system with three key components: planning, control, and improvement.
Chapter 2: History of Performance Improvement 35
as Deming, Juran, Philip B. Crosby, Armand V. Feigenbaum, Kaoru Ishikawa, and Donabedian. Perhaps because TQM implementation and vocabulary vary from one organization to the next, TQM programs have decreased in popularity in the United States and have been replaced with more codified programs such as Six Sigma, Lean, and the Malcolm Baldrige National Quality Award criteria.
Six Sigma and TQM are both based on the teachings of Shewhart, Deming, Juran, and other quality experts. Both methodologies emphasize the importance of top management support and leadership, and both focus on continuous improvement as a means to ensure the long-term viability of an organization. The define-measure-analyze-improve-control cycle of Six Sigma (see chapter 9) has its roots in the PDCA cycle of TQM. Six Sigma and TQM have been described as both philosophies and methodologies. Six Sigma can also be defined as a metric, or goal, of 3.4 defects per million opportunities, represented by its unit-based form, 6σ; TQM does not specify a numeric goal to achieve. TQM is not defined as Six Sigma and is not supported by or associ-ated with any certification programs.
The definition of TQM was shaped mainly by academics and is abstract and general, whereas Six Sigma has its base in industry—Motorola and General Electric were early developers—and is specific, providing a clear framework for organizations to follow. Early TQM efforts focused on quality as the primary goal; improved business performance was thought to be a natural outcome of this goal. Quality departments were mainly responsible for TQM throughout the organization. While Six Sigma sets quality (again, as defined by the customer in terms of satisfaction) as a primary goal and focuses on tangible results, it also takes into account the effects of a Six Sigma initiative on business performance. No longer is the focus on quality for quality’s sake; instead, a quality focus is seen as a means to improve organizational performance. Six Sigma training in the use of specific tools and techniques provides common understanding and common vocabulary across organizations. In other words, this method makes quality the goal of the entire organization, not just the quality department.
In essence, Six Sigma took the theory and tools of TQM and codified their implementation, providing a well-defined approach to quality that orga-nizations can quickly and easily adopt.
ISO 9000The ISO 9000 series of standards, first published in 1987 by the Interna-tional Organization for Standardization (ISO), is primarily concerned with quality management, or how the organization ensures that its products and services satisfy the customer’s quality requirements and comply with applicable regulations. In 2002, the ISO 9000 standard was renamed ISO 9000:2000, consolidating the ISO 9001, 9002, and 9003 standards into the set.
The standards are specifically concerned with the processes of ensuring quality rather than the products or services themselves. ISO standards give
ISO 9000A series of process standards developed by the International Organization for Standardization to give organizations guidelines for developing and maintaining effective quality systems.
Healthcare Operat ions Management36
organizations guidelines by which to develop and maintain effective quality systems.
A significant number of US hospitals are now using the ISO 9001 Quality Management Program to achieve Medicare accreditation. This deeming author-ity, whereby the Centers for Medicare & Medicaid Services confers accreditation authority on a third party, was granted to DNV GL (2016) in 2008.
Many organizations require that their vendors be ISO certified. For an organization to be registered as an ISO 9001 supplier, it must demonstrate to an accredited registrar (a third-party organization that is itself certified) its compliance with the requirements specified in the standard(s). Organizations that are not required by their vendors to be certified can still use the standards to develop quality systems without attempting to be certified.
Baldrige AwardJapanese automobiles and electronics gained market share in the United States during the 1970s because their quality was higher and their costs were lower than those manufactured in the United States. In the early 1980s, both US government and industry believed that the only way for the country to stay competitive was to increase industry focus on quality. The Malcolm Baldrige National Quality Award was established by Congress in 1987 to recognize US organizations for their achievements in quality. Its aim was to raise aware-ness about the importance of quality as a competitive priority and help dis-seminate best practices by providing examples of how to achieve quality and performance excellence.
The award was originally given annually to a maximum of three organi-zations in each of three categories: manufacturing, service, and small business. In 1999, the categories of education and healthcare were added, and in 2002, the first Baldrige Award in healthcare was bestowed. The healthcare category includes hospitals, health maintenance organizations, long-term care facilities, healthcare practitioner offices, home health agencies, health insurance compa-nies, and medical and dental laboratories.
The program is a cooperative effort of government and the private sec-tor. The evaluations are performed by a board of examiners, which includes experts from industry, academia, government, and the not-for-profit sector. The examiners volunteer their time to review applications, conduct site visits, and provide applicants with feedback on their strengths and opportunities for improvement in seven categories. Additionally, board members give presenta-tions on quality management, performance improvement, and the Baldrige Award.
A main purpose of the award is the dissemination of best practices and strategies. Recipients are asked to participate in conferences, provide basic mate-rials on their organizations’ performance strategies and methods to interested parties, and answer inquiries from the media. Baldrige Award recipients have
Malcolm Baldrige National Quality AwardAn annual award established by the US Congress in 1987 to recognize organizations in the United States for their achievements in quality.
Chapter 2: History of Performance Improvement 37
gone beyond these expectations to give thousands of presentations aimed at educating other organizations on the benefits of using the Baldrige framework and disseminating best practices. In fact, many organizations now use the application process as a structure for their comprehensive quality improve-ment programs.
Just-in-Time, Leading to Lean and AgileJust-in-time (JIT) is an inventory management strategy aimed at reducing or eliminating inventory. It is one aspect of Lean manufacturing, whose goal is to eliminate waste, of which inventory is one form. JIT was the term originally used for Lean production in the United States, where industry leaders noted the success of the Japanese auto manufacturers and attempted to copy it by adopting Japanese practices. As academics and organizations realized that Lean production was more than JIT, inventory management terms such as big JIT and little JIT were employed, and JIT production became synonymous with Lean production. For clarity, the term JIT refers to the inventory management strategy in this text.
After World War II, Japanese industry needed to rebuild and grow, and its leaders wanted to copy the assembly line and mass production systems found in the United States. However, the country had limited resources and limited storage space. At Toyota Motor Corporation, Taiichi Ohno and Shigeo Shingo developed what has become known as the Toyota Production System (TPS). They began by realizing that large amounts of capital dollars were tied up in inventory in the mass production system typical at that time.
Ohno and Shingo sought to reduce inventory by various means, most importantly by increasing the rate at which autos were assembled (known as flow rate). Standardization reduced the number of parts in inventory and the number of tools and machines needed. Processes such as single-minute exchange of die allowed for quick changeovers of tooling, increasing the amount of time that could be used for production by reducing setup time. As in-process inventory was reduced, large amounts of capital were freed for other purposes.
Customer lead time (the time a customer spends waiting for his vehicle once it has been ordered) was reduced as the speed of product flow increased throughout the plant. Because inventory provides a buffer for poor quality, reducing inventory forced Toyota to pay close attention to not only its own quality but suppliers’ quality as well. To discover the best ways to reduce inven-tory, management and line workers needed to cooperate, and teams became an integral part of Lean.
When the US auto industry began to be threatened by the increased popularity of Japanese automobiles, management and scholars began to study this Japanese system. However, what they brought back were usually the most visible techniques of the program—JIT, kanbans, quality circles (discussed in more depth later in the book)—rather than the underlying principles of
Just-in-time (JIT)An inventory management system designed to improve efficiency and reduce waste. Part of Lean manufacturing.
Toyota Production System (TPS)A quality improvement system developed by Toyota Motor Corporation for its automobile manufacturing lines. TPS has broad applicability beyond auto manufacturing and is now commonly known as Lean manufacturing.
Healthcare Operat ions Management38
Lean. Not surprisingly, many of the first US firms that attempted to copy this system failed; however, some were successful. The Machine That Changed the World (Womack, Jones, and Roos 1990), a study of Japanese, European, and American automobile manufacturing practices, first introduced the term Lean manufacturing and brought the theory, principles, and techniques of Lean to a broad audience.
Lean is both a management philosophy and a strategy. Its goal is to eliminate all waste in the system. Although Lean production originated in manufacturing, the goal of eliminating waste is easily applied to the service sector. Many healthcare organizations are using the tools and techniques asso-ciated with Lean to improve efficiency and effectiveness.
Sometimes seen as a broader strategy than TQM or Six Sigma, Lean requires an organization to be defined by quality. To operate as a quality orga-nization, it does not necessarily need to be Lean. However, if customers value speed of delivery and low cost, and quality is defined as customer satisfaction, a quality focus should lead an organization to implement Lean. That said, either a Lean initiative or another type of quality improvement program can result in the same outcome.
Bringing Together Baldrige, Six Sigma, Lean, and ISO 9000All of these systems or frameworks are designed for performance improvement, and each differs in area of emphasis, tools, and techniques. However, they all emphasize customer focus, process or system analysis, teamwork, and quality, and they all are compatible.
The importance of the organization’s culture, and management’s ability to shape that culture, cannot be overstated. The successful implementation of any program or deployment of any technique requires a culture that supports those changes. The leading causes of failure of new initiatives are lack of top management support and absence of buy-in on the part of employees.
Management must believe that a particular initiative will make the organi-zation better and must demonstrate its support in that belief, both ideologically and financially, to ensure the success of the initiative. Employee buy-in and support only occur when top management commitment is evident. Communi-cation and training can aid in this process, but only unequivocal management commitment ensures success.
Supply Chain Management
The term supply chain management (SCM) was first used in the early 1980s. In 2005, the Council of Supply Chain Management Professionals (2016) agreed on the following definition of SCM:
Chapter 2: History of Performance Improvement 39
Supply chain management encompasses the planning and management of all activi-
ties involved in sourcing and procurement, conversion, and all logistics management
activities. Importantly, it also includes coordination and collaboration with channel
partners, which can be suppliers, intermediaries, third party service providers, and
customers. In essence, supply chain management integrates supply and demand
management within and across companies.
This definition makes apparent that SCM is a broad discipline, encompassing activities outside as well as inside an organization.
SCM has its roots in systems thinking. Systems thinking is based on the idea that everything affects everything else. The need for systems thinking comes from the notion that optimizing one part of a system is possible, and even likely, if the whole system is suboptimal. A current example of a suboptimal system in healthcare can be seen in one purchasing avenue for prescription drugs. In the United States, the customer can optimize his drug purchases (minimize cost) by purchasing drugs from pharmacies located in foreign countries (e.g., Canada, Mexico). Often, these drugs are manufactured in the United States. While the customer has minimized his costs, the total supply chain has incurred additional costs, as with the extra transportation that takes place shipping drugs to Canada or another foreign country and then back to the United States.
SCM became increasingly important to manufacturing organizations in the late 1990s, driven by the need to decrease costs in response to competitive pressures and enabled by technological advances. As manufacturing became more automated, labor costs as a percentage of total costs decreased, and the percentage of material and supply costs increased. In 2006, 70 to 80 percent of the cost of a manufactured good was expended in purchased materials and services, and less than 25 percent was spent on labor (BEA 2006); this trend continues today. Consequently, fewer opportunities are available for reducing the cost of goods through decreasing labor and more opportunities are associ-ated with managing the supply chain. Additionally, advances in information technology allow firms to collect and analyze the information needed to be increasingly efficient in managing their supply chains.
Indeed, SCM was significantly enabled by technology, beginning with the inventory management systems of the 1970s—including materials require-ments planning—followed by the enterprise resource planning systems of the 1990s. As industry moved to increasingly sophisticated technological systems for managing the flow of information and goods, its ability to collect and respond to information about the entire supply chain expanded and firms could now actively manage their supply chains.
SCM is becoming increasingly important in healthcare as well, with its growing focus on reducing costs and the need to reduce those costs through the development of efficient and effective supply chains.
Systems thinkingA view of reality that emphasizes the relationships and interactions of each part of the system to all of the other parts.
Healthcare Operat ions Management40
Big Data and Analytics
Business has always embraced computing technologies as they become available and reliable. In a 2001 article published in The Economist, the magazine looks back at the first use of computers in business, for example:
[T]he Lyons Electronic Office (LEO), was built by Lyons, a British catering company. On
November 17th 1951, it ran a program to evaluate the costs, prices and margins for
that week’s output of bread, cakes and pies, and ran the same program each week
thereafter. In February 1954 LEO took on the weekly calculation of the company’s
payroll, prompting an article in these pages [referring to Economist (1954)].
Other computers had been used to run one-off calculations for businesses,
and many firms used mechanical or electrical calculators. But LEO was the first dedi-
cated business machine to operate on the “stored program” principle, meaning that
it could be quickly reconfigured to perform different tasks by loading a new program.
Between 1950 and 1970, business use of computers was essentially con-fined to databases and computing machines that were physically located in the business enterprise and that only operated on the organization’s owned data. In the 1970s, the personal computer was created, which allowed individuals in business to conduct their own analysis using a desktop machine. The year 1991 gave rise to the use of the Internet, freeing analysts to access data from both their own company and other sources throughout the world. In 1997, Google launched its search engine and the term big data began to appear.
Big data is typically characterized by the so-called three Vs (Marr 2015):
• Volume. Data sets were becoming very large—in 2008, 9.57 trillion gigabytes of data were processed by the world’s computers.
• Variety. Many types of data are now being stored (e.g., text, video, clinical equipment outputs).
• Velocity. The data enter computer databases at an increasing rate of speed.
In 2005, HaDoop, an open source data framework developed to process big data, was widely deployed (Bappalige 2014). HaDoop software allowed very large clusters of multiple computers to work as one and thereby provide the computing power necessary for the analysis of very large data sets. In 2014, mobile Internet usage (e.g., via tablets and smartphones) surpassed desktop usage, and the connection of many devices (e.g., thermostats, lights, refrigera-tors, pacemakers) to the Internet continues to increase (Marr 2015).
As these new technologies came online, opportunities for increasingly sophisticated analysis emerged. Many of these new and powerful tools are described throughout the remainder of this book.
Chapter 2: History of Performance Improvement 41
Conclusion
Service organizations in general, and healthcare organizations in particular, have lagged in their adoption of process improvement philosophies, techniques, and tools of operations management, but they no longer have this option. Hospitals, health systems, and other healthcare delivery organizations face increasing pressures from consumers, industry, and government to provide their services in an efficient and effective manner, and they must adopt these philosophies to remain competitive.
In healthcare today, organizations such as the Institute for Healthcare Improvement and AHRQ are leading the way in the development and dis-semination of tools, techniques, and programs aimed at improving the quality, safety, efficiency, and effectiveness of the US healthcare system.
Discussion Questions
1. What is the difference between data, information, knowledge, understanding, and wisdom? Give specific examples of each in your own organization.
2. How has operations management changed since its early days as scientific management?
3. What are the major factors leading to increased interest in the use of operations management tools and techniques in the healthcare sector?
4. Why has ISO 9000 certification become important to healthcare organizations?
5. Research those organizations that have won the Baldrige Award in the healthcare category. What factors led to their success in winning the award?
6. What are some of the reasons for the success of Six Sigma?7. What are some of the reasons for the success of Lean?8. Compare and contrast ISO 9000, the Baldrige
criteria, and Six Sigma. (More information on each of these programs is available on the book’s companion website.) Which would you find most appropriate to your organization? Why?
9. How are Lean initiatives similar to total quality management and Six Sigma initiatives? How are they different?
10. Why is supply chain management increasing in importance for healthcare organizations?
11. What are some new opportunities for the use of big data and analytics in healthcare?
On the web at ache.org/books/OpsManagement3
Healthcare Operat ions Management42
References
Auerbach, A. D. 2001. “Making Health Care Safer: A Critical Analysis of Patient Safety Practices.” In Localizing Care to High-Volume Centers. Evidence Report/Technol-ogy Assessment No. 43, AHRQ Publication No. 01-E058. Rockville, MD: Agency for Healthcare Research and Quality.
Bappalige, S. P. 2014. “An Introduction to Apache Hadoop for Big Data.” Published August 26. https://opensource.com/life/14/8/intro-apache-hadoop-big-data.
Bellis, M. 2006. “Henry Ford (1863–1947).” About.com. Accessed June 8. http://inventors. about.com/library/inventors/blford.htm.
Bureau of Economic Analysis, US Department of Commerce (BEA). 2006. “Benchmark Input-Output Accounts.” Accessed December 20. http://bea.gov/bea/dn2/home/benchmark.htm.
Cohen, I. B. 1984. “Florence Nightingale.” Scientific American 250 (3): 128–37.Council of Supply Chain Management Professionals. 2016. “CSCMP Supply Chain Manage-
ment Definitions and Glossary.” Accessed August 11. www.cscmp.org/Website/AboutCSCMP/Definitions/Definitions.asp.
Deming, W. E. 1994. The New Economics for Industry, Government, Education, 2nd edition. Cambridge, MA: MIT Press.
———. 1986. Out of the Crisis. Cambridge, MA: Cambridge University Press.DNV GL. 2016. “Hospital Accreditation: DNV GL’s Pioneering NIAHO® Program Inte-
grates ISO 9001 with the Medicare Conditions of Participation.” Accessed April 4. http://dnvglhealthcare.com/accreditations/hospital-accreditation.
Donabedian, A. 1985. The Methods and Findings of Quality Assessment and Monitoring: An Illustrated Analysis. Exploration in Quality Assessment and Monitoring, vol. 3. Chicago: Health Administration Press.
———. 1982. The Criteria and Standards of Quality. Exploration in Quality Assessment and Monitoring, vol. 2. Chicago: Health Administration Press.
———. 1980. Definition of Quality and Approaches to Its Assessment. Exploration in Quality Assessment and Monitoring, vol. 1. Chicago: Health Administration Press.
———. 1966. “Evaluating the Quality of Medical Care.” Milbank Memorial Fund Quarterly 44 (3): 166–206.
Economist, The. 2001. “It All Started with Pies: The Routine Use of Computers in Busi-ness Is 50 Years Old This Week.” Published November 15.www.economist.com/node/863267.
———. 1954. “Electronic Abacus.” Published March 13. www.economist.com/node/863244.Finnie, W. 1997. “Strategies, Systems, and Organizations: An Interview with Russell L.
Ackoff.” Strategy and Leadership 25 (2): 22–28.Gilbreth, F. B., and E. G. Carey. 1948. Cheaper by the Dozen. New York: T. Y. Crowell.Halm, E. A., C. Lee, and M. R. Chassin. 2002. “Is Volume Related to Outcome in Health
Care? A Systematic Review and Methodologic Critique of the Literature.” Annals of Internal Medicine 137 (6): 511–20.
Chapter 2: History of Performance Improvement 43
Hefkin, D. C. 1993. “The Navy’s Quality Journey: Operational Implementation of TQL.” Executive Research Project CS5. Washington, DC: Industrial College of the Armed Forces.
Heskett, J. L. 2003. Shouldice Hospital Ltd. Harvard Business School Case 9-683-068. Boston: Harvard Business School Publishing.
Institute for Healthcare Improvement (IHI). 2006. “Protecting 5 Million Lives from Harm.” Accessed June 20, 2007. www.ihi.org/IHI/Programs/Campaign/Campaign.htm.
Institute of Medicine. 1999. To Err Is Human: Building a Safer Health System. Washington, DC: National Academies Press.
International Work Simplification Institute. 1968. “Pioneers in Improvement and Our Modern Standard of Living.” IW/SI News, September 18. Accessed June 8, 2006. http://gilbrethnetwork.tripod.com/bio.html.
Juran, J. M. 1986. “The Quality Trilogy.” Quality Progress 19 (8): 19–24.Juran, J. M., and A. B. Godfrey. 1998. Juran’s Quality Handbook, 5th edition. New York:
McGraw-Hill.Kansal, B. B., and P. C. K. Rao. 2006. Preface to Management. New Delhi, India: Paragon.Lienhard, J. H. 2016. “Nightingales Graph.” Engines of Our Ingenuity. Accessed September
24. www.uh.edu/engines/epi1712.htm.Marr, B. 2015. “A Brief History of Big Data.” Published February 23. www.slideshare.net/
BernardMarr/a-brief-history-of-big-data.Mayo, E. 1933. The Human Problems of an Industrial Civilization. New York: MacMillan.McCormick, M. 2006. “Immanuel Kant: Metaphysics.” Internet Encyclopedia of Philosophy.
Accessed August 8, 2016. www.iep.utm.edu/k/kantmeta.htm.NationMaster.com. 2004. “Venetian Arsenal.” NationMaster. Accessed June 8, 2006. www.
nationmaster.com/encyclopedia/Venetian-Arsenal.Nelson, D. 1980. Frederick Taylor and the Rise of Scientific Management. Madison, WI:
University of Wisconsin Press.NetMBA.com. 2005. “Frederick Taylor and Scientific Management.” Accessed June 8,
2006. www.netmba.com/mgmt/scientific.Neuhauser, D. 2003. “Florence Nightingale Gets No Respect: As a Statistician That Is.”
Quality and Safety in Health Care 12 (4): 317.Nightingale, F. 1858. Notes on Matters Affecting the Health, Efficiency and Hospital Admin-
istration of the British Army. London: Harrison and Sons.Perkins, J. S. 1997. “Frank B. Gilbreth’s Research: The Quest of the One Best Way.” The
Quest, Accessed October 29, 2007. http://gilbrethnetwork.tripod.com/qv1n2.html.Ressler, T., and M. Ahrens. 2006. The Decision Making Book. Minneapolis, MN: University
of St. Thomas.Schmenner, R. W. 2001. “Looking Ahead by Looking Back: Swift, Even Flow in the His-
tory of Manufacturing.” Production and Operations Management 10 (1): 87–96.Shewhart, W. A. 1931. Economic Control of Quality of Manufactured Product. New York:
D. Van Nostrand.
Healthcare Operat ions Management44
Shewhart, W. A., and W. E. Deming. 1939. Statistical Method from the Viewpoint of Quality Control. Washington, DC: The Graduate School, US Department of Agriculture.
Simkin, J. 2005. “Henry Ford.” Spartacus. Accessed June 8, 2006. www.spartacus.schoolnet.co.uk/USAford.htm.
Sunol, R. 2000. “Avedis Donabedian.” International Journal for Quality in Health Care 12 (6): 451–54.
Taylor, F. W. 1911. The Principles of Scientific Management. New York: Harper & Row.Womack, J. P., D. T. Jones, and D. Roos. 1990. The Machine That Changed the World:
Based on the Massachusetts Institute of Technology 5-Million Dollar 5-Year Study on the Future of the Automobile. New York: Rawson Associates.
Zeleny, M. 1987. “Management Support Systems: Towards Integrated Knowledge Manage-ment.” Human Systems Management 7 (1): 59–70.
CHAPTER
45
EVIDENCE-BASED MEDICINE AND VALUE-BASED PURCHASING
Operations Management in Action
MultiCare Health System is an inte-grated delivery system serving com-munities throughout Washington state. After reviewing its patient populations, it undertook an initiative to lower the costs of care and improve the care expe-rience for pneumonia patients.
This initiative included build-ing an evidenced-based order set and assigning a team of social workers, called personal health partners, to research and improve patient follow-up and communication processes. It also deployed an analytics application to provide near real-time feedback on com-pliance and performance while offering a single view of patient-specific data across multiple visits and care settings.
The MultiCare team determined that a standardized electronic order set was the easiest and most effective way to define best practices while leverag-ing informatics to help clinicians “do the right thing.” This effort required bring-ing its clinicians together to review the evidence on best practices in the treat-ment of pneumonia and to arrive at a consensus on the treatment protocols.
Advanced analytics provided new capabilities to correlate processes with outcomes. MultiCare used an analytics
3OVE RVI EW
The science of medicine progressed rapidly through the latter half
of the twentieth century, with advances in pharmaceuticals, surgi-
cal techniques, and laboratory and imaging technology promoting
the rapid subspecialization of medicine itself. This “age of miracles”
improved health and lengthened life spans.
In the mid-1960s, the federal government began the Medi-
care and Medicaid programs. This new source of funding fueled the
explosive growth and expansion of the US healthcare delivery system.
However, in this vastly expanded care environment, many new tools
and clinical approaches that had little scientific merit were initiated
alongside those with great promise. As these clinical approaches were
used broadly, they became community standards. At the same time,
many simple yet highly effective tools and techniques either fell out
of favor or were not used consistently.
In response to these trends, a number of clinicians began the
movement that has become known today as evidence-based medicine
(EBM). As defined earlier, EBM is the conscientious and judicious use
of the best current evidence in making decisions about the care of
individual patients. In almost all cases, the broad application of EBM
not only improves clinical outcomes for patients but reduces costs in
the system as well.
This chapter reviews
• the history, current status, and future of EBM;
• public reporting;
• pay for performance (P4P) and payment reform; and
• value purchasing, including Medicare’s Hospital Value-Based
Purchasing (VBP) program
EBM is explored in depth, followed by an examination of how
payers use its principles to encourage the use of EBM by clinicians.
(continued)
Healthcare Operat ions Management46
application that could mine the data related to pneumonia patients and provide near real-time, interactive data that showed the im- pact of interventions on the high-level out-come metrics: mortal-ity, readmissions, length of stay (LOS), and cost. The feedback generated through these analytic tools provided the platform for continuous improvement in the order sets and protocols.
Through these efforts, MultiCare has realized significant outcome improve-ments, including the following:
• 28 percent reduction in pneumonia mortality rate
• 23 percent reduction in pneumonia readmissions
• 2 percent decrease in LOS for pneumonia patients
• 6.4 percent reduction in average variable cost per patient
Source: Health Catalyst (2016).
Evidence-Based Medicine
The expansion of clinical knowledge has three major phases. First, basic research is undertaken in the lab and with animal models. Second, carefully controlled clinical trials are conducted to demonstrate the efficacy of a diagnostic or treat-ment methodology that emerges from the preliminary research. Third, the successful or promising clinical trial results are translated to clinical practice.
The final phase, translation, is where the system often breaks down. A major study by the United Health Foundation examined the transfer of clinical research knowledge to the so-called bedside and reported (Ellis et al. 2012)
both quality and actual medical costs for episodes of care provided by nearly 250,000
US physicians serving commercially insured patients nationwide. Overall, episode
costs for a set of major medical procedures varied about 2.5-fold, and for a selected
set of common chronic conditions, episode costs varied about 15-fold. Among doc-
tors meeting quality and efficiency benchmarks, however, costs for episodes of care
were on average 14 percent lower than among other doctors.
The cure for this wide variation in practice is the consistent application of EBM. The key tool for doing so is the clinical guideline (Shekelle 2016):
OVE RVI EW (Continued)
The operations tools presented in other chapters of this book are
introduced in terms of how they are linked to achieving EBM goals.
The chapter concludes with an illustration of the chartering of a
project team to improve implementation of EBM at Vincent Valley
Hospital and Health System (VVH).
Chapter 3: Evidence-Based Medicine and Value-Based Purchasing 47
Clinical practice guidelines are recommendations for clinicians about the care of
patients with specific conditions. They should be based upon the best available
research evidence and practice experience.
The Institute of Medicine [2011] defines clinical practice guidelines as “state-
ments that include recommendations, intended to optimize patient care, that are
informed by a systematic review of evidence and an assessment of the benefits and
harms of alternative care options.”
Based on this definition, guidelines have two parts:
• The foundation is a systematic review of the research evidence bearing on a
clinical question, focused on the strength of the evidence on which clinical
decision-making for that condition is based.
• A set of recommendations, involving both the evidence and value judgments
regarding benefits and harms of alternative care options, addressing how
patients with that condition should be managed, everything else being equal.
A comprehensive source for such information is the National Guideline Clearinghouse (NGC 2016), a database of evidence-based clinical practice guidelines and related documents that contains more than 4,000 guidelines. NGC is a joint project of the Agency for Healthcare Research and Quality (AHRQ), the American Medical Association, and America’s Health Insurance Plans. In addition, AHRQ (2016b) provides easy-to-use resources for clinicians and patients through its Effective Health Care Program.
What are the barriers to the wider application of EBM? Baiardini and colleagues (2009) reviewed the literature and identified 293 potential obstacles to the use of guidelines by physicians. They then grouped these into seven barriers:
1. Lack of knowledge that guidelines exist for a specific condition2. Lack of familiarity with the details of specific guidelines3. Disagreement with the guideline recommendations4. Inability to effectively apply a guideline’s recommendation due to lack
of skill, resources, or training5. Lack of trust in the effectiveness of a guideline to improve outcomes—
particularly with an individual patient’s condition6. Resistance to change and reliance on habits7. External factors (lack of resources, financial barriers or incentives,
organizational factors)
The application of EBM is a two-way street that requires the involve-ment of the patient as well as the physician. Baiardini and colleagues (2009) also identified the following barriers to patients’ compliance with guidelines:
Healthcare Operat ions Management48
• Presence of confounding characteristics, such as a psychiatric or psychological comorbidity or lack of social support
• Difficulty in recognizing symptoms and adhering to therapies prescribed for the symptoms
• Complex therapeutic regimens• Relationship and personal interaction issues between patient and physician
Standard and Custom Patient CareOne historical criticism of EBM is that all patients are unique and EBM is “cookbook” medicine that only applies to a few patients. EBM proponents counter this argument with simple examples of well-accepted and effective clinical practices that are inconsistently followed. A more productive view of the mix of art and science in medicine is provided by Bohmer (2005), who suggests that all healthcare is a blend of custom and standard care. Exhibit 3.1 shows the four currently used models that blend these two approaches.
Model A (separate and select) provides an initial sorting by patients themselves. Those with standard problems are treated with standard care using EBM guidelines. Examples of this type of system are specialty hospitals for laser eye surgery and walk-in clinics operating in pharmacies and retail outlets. Patients who do not fit the provider’s homogeneous clinical conditions are referred to other providers who can deliver customized care (Bohmer 2005).
OutputInputReasoningprocess
Sortingprocess
Standardsubprocess
Customizedsubprocess OI
(A) Separateand select
O
I
(B) Separate and accommodate
OO
I
O
I(D) Integrated
O
I(C) Modularized
EXHIBIT 3.1Four
Approaches to Blending Custom and
Standard Processes
Source: Bohmer (2005). Used with permission.
Chapter 3: Evidence-Based Medicine and Value-Based Purchasing 49
Model B (separate and accommodate) combines the two methods inside one provider organization. Duke University Health System, for example, has developed standard protocols for its cardiac patients. Patients are initially sorted, and those who can be treated with the standard protocols are cared for by nurse practitioners using a standard care model. Cardiologists care for the remainder using custom care. However, on every fourth visit to the nurse practitioner, the cardiologist and nurse practitioner review the patient’s case together to ensure that standard care is still the best treatment approach (Bohmer 2005).
Model C (modularized) is used when the clinician moves from the role of care provider to that of architect of care design for the patient. In this case, a number of standard processes are assembled to treat the patient. The Andrews Air Force Base clinic uses this system to treat hypertension patients. “After an initial evaluation, treatment may include weight control, diet modification, drug therapy, stress control, and ongoing surveillance. Each component may be provided by a separate professional and sometimes a separate organization. What makes the care uniquely suited to each patient is the combination of components” (Bohmer 2005, 326).
Model D (integrated) combines standard care and custom care in a single organization. In contrast to Model B, each patient receives a mix of both custom and standard care as determined by her condition. Intermountain Healthcare (IHC) employs this model through the use of 62 standard care processes available as protocols in its electronic health record (EHR). These processes cover “the care of over 90 percent of patients admitted in IHC hos-pitals” (Bohmer 2005, 326). Clinicians are encouraged to override elements in these protocols when it is in the best interest of the patient. All of these overrides are collected and analyzed, and changes are made to the protocol, which is an effective method to continuously improve clinical care.
All of the tools and techniques of operations improvement included in the remainder of this book can be used to make standard care processes oper-ate effectively and efficiently.
EBM and Cost ReductionEBM has the potential to not only improve clinical outcomes but also decrease total cost in the US healthcare system. Potentially preventable hospitalizations, which might be avoided with high-quality outpatient treatment and disease management, provide just one significant opportunity for financial savings.
AHRQ (2015) developed a set of prevention quality indicators (PQIs) to assist providers in reducing the number of potentially preventable hospi-talizations for chronic and acute conditions throughout the United States. A patient who is admitted to a hospital and has a PQI code is an individual whose hospitalization or other severe complication is potentially preventable when good, evidence-based outpatient care is delivered.
Prevention quality indicator (PQI)A set of measures that can be used with hospital discharge data to identify patients whose hospitalizations or complications might have been avoided with the use of evidence-based ambulatory care.
Healthcare Operat ions Management50
The PQI system is now integrated with many other federal healthcare improvement efforts (exhibit 3.2).
Chronic Disease Management One of the most expensive aspects of all healthcare systems is the care of patients with chronic disease (e.g., diabetes, chronic obstructive pulmonary disease, congestive heart failure). Much of the variation in the outcomes of this care can be attributed to providers’ and patients’ lack of adherence to EBM.
Fortunately, many investigators now look beyond determining which clinical interventions provide good results (e.g., the use of statins) to identify-ing those systems of care that produce superior results. (Chapter 9 provides more details and examples of the use of business process improvements to achieve high-quality care.)
Federal Initiatives Using AHRQ QIs*
Indicator Module
Inpatient (IQI)
Patient Safety (PSI)
Pediatric (PDI)
Prevention (PQI)
HAC Reduction Program Hospital Inpatient Quality Reporting Program
Hospital VBP Shared Savings Program Partnership for Patients Healthcare Innovation Awards (CMMI)
Hospital Compare ACO: Accelerated Development Learning Sessions (CMMI)
Home and Community Based Services
* A sample of CMS and CMMI initiatives that use the AHRQ QIs.
Source: Reprinted from AHRQ (2015).
Note: AHRQ = Agency for Healthcare Research and Quality; CMMI = Center for Medicare & Medicaid Innovation; CMS = Centers for Medicare & Medicaid Services; Hospital VBP = Medicare Hospital Value-Based Purchasing program; IQI = inpatient quality initiative; PDI = pediatric initiative; PQI = prevention quality initiative; PSI = patient safety initiative; QI = quality initiative.
EXHIBIT 3.2PQIs and
Other Federal Initiatives
Chapter 3: Evidence-Based Medicine and Value-Based Purchasing 51
The Chronic Care ModelDr. Edward Wagner of the MacColl Center for Health Care Innovation, a leader in the improvement of chronic care, has developed one of the most widely accepted models for chronic disease management (Wagner et al. 2001). The first important element of Wagner’s chronic care model (CCM) is population-based outreach, which ensures that all patients in need of chronic disease management receive it. Next, treatment plans are created that are sensitive to each patient’s preferences. The most current evidence-based medicine is employed, and this process is aided by clinical information systems with built-in decision support. The patient is encouraged to change risky behaviors and improve the management of his health.
The clinical visit itself differs in the Wagner model to allow more time for interaction between the physician and patients with complicated clinical issues. Visits for routine or specialized matters are handled by other healthcare profes-sionals (e.g., nurses, pharmacists, dieticians, lay health workers). Close follow-up, supported by clinical information system registries and patient reminders, is also characteristic of effective chronic disease management (Wagner et al. 2001).
The CCM has now been widely deployed. In a review of 16 studies of the care of diabetes patients, for example, Stellefson, Dipnarine, and Stopka (2013) found
evidence that CCM approaches have been effective in managing diabetes in US
primary care settings. Organizational leaders in health care systems initiated sys-
tem-level reorganizations that improved the coordination of diabetes care. Disease
registries and electronic medical records were used to establish patient-centered
goals, monitor patient progress, and identify lapses in care. Primary care physicians
(PCPs) were trained to deliver evidence-based care, and PCP office–based diabetes
self-management education improved patient outcomes.
Patient-Centered Medical Homes The patient-centered medical home (PCMH) concept has emerged as an effec-tive tool in the delivery of care to patients with chronic disease. The Affordable Care Act (ACA) supported this innovation with additional payment for Medicaid patients (§2703). Also known as the healthcare home, the PCMH has proven to be a valuable addition to the care management approach for patients with chronic diseases and is now being funded by both government and private payers.
AHRQ (2016a) defines the PCMH as
a model of the organization of primary care that delivers the core functions of pri-
mary health care.
The medical home encompasses five functions and attributes:
1. Comprehensive Care
The primary care medical home is accountable for meeting the large majority of
each patient’s physical and mental health care needs, including prevention and
Patient-centered medical home (PCMH) Care that is accessible, continuous, comprehensive, family centered, coordinated, compassionate, and culturally effective.
Healthcare Operat ions Management52
wellness, acute care, and chronic care. Providing comprehensive care requires a
team of care providers. This team might include physicians, advanced practice
nurses, physician assistants, nurses, pharmacists, nutritionists, social workers,
educators, and care coordinators. Although some medical home practices may
bring together large and diverse teams of care providers to meet the needs of their
patients, many others, including smaller practices, will build virtual teams linking
themselves and their patients to providers and services in their communities.
2. Patient-Centered
The primary care medical home provides health care that is relationship-based
with an orientation toward the whole person. Partnering with patients and their
families requires understanding and respecting each patient’s unique needs,
culture, values, and preferences. The medical home practice actively supports
patients in learning to manage and organize their own care at the level the patient
chooses. Recognizing that patients and families are core members of the care
team, medical home practices ensure that they are fully informed partners in
establishing care plans.
3. Coordinated Care
The primary care medical home coordinates care across all elements of the
broader health care system, including specialty care, hospitals, home health
care, and community services and supports. Such coordination is particularly
critical during transitions between sites of care, such as when patients are being
discharged from the hospital. Medical home practices also excel at building clear
and open communication among patients and families, the medical home, and
members of the broader care team.
4. Accessible Services
The primary care medical home delivers accessible services with shorter waiting
times for urgent needs, enhanced in-person hours, around-the-clock telephone
or electronic access to a member of the care team, and alternative methods of
communication such as email and telephone care. The medical home practice
is responsive to patients’ preferences regarding access.
5. Quality and Safety
The primary care medical home demonstrates a commitment to quality and qual-
ity improvement by ongoing engagement in activities such as using evidence-
based medicine and clinical decision-support tools to guide shared decision
making with patients and families, engaging in performance measurement and
improvement, measuring and responding to patient experiences and patient
satisfaction, and practicing population health management. Sharing robust
quality and safety data and improvement activities publicly is also an important
marker of a system-level commitment to quality.
The PCMH model has been shown to increase quality and reduce costs. A University of Minnesota evaluation of the Health Care Homes initiative
Chapter 3: Evidence-Based Medicine and Value-Based Purchasing 53
in that state found that over a five-year evaluation period, the use of medical homes reduced inpatient admissions by 29 percent and hospital outpatient visits by 38 percent. The study also reported improvements in the quality of care for patients with diabetes, vascular disease, asthma, and depression (Wholey et al. 2016, i, 43).
EBM and Comparative Effectiveness ResearchThe source of evidence for EBM has long been medical research that is pub-lished in respected and refereed journals. However, these studies usually are initiated by a single investigator’s interest, and thus the efficacy of many com-mon clinical approaches has never been adequately tested. The medical research community has held historical and understandable biases toward developing technologies that are designed to address intractable diseases and mysterious diagnostic challenges. Many aspects of routine healthcare have therefore never been sufficiently evaluated.
To address this problem, the ACA (and the American Recovery and Reinvestment Act [ARRA]) contained significant policy direction for the establishment and funding of a nonprofit corporation, the Patient-Centered Outcomes Research Institute (PCORI). ACA Section 6301 states that the mission of PCORI is
to assist patients, clinicians, purchasers, and policy-makers in making informed
health decisions by advancing the quality and relevance of evidence concerning the
manner in which diseases, disorders, and other health conditions can effectively and
appropriately be prevented, diagnosed, treated, monitored, and managed through
research and evidence synthesis that considers variations in patient sub-populations,
and the dissemination of research findings with respect to the relative health outcomes,
clinical effectiveness, and appropriateness of the medical treatments, and services.
PCORI’s focus is on the application of EBM to specific healthcare technologies and treatments to ascertain which, among alternative therapies for a given medical condition, produce the best clinical outcomes. This specific focus is known as comparative effectiveness research (CER). PCORI’s (2014) CER agenda has five priorities:
• Assessing prevention, diagnosis, and treatment options • Improving healthcare systems• Communicating and disseminating research• Addressing disparities across patient populations and the healthcare
required to achieve best outcomes in each population• Accelerating patient-centered outcomes research and methodological
research
Healthcare Operat ions Management54
PCORI complements the work of the National Institutes of Health and AHRQ—both part of the US Department of Health and Human Services (HHS). One of AHRQ’s responsibilities is to assist users of health information technology that is focused on clinical decision support to incorporate research findings into clinical practices and to promote the technology’s ease of use. A major focus for the research topics addressed by PCORI is related to chronic disease management.
Tools to Expand the Use of Evidence-Based Medicine
Organizations that are outside the healthcare delivery system itself, such as pay-ers and government, have used the increased acceptance of EBM as the basis for new programs designed to encourage its implementation. These programs, referred to as value purchasing, feature public reporting of clinical results and pay-for-performance (P4P) elements to help third-party payers determine the value delivered by healthcare providers.
Public ReportingAlthough strongly resisted by clinicians for many years, public reporting has come of age. The Centers for Medicare & Medicaid Services (CMS) now reports the performance of hospitals, long-term care facilities, and medical groups online at Hospital Compare (www.hospitalcompare.hhs.gov). Many private health insurance plans also report performance and the prices charged by providers in their networks to assist their plan members, particularly those with consumer-directed health insurance products, in choosing how and from whom they receive treatment or preventive care.
As with any growing field, a number of issues surround public report-ing. The first and most prominent is risk adjustment. Most clinicians feel their patients are “sicker” than average and that contemporary risk adjustment systems do not adequately account for this factor in reimbursement. Patient compliance is another challenging aspect of public reporting. If a doctor follows EBM guidelines for diagnosis and treatment but the patient does not take her medication, for example, the public reporting mechanism may trigger an unwarranted poor grade.
One anticipated impact of public reporting is that patients will use the Internet to shop for quality healthcare products as they might for an automobile or a television. Currently, however, few patients do so to guide their health-care buying decisions. That said, clinical leaders do review the public reports and target improvement efforts to areas where they have poor performance compared to their peers.
AHRQ (2012) conducted a comprehensive review of the impact of public reporting on the healthcare system. Select findings from its research include the following:
Value purchasing A system using payment as a means to reward providers who publicly report results and achieve high levels of clinical care. Also known as value-based purchasing.
Public reportingA statement of healthcare quality made by hospitals, long-term care facilities, and clinics. May also include patient satisfaction and provider charges.
Risk adjustmentRaising or lowering fees paid to providers on the basis of factors that may increase medical costs, such as age, sex, or illness.
Chapter 3: Evidence-Based Medicine and Value-Based Purchasing 55
• Public reporting has a positive impact on mortality reduction and specific clinical outcomes such as pain reduction, decreased pressure ulcers, and increased patient satisfaction.
• Changes in the delivery structure were observed as a result of public reporting, including the addition of new services, policy revisions, departure of surgeons with poor outcomes, and increases in quality improvement activities.
• Public reports seemed to have little to no impact on selection of providers by patients and families or their representatives.
• Public reporting does have an impact in competitive markets, and improvements are more likely to occur in the subgroup of providers with low scores in initial public reports than for those with high or moderate scores.
Pay for Performance and Payment ReformAnother logical tool to expand the use of EBM is the financing system. Many buyers of healthcare are installing P4P systems to encourage providers to deliver EBM care.
P4P MethodsIn general, P4P systems add payments to the amount that would otherwise be reimbursed to a provider. To obtain these additional payments, the provider must demonstrate that he is delivering care that meets clinical EBM goals. These clinical measures can be either process or outcome measures.
Although many providers prefer to be measured on outcomes, this approach is difficult to use, as some outcomes need to be measured over many years. In addition, some providers have a small number of patients in a particu-lar clinical group, so outcome results can vary dramatically. Therefore, process measures backed by extensive EBM literature are used to assess performance in the treatment of many conditions. For example, a patient with diabetes whose blood pressure is maintained in a normal range tends to experience fewer complications than one whose blood pressure is uncontrolled. Blood pressure can be measured and reported at every visit, whereas complications occur infrequently.
In a study sponsored by the National Quality Forum, Schneider, Hussey, and Schnyer (2011) surveyed the breadth of payment reform methods and found nearly 100 implemented and proposed payment reform programs. They then classified these methods into 11 payment reform models. Many of these models are included in the ACA, and the goals for the reforms are illustrated in exhibit 3.3.
Exhibit 3.4 lists and describes each model, and chapter 14 examines how organizations can apply the operations management tools contained throughout this book to succeed financially with any of these payment models.
Healthcare Operat ions Management56
Cost containment goals• Reverse the fee-for-service
incentive to provide more services• Provide incentives for efficiency• Manage financial risk• Align payment incentives to
support quality goals
Quality goals• Increase or maintain appropriate
and necessary care• Decrease inappropriate care• Make care more responsive to
patients• Promote safer care
Source: Schneider, Hussey, and Schnyer (2011).
EXHIBIT 3.3General
Payment Reform Model
Model Description
1. Global payment A single per-member per-month payment is made for services delivered to a patient, with payment adjustments based on measured performance and patient risk.
2. ACO shared sav-ings program
Groups of providers (known as accountable care organizations [ACOs]) that voluntarily assume responsibility for the care of a population of patients share payer savings if they meet quality and cost performance benchmarks.
3. Medical home payments
A physician practice or other provider is eligible to receive additional pay-ment if medical home criteria are met. Payment may include calculations based on quality and cost performance using a P4P-like mechanism.
4. Bundled payment
A single bundled payment, which may include multiple providers in mul-tiple care settings, is made for services delivered during an episode of care related to a medical condition or procedure.
5. Hospital–physician gainsharing
Hospitals are permitted to provide payments to physicians that represent a share of savings resulting from collaborative efforts between the hospital and physicians to improve quality and efficiency.
6. Payment for coordination
Payments are made to providers furnishing care coordination services that integrate care between providers.
7. Hospital P4P Hospitals receive differential payments for meeting or missing perfor-mance benchmarks.
8. Payment adjustment for readmissions
Payments to hospitals are adjusted based on the rate of potentially avoid-able readmissions.
9. Payment adjust-ment for hos-pital-acquired conditions
Hospitals with high rates of hospital-acquired conditions are subject to a payment penalty, or treatment of hospital-acquired conditions or serious reportable events is not reimbursed.
10. Physician P4P Physicians receive differential payments for meeting or missing perfor-mance benchmarks.
11. Payment for shared decision making
Payment is made for the provision of shared decision-making services.
Source: Schneider, Hussey, and Schnyer (2011).
EXHIBIT 3.4Payment Reform
Model Details
Chapter 3: Evidence-Based Medicine and Value-Based Purchasing 57
Value-Based Purchasing1 The ACA calls for establishment of a value purchasing program on the basis of much of the research, practical experience, and analysis in both public reporting and P4P described in the previous section. (If portions of the ACA are repealed or changed, value purchasing is likely to remain intact in some form because it is so strongly supported by research.) Medicare’s Hospital VBP program is CMS’s (2015) answer to that call. Forms of payment such as value purchasing, as alternatives to the traditional fee-for-service (FFS) reimbursement scheme, are accelerating, and soon the majority of financing systems for health services in the United States will move completely from FFS to value purchasing.
Although FFS has served the health industry well for many years, poli-cymakers have come to understand that perverse incentives accompany this type of payment system. Insurer UnitedHealth Group’s UnitedHealth Center for Health Reform & Modernization (2012) conducted a review of the many studies on FFS and found three major problems:
• FFS encourages providers to deliver more, and more expensive, services to maximize reimbursement.
• FFS facilitates fragmented and uncoordinated care delivery.• FFS does not offer incentives for high-quality care.
These problems have been well known for many years, and policymak-ers have searched for new payment models through Medicare demonstration projects—many of which were included in the ACA. For example, the Medi-care Shared Saving Program (§3022 of the ACA) was based on the Physician Group Practice Demonstration (CMS 2011), and the Bundled Payments for Care Improvement Initiative in the Center for Medicare & Medicaid Innova-tion (§3021) is based on the Acute Care Episode Demonstration (CMS 2016).
Today, alternative payment schemes are founded on one of two distinc-tive methodologies: bundled payments for services or additional payments or penalties for quality.
Medicare Value Purchasing As mentioned earlier, the transition from FFS to value-based systems is accel-erating. In 2015, then Secretary of HHS Sylvia Mathews Burwell announced, “Our goal is for 30% of all Medicare provider payments to be in alternative payment models that are tied to how well providers care for their patients, instead of how much care they provide in 2016. Our goal would then be to get to 50% by 2018.” The independent, not-for-profit organization Catalyst for Payment Reform (2014), which evaluates payment systems throughout the United States, found that the percentage of payments meeting its definition of value-oriented payment methods had reached 40 percent for 2014—up from 11 percent in 2013. This accelerated transformation is likely to continue.
Healthcare Operat ions Management58
Policy Issues in Value PurchasingThe rapid movement to value purchasing presents a number of policy issues.
Attribution, or Whose Patient Is This? In a complex delivery system, the connection of one patient’s care outcomes to a specific provider can be problematic. The Center for Healthcare Quality & Payment Reform has identified a number of these types of issues (Miller 2014). The following are just a few examples:
• Patients who lack a primary care physician can cause distortions in spending comparisons.
• As a function of EHR system structures, a physician can be assigned accountability for services a patient received from another provider.
• The cost of caring for a patient with a preventable conditions may be assigned to the physician treating the condition rather than the provider who caused it.
Too Many Measures The use of quality measures as the basis for payment is increasing the complexity of the system. For example, the number of ways that quality is measured has grown dramatically. In 2015, the Washington Post reported that 33 different care programs in Medicare used a combined 1,676 reporting measures the previous year (Millman 2015). A 2013 Health Affairs study of 23 commercial health plans found 546 distinct quality measures—with very little overlap to Medicare programs (Delbanco 2015).
Unintended ConsequencesComplex systems can have unintended consequences. For example, in 2008 the ARRA provided significant funding to assist with the installation of EHRs in hospitals and clinics. A clear aim of this policy was to enable providers to track patients with chronic disease, improve their care, and reduce costs in the system. However, as a consequence of more complete records arising from the use of EHRs, hospitals received $1 billion more in Medicare reimbursements in 2010 than they had five years earlier through improved billing of emergency department coding alone, according to a New York Times analysis of Medicare data (Abeslson, Creswell, and Palmers 2012). The article also notes that clinics have similarly changed the way they bill for office visits, increasing their pay-ments by billions of dollars. The consequence of increased Medicare billings was not an aim of the ARRA.
Considering that history, value purchasing’s impact on the care system will also likely produce outcomes that have not been anticipated by its architects.
Chapter 3: Evidence-Based Medicine and Value-Based Purchasing 59
Implications for Operations ManagementOne clear advantage of FFS was its clean lines of accountability for services—if you provided the service, you got paid. Value purchasing breaks this link as, in many cases, the service provider does not get paid directly. Hence, improved operational structures need to be built to accommodate these payment systems.
Strategy Execution The value purchasing environment leads to growth in the number of quality improvement projects required to respond to the new incentive opportunities. A useful management strategy is the blended balanced scorecard–strategy map-ping approach developed by Kaplan and Norton (2001). This method converts general strategies (e.g., reduce readmission rates) into specific projects (e.g., acquire predictive analytics capability), which are then connected in a strategy map. Each project establishes metrics that then can be displayed as a scorecard. This disciplined execution method is used by many large organizations both inside and outside healthcare. The balanced scorecard methodology is outlined in detail in chapter 4.
Improved Modeling and AnalyticsThe new environment requires more sophisticated systems of analysis than in the past. While traditional accounting systems were adequate for the Medicare FFS environment, much more detailed costing systems are now needed, such as activity-based accounting. Patient behavior models were historically built on groups (e.g., males over age 65) but now must be built with individual predictive modeling capabilities. Modeling and analytics tools can be used to finely align delivery system resources with patient needs. Analytics is addressed in chapter 8, and activity-based accounting is covered in chapter 14.
Innovation CentersThe new value purchasing environment is also sparking creativity. Many health-care organizations have launched innovation centers to coalesce creative energy toward developing new approaches to care delivery. Innovation centers are addressed in chapter 5.
Clinical Decision Support
One development in the use of guidelines is the spread of clinical decision sup-port systems, which are now becoming a standard part of EHRs. As a clinician accesses a specific patient’s medical record, the automated system provides advice on recommended treatments and needed follow-up (see the Operations Management in Action section at the beginning of this chapter).
Healthcare Operat ions Management60
Institute for Clinical Systems Improvement and High-Tech Diagnostic ImagingClinical decision support can be applied across multiple EHR systems and need not be vendor specific. The Institute for Clinical Systems Improvement (ICSI 2012), for example, undertook a project in 2007 to improve the appropriate utilization of CT (computed tomography), MRI (magnetic resonance imaging), PET (positron emission tomography), and nuclear cardiology diagnostic scans.
ICSI (2009) noted:
[The approach of those organizations we studied] consists of deploying a common
set of appropriateness criteria that would be:
• available in the physician’s office to provide clinical decision support at the
time care is being discussed with the patient and prior to ordering HTDI [high-
tech diagnostic imaging] tests
• embedded into an electronic medical record (EMR), or made available via a
Web site
• continually enriched and expanded for improved outcomes.
The ordering guidance screen is shown in exhibit 3.5. The ICSI (2009) project analysis continues, noting:
[The simple 1 through 9 rating on] the level of diagnostic utility of the provider’s selec-
tion carries multiple benefits, offering guidance to ordering providers and supporting
shared decision making between providers and patients. For those organizations with
Provider sees appropriateness of test and higher utility options—opportunity to engage patient.
Chest CT has marginal utility for clinical indications provided.
Alternate procedures to consider:
23456789
229MRACTA
Indicated 7−9 Marginal 4−6 Low utility 1−3
MR
1
EXHIBIT 3.5Decision
Support Process Embedded
in Electronic Health Record
Source: Copyright © 2011 Institute for Clinical Systems Improvement. Used with permission.
Note: CT = computed tomography; CTA = computed tomography angiography; MR = magnetic reso-nance; MRA = magnetic resonance angiography.
Chapter 3: Evidence-Based Medicine and Value-Based Purchasing 61
full EHRs, the patient’s clinical information is loaded automatically into this system
which then makes its recommendation based on guidelines from the American Col-
lege of Radiology and the American College of Cardiology.
When a test of a value that is below 6 is ordered, additional information is
provided to the ordering physician, who may choose to continue and order the test
or switch to another. All payers in the system have agreed to make payments no
matter what level of test is ordered. In some cases the recommended test is, in fact,
more expensive than the test originally ordered.
The project has been successful in making appropriate recommenda-tions to providers. Exhibit 3.6 shows the actual use of HTDI versus the trend that would have been seen had the existing radiology management systems remained in place.
As determined by ICSI (2010):
The summary of the benefits of this system over three years among five large medi-
cal groups is:
• $84 million savings based on reduction of HTDI scans against projected trend
line without decision-support
• 11,000 fewer administrative hours for just one medical group by having
electronic decision support accepted versus calling the radiology benefits
manager
• Decreased exposure to radiation—potentially preventing cancers
60
55
45
50
40
35
3032.03
36.12
39.1940.84
44.8947.52
51.26
54.2656.35
39.47
43.94
40.3040.2139.77
42.1342.3942.54
38.09
State legislativemandate for MN DHS
to address HTDI.
Pilot ends; medical groupscontinue to use decision support.
Yearlong ICSIdecision support
pilot begins.25
1Q03
2Q03
3Q03
4Q03
1Q04
2Q04
3Q04
4Q04
1Q05
2Q05
3Q05
4Q05
1Q06
2Q06
3Q06
4Q06
1Q07
2Q07
3Q07
4Q07
1Q08
2Q08
3Q08
4Q08
1Q09
2Q09
3Q09
4Q09
1Q10
2Q10
3Q10
Aggregate Utilization per 1,000 Members
Projected utilization at 1Q03–2Q06 average rate of changeProjected utilization at 2Q06–3Q10 average rate of changeActual utilization
Source: Copyright © 2011 Institute for Clinical Systems Improvement. Used with permission.
Note: 1Q03 = first quarter of 2003, 2Q03 = second quarter of 2003, etc.; ICSI = Institute for Clinical Systems Improvement; MN DHS = Minnesota Department of Health Services.
EXHIBIT 3.6Utilization of High-Tech Digital Imaging (HTDI)—Actual Versus Trend
Healthcare Operat ions Management62
The Future of Evidence-Based Medicine and Value Purchasing
One challenge of the increasingly widespread use of EBM is the fact that it is based on averages resulting from clinical studies of many patients. No specific patient is ever completely average, and clinicians frequently vary from guidelines to compensate for this difference. As described next, Optum Labs is a leading example of how big data can be used to address this challenge.
The second major obstacle that arose with the increased use of EBM relates to the clinicians themselves. What systems can be created to support professionalism and fair compensation and yet encourage the use of the most current and effective healthcare methods and technologies? A brief look at physician compensation and process improvement later in this section helps set the stage for answering this question, which we return to throughout the remainder of the book.
Optum LabsVery large databases are now being created to more fully research the impact of EBM. Optum Labs is a partnership of Optum and the Mayo Clinic that, as of 2016, included 19 additional industry partners. A key asset of Optum Labs is its high-quality, integrated healthcare database, which contains deidentified claims and clinical data for more than 150 million people, gathered from multiple health plans and healthcare providers. The database also includes plan enrollment information, medical and pharmacy claims, and lab results from multiple payers that have been integrated across care settings and longitudinally linked at the patient level. This database allows Optum Labs to perform fine-grained CER.
An Optum Labs Example: DiabetesWallace and colleagues (2014) offer an example of Optum Labs’ effectiveness in diabetes management:
Metformin is consistently recommended as the initial intervention for patients newly
diagnosed with uncomplicated type 2 diabetes. However, there are a number of
choices for second-line medication treatment, including older sulfonylurea drugs
and newer oral agents plus insulin.
An observational study using the Optum Labs database that compared alter-
native medication management strategies across 37,501 patients showed similar
effects for all drugs in achieving glucose control, longevity, and overall quality of
life. However, the cost of this benefit was less in patients who were treated with
sulfonylureas. These drugs were also associated with a longer interval until insulin
was required than was the case when other oral agents were used. These findings
are being translated into potential revisions of guidelines used by care providers.
Chapter 3: Evidence-Based Medicine and Value-Based Purchasing 63
As the size and scope of these large databases increase, the ability to perform highly detailed analysis will improve. These new studies will lead to ever more precise evidence-based guidelines and accurate clinical effectiveness data.
Physician Compensation and Value PurchasingA major emphasis of value purchasing is to change physician behavior through payment systems. Physician compensation is a complex and frequently contro-versial topic in healthcare organizations, and value purchasing alone will not resolve this challenge. Because CMS and private payers continue to introduce many new metrics and publicly reported quality measures, an organization might be tempted to directly link physician payment to these metrics—this linkage may actually be happening in some small practices.
However, in large systems, the number and complexity of the met-rics and their relationship to all the supporting clinical systems render both accountability and transparency difficult. A basic rule of compensation systems is that the “line of sight” should be clear between a goal and a reward; value purchasing does not allow line of sight to be achieved easily.
In a report created for the Medicare Payment Advisory Commission, Zis-mer and colleagues interviewed 15 senior leaders of integrated health systems on reimbursement models and the alignment of incentives in physician compensation (Zismer 2013). A key finding was that stability in provider compensation was a major factor in retaining and recruiting physicians. Zismer comments that to bring about such stability, payment systems must disconnect how the organization is paid from how the physician is paid. Although quality outcomes are important, many physicians in integrated systems have other obligations, such as treating expanded panels of patients, managing mid-level practitioners, and teaming with colleagues to manage the care of complex patients. Hence, compensation needs to take into account payment for the many actual duties of physicians today.
A clear strategy outlined in the ACA is to encourage the formation of systems of care. To respond effectively to value purchasing will take teams of highly skilled clinicians and process improvement personnel working diligently to meet the performance goals. The remaining chapters in this book provide the tools for this ongoing journey.
Vincent Valley Hospital and Health System and Pay for Performance
The leaders of VVH feel they have a number of opportunities to succeed with the Medicare Hospital Value-Based Purchasing program. They begin by creating a project team to improve the care of patients with pneumonia. The specific measures the team targets for improvement are those delineated in the VBP:
Healthcare Operat ions Management64
• Pneumonia patients assessed and given pneumococcal vaccination• Pneumonia patients whose initial emergency department blood culture
was performed prior to the administration of the first hospital dose of antibiotics
• Pneumonia patients given smoking cessation advice and counseling• Pneumonia patients given initial antibiotic(s) within six hours of arrival• Pneumonia patients given the most appropriate initial antibiotic(s)• Pneumonia patients assessed and given influenza vaccination
The operations management tools and approaches detailed in this book were used to improve performance for each of these measures, culminating in chapter 15, which describes how VVH accomplishes this goal.
Conclusion
The use of EBM to develop systems of care is becoming well accepted by most clinicians. Clinical results are being made transparent and easily accessible to the general public. Payers are implementing systems that reward value, and providers are installing clinical decision support systems to help in their practices. The effective use of EBM identifies high-performance healthcare organizations, and its widespread use is a key to the provision of high-quality, cost-effective care throughout the world.
Discussion Questions
1. In addition to those mentioned in the chapter, what are some examples of a care delivery setting offering a mix of standard and custom care?
2. Access the CMS Hospital Compare website and review three local hospitals’ quality scores. At which hospital would you choose to receive care, and why? Which hospital would you choose for your parents or your children? Did your answers differ? Why or why not?
3. Review the 11 payment reform methodologies (exhibit 3.4) and rank them on two scales: ability to improve quality and ability to reduce healthcare inflation. Provide a rationale for your ranking.
4. What are three strategies to maximize P4P revenue?
Note
1. Portions of this section were adapted from McLaughlin (2015) with permission from the American College of Healthcare Executives.
Chapter 3: Evidence-Based Medicine and Value-Based Purchasing 65
References
Abelson, R., J. Creswell, and G. Palmers. 2012. “Medicare Bills Rise as Records Turn Elec-tronic.” New York Times. Published September 21. www.nytimes.com/2012/09/22/business/medicare-billing-rises-at-hospitals-with-electronic-records.html?_r=0.
Agency for Healthcare Research and Quality (AHRQ). 2016a. “Defining the PCMH. Accessed August 12. https://pcmh.ahrq.gov/page/defining-pcmh.
———. 2016b. “Effective Health Care Program.” Accessed August 12. www.effective healthcare.ahrq.gov/.
———. 2015. “Panel Discussion: Lessons Learned in Using the AHRQ QIs to Improve the Quality and Safety of Care.” PowerPoint slides. Presented December 9. www.qualityindicators.ahrq.gov/Downloads/Resources/Webinars/2015/AHRQ_QI_Impact_Webinar.pdf.
———. 2012. “Public Reporting as a Quality Improvement Strategy. Closing the Quality Gap: Revisiting the State of the Science.” Published July. https://effectivehealthcare.ahrq.gov/ehc/products/343/1198/Evidencereport208_CQG-PublicReporting_ExecutiveSummary_20120724.pdf.
Baiardini, I., F. Braido, M. Bonini, E. Compalati, and G. W. Canonica. 2009. “Why Do Doctors and Patients Not Follow Guidelines?” Current Opinion in Allergy and Clinical Immunology 9 (3): 228–33.
Bohmer, R. M. J. 2005. “Medicine’s Service Challenge: Blending Custom and Standard Care.” Health Care Management Review 30 (4): 322–30.
Burwell, S. M. 2015. “Progress Towards Achieving Better Care, Smarter Spending, Healthier People.” Published January 26. www.hhs.gov/blog/2015/01/26/progress-towards-better-care-smarter-spending-healthier-people.html.
Catalyst for Payment Reform. 2014. “National Scorecard on Payment Reform.” Accessed August 14, 2016. www.catalyzepaymentreform.org/images/documents/national scorecard2014.pdf.
Centers for Medicare & Medicaid Services (CMS). 2016. “Medicare Acute Care Demonstra-tion Project for Orthopedic and Cardiovascular Surgery.” Accessed September 8. www.cms.gov/Medicare/Demonstration-Projects/DemoProjectsEvalRpts/down loads/ACE_web_page.pdf.
———. 2015. “Hospital Value-Based Purchasing.” Modified October 30. www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/hospital-value-based-purchasing/index.html?redirect=/Hospital-Value-Based-Purchasing/.
———. 2011. “Medicare Physician Group Practice Demonstration.” Published July. www.cms.gov/Medicare/Demonstration-Projects/DemoProjectsEvalRpts/downloads/PGP_Fact_Sheet.pdf.
Delbanco, S. 2015. “The Payment Reform Landscape: Everyone Has a Goal.” Pub-lished March 6. http://healthaffairs.org/blog/2015/03/06/the-payment- reform-landscape-everyone-has-a-goal/.
Healthcare Operat ions Management66
Ellis, P., L. G. Sandy, A. J. Larson, and S. L. Stevens. 2012. “Wide Variation in Episode Costs Within a Commercially Insured Population Highlights Potential to Improve the Efficiency of Care.” Health Affairs 31 (9): 2084–93.
Health Catalyst. 2016. “One Healthcare System’s Effective Strategy to Improve Pneu-monia Outcomes.” Accessed August 1. www.healthcatalyst.com/success_stories/reducing-pneumonia-readmissions-multicare.
Institute for Clinical Systems Improvement (ICSI). 2012. “Diagnostic Imaging.” Accessed January 26. www.icsi.org/health_care_redesign_/diagnostic_imaging_35952.
———. 2010. “ICSI High Tech Diagnostic Imaging Enrollment and Next Steps.” Accessed May 18, 2012. www.icsi.org/htdi_slide_presentation__35982/htdi_slide_presentation_. html.
———. 2009. “Transforming High-Tech Diagnostic Imaging: Appropriate, Easy, and Efficient Ordering of Scans at the Point of Care.” Accessed May 18, 2012. www.icsi.org/overview_of_icsi_s_htdi_solution/htdi_brochure.html.
Institute of Medicine. 2011. “Clinical Practice Guidelines We Can Trust.” Published March 23. www.iom.edu/Reports/2011/Clinical-Practice-Guidelines-We-Can-Trust.aspx.
Kaplan, R. S., and D. P. Norton. 2001. The Balanced Scorecard: Translating Strategy to Action. Boston: Harvard Business Review Press.
McLaughlin, D. B. 2015. “Value Purchasing Turns the Corner.” Healthcare Executive 30 (4): 56–58.
Miller, H. D. 2014. Measuring and Assigning Accountability for Healthcare Spending: Fair and Effective Ways to Analyze the Drivers of Healthcare Costs and Transition to Value-Based Payment. Center for Healthcare Quality & Payment Reform. Accessed August 14, 2016. www.chqpr.org/downloads/AccountabilityforHealthcareSpending.pdf.
Millman, J. 2015. “Health Care’s Trillion Dollar Question: How to Define ‘Quality.’” Pub-lished January 30. www.washingtonpost.com/blogs/wonkblog/wp/2015/01/30/the-biggest-challenge-facing-the-governments-new-plan-for-better-health-care/.
National Guideline Clearinghouse (NGC). 2016. Home page. Accessed August 12. https://guideline.gov/.
Patient-Centered Outcomes Research Institute (PCORI). 2014. “National Priorities and Research Agenda.” Updated August 21. www.pcori.org/research-results/research-we-support/national-priorities-and-research-agenda.
Schneider, E. C., P. S. Hussey, and C. Schnyer. 2011. Payment Reform: Analysis of Models and Performance Measurement Implications. Santa Monica, CA: RAND.
Shekelle, P. 2016. “Overview of Clinical Practice Guidelines.” UpToDate. Updated July 28. www.uptodate.com/contents/overview-of-clinical-practice-guidelines.
Stellefson, M., K. Dipnarine, and C. Stopka. 2013. “The Chronic Care Model and Diabe-tes Management in US Primary Care Settings: A Systematic Review.” Preventing Chronic Disease 10: E26.
UnitedHealth Center for Health Reform & Modernization. 2012. “Farewell to Fee-for-Service? A ‘Real World’ Strategy for Health Care Payment Reform.” Published December. www.unitedhealthgroup.com/~/media/UHG/PDF/2012/UNH-Working-Paper-8.ashx.
Chapter 3: Evidence-Based Medicine and Value-Based Purchasing 67
Wagner, E. H., B. T. Austin, C. Davis, M. Hindmarsh, J. Schaefer, and A. Bonomi. 2001. “Improving Chronic Illness Care: Translating Illness into Action.” Health Affairs 20 (6): 64–78.
Wallace, P. J., N. D. Shah, T. Dennen, P. A. Bleicher, and W. H. Crown. 2014. “Optum Labs: Building a Novel Node in the Learning Health Care System.” Health Affairs 33 (7): 1187–94.
Wholey, D. R., M. Finch, N. D. Shippee, K. M. White, J. Christianson, R. Kreiger, B. Wagner, and L. Grude. 2016. “Evaluation of the State of Minnesota’s Health Care Homes Initiative: Evaluation Report for Years 2010–2014. Published February. www.health.umn.edu/sites/default/files/UM%202015%20HCH%20Evaluation%20Final%2007Feb2016.pdf.
Zismer, D. K. 2013. “Physician Compensation in a World of Health System Consolidation and Integration.” Journal of Healthcare Management 58 (2): 87–91.
PART
IISETTING GOALS AND
EXECUTING STRATEGY
CHAPTER
71
STRATEGY AND THE BALANCED SCORECARD
Operations Management in Action
The Malcolm Baldrige National Quality Award is the nation’s highest honor for innovation and performance excellence. In 2008, Poudre Valley Health System (PVHS) was one of three organiza-tions to receive the award and the only healthcare recipient, classifying it as one of the best hospitals in the United States. The Baldrige Award (see chap-ter 2) judges evaluate each healthcare applicant’s performance on a number of dimensions: leadership; strategy; cus-tomer focus; measurement, analysis, and knowledge management; workforce focus; operations; and results.
PVHS is particularly strong in its use of the balanced scorecard to mea-sure its performance and share best practices among departments. The met-rics PVHS uses to track its performance are gathered from the following areas:
• Employee culture
• Market share
• Physician engagement
• Clinical outcomes
• Customer service and patient satisfaction
• Financial performance
Winning the Baldrige Award brings with it an expectation to share
4OVE RVI EW
Most healthcare organizations have good strategic plans; what fre-
quently fails is their execution. This chapter demonstrates how the bal-
anced scorecard can be an effective tool to consistently move strategy
to execution. First, we examine traditional management systems and
explore their failures. Next, we review the theory behind the balanced
scorecard and strategy mapping and explain the tools’ application to
healthcare organizations. Practical steps to implement and maintain
a balanced scorecard system are provided, and detailed examples
from Vincent Valley Hospital and Health System (VVH) demonstrate
the application of these tools. The companion website to this book
contains templates and explanatory videos that can be used for stu-
dent exercises or to implement a balanced scorecard in a healthcare
organization. In addition, a case study on the website includes data
that can be used to develop a realistic dashboard.
This chapter gives readers a basic understanding of balanced
scorecards that enables them to
• explain how a balanced scorecard can be used to move strategy
to action,
• explain how to monitor strategy from the four stakeholder
perspectives,
• identify key initiatives to achieve a strategic objective,
• develop a strategy map that links relevant initiatives,
• identify and measure leading and lagging indicators for each
initiative,
• understand the use of business intelligence tools to extract data
for scorecards, and
• demonstrate the connection of value purchasing metrics to
strategy and execution.
On the web at ache.org/books/OpsManagement3
Healthcare Operat ions Management72
the organization’s journey with the greater community. In accordance with this obligation, PVHS established the Center for Performance Excellence to provide consulting, coaching, and presentation services to other organizations pursuing performance excellence. The Center’s consultants apply the lessons learned over the past decade from the perspective of a Baldrige Award recipient.
Source: Nuwash (2010).
Moving Strategy to ExecutionThe Challenge of ExecutionEnvironmental causes that are commonly cited for the failure to execute in healthcare organizations include intense financial pressures, complex operating structures, and cultures with multistakeholder leadership that resists change. New and redefined relationships among healthcare providers—particularly physicians, hospitals, and health plans—are accompanied by a rapid growth in medical treatment knowledge and technology. Increased public scrutiny of how healthcare is delivered is leading to an associated rise of consumer-directed healthcare. The Affordable Care Act (ACA) is also altering strategy significantly.
No matter how significant these external factors are, however, most organizations founder on internal factors. Outram (2014) identifies a number of internal issues that prevent effective strategy execution in industry at large:
• The leadership team does not understand the strategy.• The leadership team is overconfident. • The organization is incapable of moving with speed and pace.• The organization focuses on short-term goals.• The strategy is too diffuse—it has too many goals.• The communication of strategy to the entire organization is poor.• The strategy is not linked to organizational mission.• Organizational leaders lack accountability.
These factors also plague healthcare organizations. To gain competi-tive advantage from its operations, an organization needs an effective system to move its strategies forward. The management systems of the past are poor tools for today’s challenging environment.
The day-to-day world of a current healthcare leader is intense (exhibit 4.1). Because of ever-present communication technologies (smartphones, e-mail, texts, blogs, social networks), managers float in a sea of inputs and daily barriers.
Healthcare leaders often focus on urgent issues rather than strategy execution. And although organizations can develop effective project managers
Chapter 4: Strategy and the Balanced Scorecard 73
(as discussed in chapter 5), they fail to compete successfully if they do not place the undertaken projects in a broader system of strategy implementation. The balanced scorecard provides a framework and sophisticated mechanisms to move from strategy to execution.
Why Do Today’s Management Tools Fail?Historically, most organizations have been managed with three primary tools: strategic plans, operational reports, and financial reports. Exhibit 4.2 shows the relationships among these tools. In this traditional system, the first step is to create a strategic plan, which is usually updated annually. Next, a budget and operations or project plan is created. The operations plan is sometimes referred to as the tactical plan; it provides a detailed level of task descriptions with timelines and expected outcomes. The organization’s performance is monitored by senior
Balanced scorecardA system of strategy links and reporting mechanisms that supports effective strategy execution.
What’s on your desk today?
Public reporting of quality and
costs
Financial pressure
Today’s urgent operating problem This year’s new
initiatives
Meetings, work/private e-mail, texts, and social
media
Employee turnover—recruiting
Last year’s initiative
EXHIBIT 4.1The Complex World of Today’s Healthcare Leader
Operatingstatistics
Strategicplan
Operations
Managementcontrol
Financialresults
EXHIBIT 4.2The Traditional Theory of Management
Healthcare Operat ions Management74
management through the financial and operational reports. Finally, if deviations from expected performance are encountered, managers take corrective action.
Although theoretically easy to grasp, this management system frequently fails for a number of reasons. Organizations are awash in operating data, and they make no effort to identify key metrics. The strategic plans, financial reports, and operational reports are all created by different departments, and each report is reviewed in different time frames, often by different managers. Finally, none of the reports connect with the others.
These are the root causes of poor execution. If strategies are not linked to action items, operations do not change, nor do the financial results. In addi-tion, strategic plans frequently are not linked to departmental or individual goals and, therefore, simply reside on a shelf in the executive suite.
Many strategic plans contain a logic hole, meaning they lack an explana-tion of how accomplishing a strategic objective provides a specific financial or operational outcome. Consider the following example.
• Strategic objective: to increase the use of evidence-based medicine (EBM)
• Expected outcome: increased patient satisfaction
Although this proposition may seem reasonable on the surface, the logic behind connecting the use of EBM to patient satisfaction is unclear. In fact, patient satisfaction may decrease if providers constantly counsel patients on personal lifestyle issues (e.g., “Will you stop smoking?” “You need to lose weight”); the providers are meeting EBM guidelines, but their patients might see these efforts as bullying or offensive behavior.
The time frame of strategy execution also tends to be problematic. Finan-cial reports are generally timely and accurate but only reflect the current reporting period. A review of these reports does not encourage the long-term strategic allocation of resources (e.g., a major capital expenditure) that may require multiple-year investments. A positive current-month financial outcome is likely the outcome of an action that occurred many months in the past. The cumulative result of these timing problems is poor execution, leading to poor outcomes.
BalanceThe key element of the balanced scorecard is, of course, balance. An organiza-tion can be viewed from many perspectives; to allow a standardized approach, the balanced scorecard methodology uses four common perspectives from which an organization examines its operations (exhibit 4.3):
• Financial stakeholders• Customers
Chapter 4: Strategy and the Balanced Scorecard 75
• Internal process and innovation (operations)• Employee learning and growth
Because an organization is viewed from each perspective, different mea-sures of performance are important. Every perspective in a complete balanced scorecard contains a set of objectives, metrics, targets, and actions. Each mea-sure in each perspective must be linked to the organization’s overall strategy.
The indicators that characterize performance in each of the four per-spectives must be both leading (predicting the future) and lagging (reporting on performance today). Indicators must also be obtained from both inside the organization and the external environment.
Although many think of the balanced scorecard as a reporting tech-nique, its true power lies in its ability to link strategy to action. Balanced scorecard practitioners develop strategy maps that connect projects and actions to outcomes in a series of road map–type graphics. These maps display the “theory of the company” and can be evaluated and fine-tuned as strategies are implemented.
The Balanced Scorecard in Healthcare
The balanced scorecard and its variations have been adopted by leading health-care organizations.
In 2012, Bob McDonald reviewed the use of the balanced scorecard in healthcare and found 87 published studies of healthcare organizations that are effectively using scorecards to improve their competitive marketing position, financial results, and customer satisfaction.
Operationsand
strategicplan
Financialstakeholders
EmployeesOperations
Customers
EXHIBIT 4.3The Four Perspectives in the Balanced Scorecard
Healthcare Operat ions Management76
In this study, he found a number of common success factors in the implementation of balanced scorecards (McDonald 2012):
• Senior management support
• Central involvement of clinicians and some flexibility at lower levels
• Demonstration of empirical benefits
• Cascading [of the balanced scorecard] to lower levels
• Ongoing communication with all staff
• Regular management review and monitoring
• Supporting information technology for monitoring and reporting performance
The Balanced Scorecard as Part of a Strategic Management System
Although it does not substitute for a complete strategic management system, the balanced scorecard is a key component in such a system and an effective tool for moving an organization’s strategy and vision into action. The development of a balanced scorecard leads to the clarification of strategy, and it communi-cates and links strategic measures throughout an organization. Organizational leaders can plan projects, set targets, and align strategic initiatives during the creation of the balanced scorecard. If used properly, the balanced scorecard can also enhance strategic feedback and learning.
Elements of the Balanced Scorecard System
A complete balanced scorecard system has the following elements, which are explained in detail in the subsequent sections:
• Organizational mission and vision, and their relationship to strategy• Perspectives
– Financial – Customer – Internal business process – Learning and growing
• Strategic alignment—linking balanced scorecard measures to strategy• Strategy maps• Implementation of the balanced scorecard, including processes for
identifying targets, resources, initiatives, and budgets• Feedback and the strategic learning process—making sure the balanced
scorecard works
Chapter 4: Strategy and the Balanced Scorecard 77
Mission and VisionThe balanced scorecard system presupposes that an organization has an effective mission, vision, and strategy in place. For example, the mission of VVH is “to provide high-quality, cost-effective healthcare to our community.” Its vision is, “Within five years, we will be financially sound and will be considered the place to receive high-quality care by the majority of the residents of our com-munity.” To accomplish this vision, VVH has identified six specific strategies:
• Recruit five new primary care physicians.• Implement the healthcare home model (also referred to as patient-
centered medical home; see chapter 3).• Expand the VVH accountable care organization. • Increase the volume of obstetric care.• Renegotiate health plan contracts to include performance incentives for
improved chronic disease management.• Improve emergency department (ED) operations and patient
satisfaction.
The VVH example is used throughout this chapter to demonstrate the use of the balanced scorecard. The two strategies examined in depth are increasing the volume of obstetric care and improving ED operations and patient satisfaction.
With an effective strategic plan in place, the next step is to evaluate the plan’s implementation as viewed from each of the four perspectives (financial, customer, operational, and learning and growing). Placing a perspective at the top of a balanced scorecard strategy map means that results in this perspective include the final outcomes desired by an organization. In most organiza-tions, the financial view is the top-most perspective. Therefore, the initiatives undertaken in the other three perspectives should result in positive financial performance for the organization.
“No margin, no mission” is still a valid assessment for nonprofit health-care organizations. They need operating margins to provide financial stability and capital. However, some organizations prefer to position the customer (patient) as the top perspective. In that case, the initiatives undertaken in the other three perspectives are intended to result in positive patient outcomes. (Modifications to the classic balanced scorecard are discussed at the end of this chapter.)
PerspectivesFinancial Perspective Viewed from the financial perspective, the customer, operational, and learn-ing and growing perspectives and their associated initiatives should lead to outstanding financial performance.
Healthcare Operat ions Management78
Although the focus of this book is not directly on healthcare finance, some general strategies should always be under consideration by a hospital or health system.
If the organization is in a growth mode, its financial focus should be placed on increasing revenue to accommodate this growth. If it is operating in a relatively stable environment, the organization may choose to emphasize profitability. If the organization is both stable and profitable, the focus can shift to investment—in both physical assets and human capital. Another major strategy in the financial domain is the diversification of both revenues and expenditures to minimize financial risk.
Exhibit 4.4 lists many common metrics used to measure performance from the financial perspective.
Customer Perspective and Market SegmentationThe second perspective is to view an organization’s operations from the cus-tomer’s point of view. In most healthcare operations, the customer is the patient. Integrated health organizations, however, may operate insurance programs and health plans; some of their customers, then, are employers or the government.
Health insurance exchanges are a new vehicle to connect insurance companies directly with customers. Many hospitals and clinics also consider
• Percent of budget—revenue
• Percent of budget—expense
• Days in accounts receivable
• Days of cash on hand
• Collection rate
• Return on assets
• Expense per relative value unit
• Cost per surgical case
• Case-mix index
• Payer mix
• Growth, revenue, expense, and profit—product line
• Growth, revenue, expense, and profit—department
• Growth of revenue from value purchasing payments
• Growth in members and profitability of accountable care organization
• Growth, revenue, and cost per adjusted patient day
• Growth, revenue, and cost per physician full-time equivalent
• Price competitiveness on selected services
• Research grant revenue
EXHIBIT 4.4Metrics of
Performance from the Financial
Perspective
Chapter 4: Strategy and the Balanced Scorecard 79
their community at large to be the customer. Finally, the physician is seen as the customer in many hospital organizations.
Once the general customers are identified, a helpful step is to segment them into smaller groups and determine the value proposition that will be delivered to each. Examples of market segments are patients with chronic ill-nesses (e.g., diabetes, congestive heart failure); patients seeking obstetric care, sports medicine services, cancer care, or emergency care; Medicaid patients; small employers; and referring primary care physicians.
Customer MeasuresOnce market segments have been determined, a number of traditional mea-sures of marketplace performance may be applied, the most prominent being market share. Customers should be individually tracked and measured in terms of both retention and acquisition, as retaining an existing customer is always easier than attracting a new one. Customer satisfaction and profitability are also useful measures. Exhibit 4.5 displays a number of common customer metrics.
Customers: The Value PropositionOrganizations create value to retain current customers and attract new ones. Each market segment may require products to have different attributes to
Value propositionA marketing term summarizing the relative cost, features, and quality of a service or good.
• Patient care volumes – By service, type, and physician – Turnover—new patients and those exiting the system
• Physician – Referral and admission rates – Satisfaction – Availability of resources (e.g., operating suite time)
• Market share by product line
• Clinical measures – Readmission rates – Complication rates – Compliance with evidence-based guidelines – Medical errors
• Customer service – Patient satisfaction – Waiting time – Cleanliness, ambience – Ease of navigation – Parking – Billing complaints
• Reputation
• Price comparisons relative to competitors
EXHIBIT 4.5Metrics of Performance from the Customer Perspective
Healthcare Operat ions Management80
maximize that segment’s particular value proposition. For example, the orga-nization may seek to be a price leader for outpatient imaging, as some patients will pay for this service via a healthcare savings account. For another seg-ment—emergency services, for example—speed of delivery may be critical. The personal relationship of provider to patient may be important in primary care but not as important in anesthesiology.
Image and reputation are particularly strong influences in consumer behavior and can be competitive advantages for specialty healthcare services. Taking care to understand the value proposition in an organization can lead to the development of effective metrics and strategy maps in the balanced scorecard system.
Vincent Valley Hospital and Health System’s Value PropositionVVH has developed a value proposition for its obstetric services. Its market segment is pregnant women aged 18 to 35. VVH believes the product attri-butes for this market should be
• quick access to care;• warm and welcoming facilities;• customer interactions characterized by strong and personal relationships
with nurses, midwives, and doctors; and• an image of high-quality care that is supported by an excellent system
for referrals and air transport for high-risk deliveries.
VVH has determined the following metrics to measure each attribute:
• The time from arrival to care in the obstetric suite• A patient survey of facility attributes• A patient survey of satisfaction with staff• The percentage of high-risk newborns referred and transported, and the
clinical outcomes of these patients
The main value proposition for emergency care has been identified as reduced waiting time. Following internal studies, competitive benchmarking, and patient focus groups, VVH has determined that its goal is to have fewer than 10 percent of its ED patients wait more than 30 minutes for care.
Internal Business Process PerspectiveThe third perspective in the balanced scorecard is internal business processes or operations—the primary focus of this book. The internal business process perspective has three major components: innovation, ongoing process improve-ment, and post-sale service.
Chapter 4: Strategy and the Balanced Scorecard 81
InnovationAny well-functioning healthcare organization has in place a purposeful innova-tion process. However, many hospitals and health systems today do not, and they can only be characterized as reactionary. They simply respond to—rather than anticipate—new reimbursement rules, government mandates, or technologies introduced through the medical staff. Bringing thoughtful innovation into the life cycle is one of the most pressing challenges contemporary organizations face.
The first step in an organized innovation process is to identify a potential market segment. Then, two primary questions must be answered: (1) What benefits will customers value in tomorrow’s market? (2) How can the orga-nization innovate to deliver those benefits? Once these questions have been researched and answered, related products can be created.
Quality function deployment (chapter 9) can be a useful tool for new product or service development. If a new service is on the clinical leading edge, it may require additional research and testing. A more mainstream service calls for competitor research and review of the clinical literature. The principles of project management (chapter 5) should be used throughout this process until the new service is operational and stable. The process of innovation and design thinking is explored in more depth in chapter 5.
Standard innovation measures used in many industries outside healthcare include percentage of total sales resulting from new products and proprietary products, number of new product introductions per year, time to develop new products, and time to break even.
Healthcare operations tend toward stability (bordering on being rigid), and therefore, a major challenge is simply ensuring that all clinical staff use the latest and most effective diagnostic and treatment methodologies. However, with the passage of the ACA, those organizations with a well-functioning product development process have a clear competitive advantage.
Ongoing Process ImprovementThe case for process improvement and operations excellence is made throughout this book. The project management system (chapter 5) and process improve-ment tools (chapters 6 through 11) are key to these activities. The strategic effect of process improvement and maintaining gains is discussed in chapter 15.
Post-sale ServiceThe final aspect of the operations perspective is the post-sales area, an element that is poorly executed in most healthcare delivery organizations. Sadly, the most common post-sale contact with a patient may be an undecipherable or incorrect bill.
Good post-service systems provide patients with follow-up information on the service they received. Patients with chronic diseases should be contacted periodically with reminders on diet, medication use, and the need to schedule
Healthcare Operat ions Management82
follow-up visits. An outstanding post-sale system also finds opportunities for improvement in the service as well as possible innovations for the future. Open-ended survey questions such as, “From your perspective, how could our organization improve?” or “How else can we serve your healthcare needs?” can point to opportunities for improvement and innovation. Exhibit 4.6 lists common metrics used to measure operational performance.
External Operational Metrics Today and into the Future We pause in our discussion of the elements of the strategic plan to revisit value purchasing, specifically in terms of its influence on the business process perspec-tive. Value purchasing (or value-based purchasing, as it is often referred to) emphasizes meeting external goals and benchmarks. This emphasis complicates strategy maps; the metrics from the Centers for Medicare & Medicaid Services (CMS) alone number more than 1,700 (IOM 2015).
In 2019, CMS will implement the Merit-Based Incentive Payment Sys-tem (MIPS) for physician compensation. As discussed in chapter 3, because MIPS introduces many new metrics and publicly reported quality measures, organizations might be tempted to develop a strategy that directly links physician
• Average length of stay—case-mix adjusted
• Full-time equivalent (FTE)/adjusted patient day
• FTE/diagnosis-related group
• FTE/relative value unit
• FTE/clinic visit
• Waiting time inside clinical systems
• Access time to appointments
• Percent value-added time
• Utilization of resources (e.g., operating room, imaging suite)
• Patients leaving emergency department without being seen
• Operating room cancellations
• Admitting process performance
• Billing system performance
• Medication errors
• Nosocomial infections
• Measures from external agencies: The Joint Commission (2016), the National Quality Forum (2016), and the Centers for Medicare & Medicaid Services (2016).
• National Quality Forum (2002) “never events”
EXHIBIT 4.6Metrics of
Performance from the
Operational Perspective
Chapter 4: Strategy and the Balanced Scorecard 83
payment to MIPS metrics (which may already be happening in some small practices).
The proliferation of metrics might also tempt an organization to develop overly complex scorecards. These data visualizations are not a substitute for a disciplined strategy featuring a strategy map that can be communicated to the entire organization and effectively executed.
Vincent Valley Hospital and Health System Internal Business ProcessesVVH is executing four major projects to move its birthing center and ED strategies forward. The birthing center projects include remodeling and redeco-rating labor and delivery suites, contracting with a regional health system for emergency transport of high-risk deliveries, and introducing predelivery tours of labor and delivery facilities by nursing staff. The ED project is to execute a Lean analysis and kaizen event to improve patient flow.
Learning and Growing PerspectiveThe final perspective from which to view an organization is employee learning and growth. To effectively execute a strategy, employees must be motivated and have the necessary tools to succeed. Therefore, a high-performing organization makes substantial investments in this aspect of its operations. Kaplan and Norton (1996) identified three critical aspects of learning and growing: employee skills and abilities, necessary information technology (IT), and employee motivation.
Employee Skills and AbilitiesAlthough employees in healthcare usually come to their jobs with general training in their technical field, continuous updating of skills is necessary. Some healthcare organizations are effective in ensuring that clinical skills are updated but neglect training in other vital processes (e.g., purchasing systems, organization-wide strategies). A good measure of the attention paid to this area is the number of classes conducted by the organization (or an outside education vendor) for the staff. Another important measure is the breadth of
Kaizen and Kaizen EventsKaizen is the Japanese term for “change for the better,” or continuous improvement. Kaizen has become the vehicle by which Lean systems make changes and improve. The philosophy of kaizen involves all employees in making suggestions for improvement, then implementing those suggestions quickly. It is based on the assumptions that everything can be improved and that many small incremental changes result in an enhanced system.
A kaizen event, sometimes referred to as a rapid process improvement workshop, is a focused, short-term project aimed at improving a particular process.
Healthcare Operat ions Management84
employee occupations attending these classes. Do all employees—from doctors to housekeepers—attend organization-wide training?
Necessary IT Most healthcare workers are considered knowledge workers. They primarily use thinking to accomplish the goals of their profession, as opposed to physical labor. The more immediately and conveniently they can obtain information, the more effectively they can perform their jobs. Facilitative IT is one key to this ability.
Process redesign projects frequently use IT as a resource for automa-tion and information retrieval. Measures of automation include the number of employees having easy access to IT systems, the percentage of individual jobs that have an automation component, and the speed of installation of new IT capabilities. The use of data and analytics is explored in depth in chapter 8.
Employee MotivationA progressive culture and motivated employees are clearly competitive advan-tages; therefore, the organization must monitor these areas with some frequency. Measures of employee satisfaction include the following:
• Level of involvement in decision making• Recognition for doing a good job• Amount of access to information• Level of encouragement of creativity and initiative• Support for staff-level functions• Overall satisfaction with the organization• Turnover rate• Absenteeism rate• Training hours per employee
Data for many of these measures are typically collected through employee surveys.
These three aspects of learning and growing—employee skills, IT, and motivation—all contribute to employee satisfaction. A satisfied employee is productive and tends to remain with the organization. Employee satisfaction, productivity, and loyalty make outstanding organizational performance possible.
Vincent Valley Hospital and Health System Learns and GrowsVVH realizes its employees need new skills to successfully execute some of its projects, so it has engaged training firms to provide classes for all staff. Exhibit 4.7 illustrates this undertaking for improvement.
Chapter 4: Strategy and the Balanced Scorecard 85
Strategic Alignment: Linking Measures to StrategyOnce expected objectives and their related measures are determined for each perspective, the initiatives to meet these goals must be developed. An initiative can be a simple action or a large project. Regardless of its scale, each initiative must be logically linked to the desired outcome through a series of cause-and-effect statements. These are usually constructed as “if–then” statements that tie initia-tives together and contribute to the outcome, as with the following examples:
• If the wait time in the ED is decreased, then the patient will be more satisfied.
• If an admitting process is improved through the use of automation, then the final collection rate will improve.
• If an optically scanned wristband is used in conjunction with an electronic health record, then medication errors will decline.
• If a discharge summary is routinely dictated and transmitted to the primary care provider within 24 hours, then the number of readmissions within 30 days will decrease.
Each initiative should have measures associated with it, and every measure selected for a balanced scorecard should be an element in a chain of cause–effect relationships that communicates the organization’s strategy.
Outcomes and Performance DriversSelecting appropriate measures for each initiative is critical. Measures can be categorized into two basic types of indicators. Outcome indicators, familiar to most managers, are also termed lagging indicators because they result from earlier actions. Outcome indicators tend to be generic instead of tightly focused. Healthcare operations examples include profitability, market share, and patient satisfaction. The other type of indicator is a performance driver, or
Lagging indicatorA performance measurement that assesses the outcome of existing actions.
Project Employees Involved Training
Begin predelivery tours of labor and delivery facilities by nursing staff
Obstetric nursing and support staff
Customer service and sales
Execute a Lean analysis and kaizen event to improve patient flow in the emergency department
Managers and key clinicians in the emergency department
Lean tools (chapter 10)
EXHIBIT 4.7VVH Improvement Projects and Associated Training
Healthcare Operat ions Management86
leading indicator. These indicators predict the future and are specific to an initiative and the organization’s strategy. One example of a performance driver is waiting time in the ED. A drop in waiting time should predict an improve-ment in a related outcome indicator, such as patient satisfaction.
A common pitfall in developing indicators is the use of measures associ-ated with the improvement project rather than with the process improvement. For example, the fact that a project to improve patient flow in a department is 88 percent complete is a less adequate indicator than a measure of the actual change in patient flow, a 12 percent reduction in waiting time. Outcome measures are always preferred, but in some cases they may be difficult or impossible to obtain.
Because the number of balanced scorecard measures should be lim-ited—ideally to fewer than 20—identifying measures that are indicators for a complex process is sometimes useful. For example, a seemingly simple indicator such as time to next appointment for patient scheduling actually tracks many complex processes in an organization.
Strategy MapsAs discussed, a set of initiatives should be linked together by if–then statements to achieve a desired outcome. Both outcome and performance driver indicators should be determined for each initiative. These can be displayed graphically in a strategy map, which may be most helpfully organized into the four perspec-tives, where learning and growing is positioned at the bottom and financial resides at the top. A general strategy map for any organization includes the following conditional statements:
• If employees have skills, tools, and motivation, then they will improve operations.
• If operations and marketing efforts are improved, then customers will buy more products and services.
• If customers buy more products and services and operations are run efficiently, then the organization’s financial performance will improve.
Exhibit 4.8 shows a strategy map in which these general initiatives are indicated.
The strategy map is enhanced if each initiative also contains the strategic objective, measure used, and results that the organization hopes to achieve (targets). Each causal pathway from initiative to initiative needs to be as clear and quantitative as possible.
Vincent Valley Hospital and Health System Strategy MapsVVH has two major areas of strategic focus—the birthing center and the ED. Exhibit 4.9 displays the strategy map for the birthing center.
Leading indicatorA performance measurement that predicts the future and is specific to an initiative or organizational strategy. Also called performance driver.
Strategy mapA set of initiatives that are graphically linked by if–then statements to describe an organization’s strategy.
Chapter 4: Strategy and the Balanced Scorecard 87
Improve marketing and customerservice
Improve financial results
Improve operations
Provide employeeswith skills, tools, andmotivation
Learningand
Growing
BusinessProcesses
Customers
Financial
EXHIBIT 4.8General Strategy Map
Learningand
Growing
BusinessProcesses
Customers
FinancialIncrease net revenue of obstetric product line Goal = 10%
Measure market shareGoal = 5% increase
Measure patient satisfaction(facilities)Goal
^
90% satisfaction
Remodel obstetric suiteGoal = complete byNovember 1
Measure patient satisfaction (perceived clinical quality)Goal
^
90% satisfaction
Contract for emergencytransportationGoal = 10 runs/month
Measure patient satisfaction (high touch)Goal
^
90% satisfaction
Begin tours and surveyGoal = patientsatisfaction
^
90%
Customer service trainingGoal = 90% averagepassing score
EXHIBIT 4.9VVH Birthing Center Strategy Map
Healthcare Operat ions Management88
Recall that VVH has decided to execute three major projects in this area. Other initiatives needed for the successful execution of each project are identi-fied on the map. For instance, for nursing staff to successfully lead expectant mothers on tours of labor and delivery suites, the staff must participate in a customer service training program. After the tours begin, the birthing center will measure potential patients’ satisfaction to ensure that the tours are being conducted effectively.
After patients deliver their babies in VVH’s obstetric unit, they will again be surveyed on their experience, with special questions on the effect of each major project. These leading satisfaction indicators should predict the lagging indicators of increased market share and net revenue.
The second major strategy for VVH is to improve patient flow in the ED. Exhibit 4.10 shows the strategy map for the department.
The first required steps in this strategy are forming a project team (chap-ter 5) and learning how to use Lean process improvement tools (chapter 10). Then the team can begin analyzing patient flow and implementing changes to improve flow. VVH has set a goal of reducing the amount of non-value-added time by 30 percent. From the time this goal is first met, waiting time for 90 percent of patients should not exceed 30 minutes. A reduced waiting time should result in patients being more satisfied and, hence, a growth in market share and increased net revenue. Following are more formal cause-and-effect statements:
Measure patientwait timeGoal ^ 30 minutes
Measure patient shareGoal = 5% increase
Increase net revenue of emergency departmentproduction lineGoal = 10%
Conduct project on patient flow andmake changesGoal = value stream increased by 30%
Learn Lean processimprovement toolsGoal = complete by December 1
Learningand
Growing
BusinessProcesses
Customers
Financial
EXHIBIT 4.10VVH Emergency
Department Strategy Map
Chapter 4: Strategy and the Balanced Scorecard 89
• If ED staff undertake educational activities to learn project management and Lean, then they can effectively execute a patient flow improvement project.
• If a patient flow project is undertaken and non-value-added time is reduced by 30 percent, then the waiting time for 90 percent of the patients should never exceed 30 minutes.
• If the waiting time for most patients never exceeds 30 minutes, then they will be highly satisfied, and this satisfaction will increase the number of patients and VVH’s market share.
• If the ED market share increases, then net revenue will increase.
The book’s companion website contains a downloadable strategy map and linked scorecard. It also includes a number of videos that demonstrate how to use and modify these tools for both student and practitioner use.
Implementation of the Balanced ScorecardLinking and CommunicatingThe balanced scorecard can be used at many different levels in an organization. However, departmental scorecards should link to the divisional, and ultimately the corporate, level. Each scorecard should be linked upward and downward. For example, an obstetric initiative to increase revenue from normal childbirths should be linked to the corporate-level objective of overall increased revenue.
Sometimes, specific linkages are difficult to establish between a depart-mental strategy map and corporate objectives. In these cases, the department head must derive a more general link by stating how a departmental initiative will influence a particular corporate goal. For example, improving the quality of the hospital laboratory testing system generally affects the corporate objec-tive that patients should perceive that the hospital provides the highest level of quality care.
The development and operation of scorecards at each level of an orga-nization require disciplined communication, which can be an incentive for action. Balanced scorecards can also be used to communicate with an organiza-tion’s external stakeholders. A well-implemented balanced scorecard system is integrated with individual employee goals and the organization’s performance management system.
Targets, Resources, Initiatives, and BudgetsAs demonstrated in this chapter, a balanced scorecard strategy map consists of a series of linked initiatives, and each initiative should have a quantitative measure and a target. Initiatives can reside in one department, but they are
On the web at ache.org/books/OpsManagement3
Healthcare Operat ions Management90
frequently cross-departmental. Many initiatives are projects, and the process for successful project management (chapter 5) should be followed.
A well-implemented balanced scorecard also links carefully to an organi-zation’s budget, particularly if initiatives and projects are expected to consume considerable operating or capital resources.
The use of the balanced scorecard does not obviate the need for addi-tional operating statistics. Many other operating and financial measures still must be collected and analyzed. If the performance of any of these measures deviates substantially from its target, a new strategy and initiative may be needed. For example, most healthcare organizations carefully track and monitor their accounts receivable. If this financial measure is within industry norms, it prob-ably will not appear on an organization’s balanced scorecard. However, if the accounts receivable balance drifts over time and begins to exceed expectations, a balanced scorecard initiative may be started to address the problem.
Displaying ResultsThe actual scorecard tracks and communicates the results of each initiative. (Chapter 7 provides several examples of visual displays.) A challenge for most organizations is to collect the data to display in the scorecard. Because the scorecard should include fewer than 20 measures, a simple solution is to assign this responsibility to one individual who develops efficient methods to collect the data and determines effective methods by which to display them. A more robust solution is to develop a data warehouse with associated analysis and reporting tools (see exhibit 4.11).
Does the Balanced Scorecard Work? Feedback and Strategic LearningOnce a balanced scorecard system is created, it must be monitored closely. Management teams should divide their routine meetings into three types: operational reviews, strategy reviews, and strategy testing and adaptation. The operational meeting is held frequently (e.g., weekly) and is designed to respond to short-term problems and promote improvements. The strategy review meet-ing is held monthly and focuses on monitoring and fine-tuning the existing strategy map. The strategy testing and adaptation meeting should be held at least annually—more frequently if the business environment is changing rapidly. These meetings are designed to improve or transform the existing strategy, develop new initiatives and revise maps, and authorize needed expenditures.
The explicit purpose of the balanced scorecard is to ensure the success-ful execution of an organization’s strategy. But what if it does not achieve the desired results? Two possible causes can be at play.
The first, most obvious, problem is that an initiative itself is not achiev-ing its targeted results. For example, the ED’s patient flow project may not be able to decrease non-valued-added time by 30 percent. In that case, the
Chapter 4: Strategy and the Balanced Scorecard 91
0
100,000
200,000
300,000
400,000
500,000
600,000
700,000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0%
20%
40%
60%
80%
100%
120%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0%
5%
10%
15%
20%
25%
30%
35%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0
20
40
60
80
100
120
140
160
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0
50
100
150
200
250
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
82%84%86%88%90%92%94%96%98%
100%102%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0%
20%
40%
60%
80%
100%
120%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Net revenue
Patient satisfaction
Increase in admissions
Customer service training tests
Profit
FT admission (%)
Cost/unit
Six Sigmatraining tests
YTD
Actual 2,877,842
Goal 3,266,667
YTD
Actual 88%
Goal 90%
YTD
Actual 83
Goal 100
YTD
Actual 94%
Goal 90%
YTD
Actual 3.2%
Goal 3.0%
YTD
Actual 16%
Goal 30%
YTD
Actual 134
Goal 115
YTD
Actual 86%
Goal 90%
Note: Download this scorecard from the book’s companion website at ache.org/books/ OpsManagement3. FT = full time; YTD = year to date.
EXHIBIT 4.11Balanced Scorecard Template
Healthcare Operat ions Management92
hospital may need to add an initiative, such as engaging a consultant. This measure must be carefully monitored and frequently posted on the scorecard.
The second, more complex, problem occurs when the successful execu-tion of an initiative does not lead to achievement of the next linked target. For example, although waiting times in the ED decrease, the department does not gain market share. The first step in solving this problem is to reconsider the cause-and-effect relationships.
An organization should review its results and strategy map at least quar-terly and revise its strategy annually, usually as part of the budgeting process.
Modifications of the Classic Balanced ScorecardThe balanced scorecard has been modified by many healthcare organizations, most commonly by placing the customer or patient at the top of the strategy map (exhibit 4.12). Finance then becomes a means to achieve superior patient outcomes and satisfaction.
Implementation IssuesTwo common challenges arise when implementing balanced scorecards: (1) determination and development of metrics, and (2) initiative prioritization.
The balanced scorecard is a quantitative tool and, as such, requires data systems that generate timely information for inclusion. Each initiative on
Improve operations
Improve patient results and satisfaction
Improve availability of financial resources
Provide employeeswith skills, tools, andmotivation
Learningand
Growing
BusinessProcesses
Customers
Financial
EXHIBIT 4.12Inverted General
Strategy Map
Chapter 4: Strategy and the Balanced Scorecard 93
a strategy map should have quantitative measures, which should represent an even mix of leading and lagging indicators. Each initiative should have a target as well. However, setting targets is an art: Too timid a goal does not move the organization forward, and too aggressive a goal is discouraging for staff.
A number of sources should be used to construct targets. They include internal company operating data, executive interviews, internal and external strategic assessments, customer research, industry averages, and benchmarking data. Targets can be incremental on the basis of current operating results (e.g., increase productivity in a nursing unit by 10 percent in the next 12 months), or they can be “stretch goals,” which are possible to achieve but require extraor-dinary effort (e.g., improve compliance with evidence-based guidelines for 98 percent of patients with diabetes). Including too many measures and initiatives renders a scorecard confusing; therefore, even the most sophisticated organiza-tions limit their measures to 20 or fewer.
Achieving perfect alignment with a balanced scorecard’s goals for all of an organization’s initiatives is difficult. However, the closer the alignment, the more likely the organization’s strategic objectives will be achieved.
Conclusion
This text is about how to get things done. The balanced scorecard with strategy mapping provides a powerful tool toward that end because it
• links strategy to action in the form of initiatives;• provides a comprehensive communication tool inside and outside an
organization; and• is quantitatively based, providing a vehicle for ongoing strategy analysis
and improvement.
Discussion Questions
1. What other indicators might be used in each of the four perspectives for public health agencies? For health plans?
2. If you were to add a perspective to the four discussed in the chapter, what would it be? Draw a strategy map of a healthcare delivery organization and include this perspective.
3. How do you manage the other operations of an organization—that is, those that do not appear on a strategy map or balanced scorecard?
4. How would a department link its balanced scorecard to the corporate scorecard?
Healthcare Operat ions Management94
5. What methods could be used to involve the customer or patient in identifying the key elements of the balanced scorecard?
Exercises
1. View the videos at the companion website for this book, and download the PowerPoint strategy map provided. Develop a strategy map and
balanced scorecard for a primary care dental clinic. Conduct Internet research to determine the challenges facing primary care dentistry, and develop a strategy map for success in this environment.
Make sure the strategy map includes at least eight initiatives and that they touch on the four perspectives. Include targets, and be sure the metrics are a mix of leading and lagging indicators. Develop a plan to periodically review your map to ascertain its effectiveness.
2. Download the data for this chapter provided on the companion website, and develop a dashboard in Excel to identify readmissions that occur within 30 days of discharge. A number of initiatives are described on the website to minimize readmissions. Conduct additional Internet research and construct a strategy map to improve this readmission rate.
References
Centers for Medicare & Medicaid Services (CMS). 2016. “Quality Measures.” Modified February 14. www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/QualityMeasures/index.html?redirect=/QUALITYMEASURES/.
Institute of Medicine (IOM). 2015. Vital Signs: Core Metrics for Health and Health Care Progress. Published April 28. http://iom.nationalacademies.org/Reports/2015/Vital-Signs-Core-Metrics.aspx#sthash.NIyVcAPm.dpuf.
Joint Commission, The. 2016. “Core Measure Sets.” Accessed February 10. www.jointcom-mission.org/core_measure_sets.aspx.
Kaplan, R. S., and D. P. Norton. 1996. The Balanced Scorecard: Translating Strategy into Action. Boston: Harvard Business School Press.
McDonald, B. 2012. “A Review of the Use of the Balanced Scorecard in Healthcare.” Pub-lished April. www.bmcdconsulting.com/index_htm_files/Review%20of%20the%20Use%20of%20the%20Balanced%20Scorecard%20in%20Healthcare%20BMcD.pdf.
National Quality Forum (NQF). 2016. “Measures, Reports & Tools.” Accessed February 10. www.qualityforum.org/Measures_Reports_Tools.aspx.
On the web at ache.org/books/OpsManagement3
Chapter 4: Strategy and the Balanced Scorecard 95
———. 2002. “Serious Reportable Events in Healthcare: A National Quality Forum Con-sensus Report.” Publication No. NQFCR01-02. Washington, DC: NQF.
Nuwash, P. 2010. “Transforming Change in Health Services Performance Through Business Intelligence.” Presentation at Midwest Healthcare Business Intelligence Summit, Bloomington, Minnesota, October 19.
Outram, C. 2014. “Ten Pitfalls of Strategic Failure.” INSEAD Blog. Published March 17. http://knowledge.insead.edu/blog/insead-blog/ten-pitfalls-of-strategic-failure-3225.
Further Reading
Cleverley, W. O., and J. O. Cleverley. 2005. “Scorecards and Dashboards.” Healthcare Financial Management 59 (7): 64–69.
Kaplan, R. S., and D. P. Norton. 2008a. The Execution Premium: Linking Strategy to Opera-tions for Competitive Advantage. Boston: Harvard Business School Publishing.
———. 2008b. “Mastering the Management System.” Harvard Business Review 86 (1): 62–77.———. 2006. “How to Implement a New Strategy Without Disrupting Your Organiza-
tion.” Harvard Business Review 84 (3): 100–109.———. 2001. The Strategy-Focused Organization: How Balanced Scorecard Companies
Thrive in the New Business Environment. Boston: Harvard Business School Press.Tarantino, D. P. 2003. “Using the Balanced Scorecard as a Performance Management Tool.”
Physician Executive 29 (5): 69–72.Wyatt, J. 2004. “Scorecards, Dashboards, and KPIs: Keys to Integrated Performance Mea-
surement.” Healthcare Financial Management 58 (2): 76–80.
CHAPTER
97
PROJECT MANAGEMENT
Operations Management in Action
The examples in the Operations Manage-ment in Action sections throughout this book generally demonstrate the effec-tive use of the principles of operations management. However, professionals can also learn from failures. One of the most visible, and nearly catastrophic, fail-ures in healthcare operations manage-ment was in the implementation of the Affordable Care Act’s health insurance exchanges by the US federal government.
Although most of the operational issues with the exchanges have now been corrected, at the outset of the implementa-tion, many of the principles of good project management were not employed. Following is a list of poor or inadequate approaches compiled by an experienced governmen-tal project manager that demonstrate the absence of good operations management (adapted from Thomson 2013):
Unrealistic requirements. This is the first time anybody has ever tried to develop a single website where diverse users could (1) establish an online identity, (2) review hundreds of health-insurance options, (3) enroll in a specific plan, and (4) determine eligibil-ity for federal subsidies—all in real time.
Technical complexity. As often occurs with poorly planned [defense] projects, unrealistic requirements for
5OVE RVI EW
Everyone manages projects, whether painting a bedroom at home or
adding a 100-bed wing to a hospital. This chapter provides grounding
in the science of project management. The major topics covered include
• selecting and chartering projects;
• using stakeholder analysis to set project requirements;
• developing a work breakdown structure and schedule;
• using Microsoft Project to develop project plans and monitor
cost, schedule, and earned value;
• managing project communications, change control, and risk; and
• creating and leading project teams.
After reading this chapter and completing the associated
exercises, readers should be able to
• create a project charter with a detailed plan for costs, schedule,
scope, and performance;
• monitor the progress of a project, make changes as required,
communicate to stakeholders, and manage risks; and
• develop the skills to successfully lead a project team.
If virtually everyone has had experience managing projects,
why devote a chapter in a healthcare operations book to the topic?
The answer lies in the question. Although everyone has life experi-
ences in project management, few healthcare professionals take the
time to understand and practice the science and discipline of project
management. The ability to successfully move a project forward while
meeting time and budget goals is a distinguishing characteristic of a
high-quality, highly competitive healthcare organization.
Effective project management provides an opportunity for
progressive healthcare organizations to quickly develop new clinical
services, fix major operating problems, reduce expenses, and provide
new consumer-directed products to their patients.
Healthcare Operat ions Management98
HealthCare.gov resulted in an extraordinarily complicated system that is difficult to main-tain. There are just too many moving pieces.
Integration re- sponsibility. Despite weak internal [infor-mation technology (IT)] capabilities, [the Centers for Medicare & Medicaid Services (CMS)] decided it would take charge of integrating all the parts in HealthCare.gov, and testing the end product to ensure functionality. The results show why the military almost always hires outside companies to serve as lead integrator. The final resolution of the problems of Health-Care.gov was led by an outside consultant.
Fragmented au- thority. There seems to have been a great deal of infighting at CMS over how the website would operate and what the user experi-ence would feel like. With three different parts of the bureau-cracy contending for control—the IT shop, the policy shop, and the communications
OVE RVI EW (Continued)
Project management as a formalized management meth-
odology came of age in the period 1958–1979. New management
science mathematical tools, such as program evaluation and review
technique (PERT) and the critical path method (introduced in chap-
ter 2 and discussed later in this chapter), were developed. In addi-
tion, the rapid development of computer systems, such as the
minicomputer, made the use of these tools accessible to project
managers (Azzopardi 2016).
Project management as a discipline continued to develop
over time, culminating in the establishment of the Project Manage-
ment Institute (PMI) in 1969 (www.PMI.org). As of 2014, PMI had
more than 2.9 million members, and more than 650,000 of those
members were certified as project management professionals
(PMPs) (PMI 2015).
PMI publishes the Project Management Body of Knowl-
edge (PMBOK) (PMI 2013), which details best practices for suc-
cessful project management.
In addition to the work of PMI, Carden and Egan (2008)
undertook a comprehensive review of the scientific basis for project
management. According to their study,
refereed research has indicated that project managers utilize
tools and techniques along with people to ensure quality
deliverables are on time, within scope, and within budget.
Additionally, project leadership and a favorable development
environment both are important to the successful delivery
of projects. Therefore, the connection between knowledge
and action can be used to frame behaviors by engaging in
transactions to plan, organize, monitor, and report findings
in order to maintain a dynamic balance with the organization,
resources, tools, and the external environment.
Much as evidence-based medicine delineates the most
effective methods to care for specific clinical conditions, PMBOK
provides science-based, field-tested guidelines for successful
project management. This chapter is based on PMBOK principles as
applied to healthcare. Healthcare professionals who spend much of
their time leading projects should consider using resources avail-
able through PMI; for some, PMP certification may be appropriate.
Chapter 5: Project Management 99
shop—key decisions were often delayed, guidance to contractors was inconsistent, and nobody was truly in charge.
Loose metrics. Perhaps the most important factor in keeping complex projects on track is for managers to utilize rigorous, unambiguous performance metrics in measuring progress. Absence of reliable metrics helps explain why federal officials didn’t realize until late in the game that HealthCare.gov might not be ready for prime time.
Inadequate testing. Despite repeated warnings from contractors that more testing of system components was needed, CMS was determined to see the site go live on its planned debut date of October 1.
Aggressive schedules. You wouldn’t think that standing up a website after literally years of planning might entail overly aggressive schedules, but in the case of HealthCare.gov the disorganized bureaucracy took so long to make design choices that the back end of the project was way too hurried for comfort.
Administrative blindness. [CMS] may not have had good management prac-tices or metrics for identifying problems, but that doesn’t mean it didn’t get plenty of warnings about potential problems with HealthCare.gov. Outside consultants and contractors on the project repeatedly warned government officials about functional difficulties with some features of the site, lack of adequate testing, poor protection of sensitive information, and the like.
Definition of a Project
A project is a one-time set of activities that culminates in a desired outcome. Therefore, activities that occur repeatedly—for example, making appointments for patients in a clinic—are not projects. However, the installation of new software to upgrade the appointment-making capability is a project, as is a major process improvement effort to reduce telephone hold time for patients.
Slack (2005) provides a useful tool for determining the need for formal project management (exhibit 5.1). Operational issues arise frequently; if they are simple, they can be fixed immediately by operations staff. More difficult problems can be addressed by using the tools detailed in chapter 6. However, projects that are complex and have high organizational value need the discipline of formal project management. Many of the strategic initiatives on an organiza-tion’s balanced scorecard should use the project management methodology.
A well-managed project includes
• a specified scope of work, • expected outcomes and performance levels, • a budget, • a detailed work breakdown tied to a schedule,
Healthcare Operat ions Management100
• a formal change procedure, • a communications plan, • a plan to deal with risk, • a project conclusion process, and • a plan for redeployment of staff.
Many high-performing organizations also have a formal, executive-level chartering process for projects and a project management office to monitor enterprisewide project activities. Some healthcare organizations (e.g., health plans) may have a substantial share of their operating resources invested in projects at any one time.
For effective execution of a project, PMI recommends that three ele-ments be in place. A project charter begins the project and addresses stakeholder needs. A project scope statement identifies the project outcomes, timelines, and budget in detail. Finally, a project plan is developed and includes scope man-agement, work breakdown, schedule management, cost management, quality control, staffing management, communications, risk management, procure-ment, and the closeout process. Exhibit 5.2 displays the relationships among these elements.
Project Selection and CharteringProject SelectionMost organizations have many projects vying for attention, funding, and senior executive support. The annual budget and strategic planning processes serve
Find it ,fix it
Problem-solvingprocess
Level ofdetail andproblemsolving
Projectmanagement
Complex
Simple
Source: Slack (2005). Used with permission.
EXHIBIT 5.1When to
Use Project Management
Chapter 5: Project Management 101
as useful vehicles for prioritizing projects in many organizations. The balanced scorecard (chapter 4) helps guide the identification of worthwhile strategic projects. Other external forces (e.g., new Medicare rules) or clinical innovations (e.g., new imaging technologies), however, conspire to present an organization’s leadership with a list of projects too long for successful implementation. When selection dilemmas present themselves, consider using a quantitative approach, such as the example provided in exhibit 5.3. In this case, each possible project is scored on six measures, including the four balanced scorecard perspectives with a predetermined weighting.
To use this tool, each potential project should be scored by a senior planning group on the following factors: how well it fits into the organization’s strategy, its financial benefit, how it affects quality, its operational impact, the key personnel requirements, and the costs and time required for the project itself. A scale of 1 (low) to 10 (high) is usually used. Each criterion is also weighted; the scores are multiplied by their weight for each criterion and summed over all of the criteria. In exhibit 5.3, project B has a higher total score because of its importance to the organization’s strategy. Such a ranking methodology helps organizations avoid committing resources to projects that may have a powerful internal champion but do not advance the organization’s overall strategy. This matrix can be modified with other categories and weights in accordance with an organization’s current needs.
Initiation and charter
Scope—requirements
Project plan
Scope management and work breakdown
Schedule management
Cost management
Quality control
Communications
Risk management
Procurement
Stakeholders
Closeout process
EXHIBIT 5.2Complete Project Management Process
Healthcare Operat ions Management102
Project CharterOnce a project is identified for implementation, it needs to be chartered. “The project charter is [a] document issued by the project initiator or sponsor that formally authorizes the existence of a project and provides the project manager with the authority to apply organizational resources to project activities” (PMI 2013). A project initiator, or sponsor external to the project, issues the charter and signs it to authorize the start of the project.
Four factors tend to constrain the execution of a project charter: time, cost, scope, and performance. A successful project has a scope that specifies the resulting performance level, how much time it will take to complete, and its budgeted cost. A change in any one of these factors affects the other three, as expressed mathematically in the following equation:
Scope = f (Time, Cost, Performance),
where f is the function of the four factors in a project. Similarly,
Time = f (Cost, Scope, Performance),
and so on.Exhibit 5.4 demonstrates these relationships graphically. Here, the area
of the triangle is a measure of the scope of the project. The length of each side of the triangle indicates the amount of time, amount of money, or level of per-formance needed to complete the project. Because each side of the triangle is connected, changing any of these parameters affects the others. Exhibit 5.5 shows this same project with an increase in required performance level and shortened timelines. With the same scope, this “new” project will incur additional costs.
Measures Possible Points Project A Project B
Strategy alignment 5 3 5
Financial impact 10 4 8
Quality and productivity impact
5 2 3
Customers/patients impact 7 4 4
Staff availability and training 2 2 1
Probability of success (time, cost)
3 3 1
Total 32 18 22
EXHIBIT 5.3Project
Management Matrix
Chapter 5: Project Management 103
Although determining the relationship between all four factors specifi-cally and exactly is difficult, the successful project manager understands this general relationship well and communicates it to project sponsors. A useful analogy is the balloon: If you push hard on one part of it, a different part bulges out. The classic project management dilemma is an increase in scope without additional time or funding (sometimes termed scope creep). Many project failures are directly attributable to ignoring this unyielding formula.
Stakeholder Identification and DialogueThe first step in developing a project charter is to identify the stakeholders—in general, anyone who has an investment in the outcome of the project. Key stakeholders on a project include the project manager; customers; users; project team members; any contracted organizations involved; the project sponsor; those who can influence the project; and the project management office, if one exists in the organization.
The project manager is the individual held accountable for the project’s success and, therefore, represents the core of the stakeholder group. The customer or user of the service or product is an important stakeholder who influences and helps determine the performance of the final product. Even if project team members serve on the project in a limited part-time role, the
StakeholderAnyone who has a vested interest in the outcome of a project, including (but not limited to) employees, customers, users, partner organizations, project sponsors, and the project manager.
Cost = f (Performance, Time, Scope)
Performance
Scope
Time
Cost
EXHIBIT 5.4Relationship of Project Scope to Performance Level, Time, and Cost
Performance
Scope
Time
Cost
EXHIBIT 5.5Project with Increased Performance Requirement and Shortened Schedule
Healthcare Operat ions Management104
success of the project reflects on them; therefore, they become stakeholders as well. A common contracting relationship in healthcare involves large IT installations provided through an outside vendor, which also is included as a project stakeholder. A project should always have a sponsor with enough executive-level influence to clear roadblocks as the project progresses; hence, such individuals need to be included in the stakeholder group. A project may be aided or hindered by many individuals or organizations that are not directly part of it; a global systems analysis should be performed (the system as depicted in exhibit 1.2, chapter 1) to identify which of these should be included as stakeholders.
Once stakeholders have been identified, they need to be interviewed by the project manager to develop the project charter. If an important stakeholder is not available, the project manager should interview someone who represents the stakeholder’s interests. At this point, differentiating between the needs and the wants of stakeholders is important. Adequate detail must be gathered in this process to construct the project charter.
When the project team is organized, it need not include all stakeholders, but the team should be vigilant in attempting to meet all stakeholder needs. The project team should also be cognizant of the culture of the organization, sometimes defined as “how things get done around here.” Projects that chal-lenge an organization’s culture encounter frequent difficulties.
Feasibility AnalysisAn important activity in developing the project charter is determining the project’s feasibility. Feasibility analysis is the review of all the elements of a project that are judged by the project’s sponsor to be acceptable, leading to approval of the project. Because the project should already have undergone an initial prioritization review by the senior management team, the link to the organization’s strategy likely has already been made. To reinforce that linkage, it should be documented in the feasibility analysis. The operational and technical feasibility should also be examined. For example, a new clinical project that requires the construction of new facilities may be impeded in its execution because its timing is contingent on completion of the new buildings.
An initial schedule should be created as part of the feasibility analysis to avoid committing to a requested completion date that is impossible to meet. Finally, both financial benefit and marketplace demand should be considered here. Conducting a financial feasibility analysis is beyond the scope of this text; consult Gapenski and Reiter (2016) to view numerous examples of financial analysis.
All elements of the feasibility analysis should be included in the project charter document, which is described next.
Chapter 5: Project Management 105
Project Charter DocumentThe project charter authorizes the project and serves as an executive summary. A formal charter document should be constructed with the following elements:
• Project mission statement• Project purpose or justification and connection to strategic goals• High-level requirements that satisfy customers, sponsors, and other
stakeholders• Assigned project manager and authority level• Summary milestones• Stakeholder influences• Functional organizations and their expected level of participation• Organizational, environmental, and external assumptions and
constraints• Financial business case, budget, and return on investment• Project sponsor with approval signature
A project charter template is provided on the companion website to this book. The initial descrip-tion of the project scope is found in the Requirements, Milestones, and Financial sections of the template.
A project charter can be illustrated with an example from Vincent Valley Hospital and Health System (VVH). The hospital operates an oncology clinic, Riverview Clinic, in the south suburban area of Bakersville. Recently, the three largest health plans in the area instituted pay-for-performance (P4P) programs to encourage the use of precision medicine in the care of patients with cancer. Precision medicine focuses on identifying which therapeutic approaches will be effective for which patients on the basis of genetic, environmental, and lifestyle factors. For cancer care, pharmacogenomics—the study of how genes affect a person’s response to particular drugs—is a key element of precision medicine. This therapeutic approach combines pharmacology (the science of drugs) and genomics (the study of genes and their functions) to develop effective, safe medi-cations and doses tailored to variations in a person’s genes (Lister Hill 2016).
The health plans will pay the clinic bonuses if Riverview achieves spe-cific levels of performance in the use of precision medicine. The P4P system is being initiated because precision medicine has been shown to provide better results for the patient and to reduce the health plans’ costs over the course of treatment (van den Akker-van Marle et al. 2006). Riverview Clinic staff have decided to embark on a project to increase their use of precision medicine; their project charter is displayed in exhibit 5.6.
On the web at ache.org/books/OpsManagement3
Healthcare Operat ions Management106
Project Mission Statement
This project for Vincent Valley Hospital and Health System (VVH) will increase the level of use of precision medicine—pharmacogenomic drugs—to improve outcomes, lower the long-term costs of care to our patients, and increase reimbursements to the clinic.
Project Purpose and Justification
Health plans in Bakersville have begun to provide additional funding to clin-ics that meet pharmacogenomic use guidelines. Although a number of condi-tions are covered by these new payment systems, VVH leadership feels that pharmacogenomic drug use should be the first project executed because it is likely to be accomplished in a reasonable time frame with the maximum financial benefit to our patients and the clinic. Once this project has been executed, the clinic will move on to more complex clinical conditions.
The project team will be able to incorporate what it has learned about some of the barriers to success and methods to succeed in pay-for-performance (P4P) projects. This project is a part of the larger VVH strategic initiative of maximizing P4P reimbursement.
High-Level Requirements
Once completed, a new prescribing process will• continue to meet patients’ clinical needs and provide high-quality care and• increase pharmacogenomic drug use by 4 percent from baseline within six
months.
Assigned Project Manager and Authority Level
Sally Humphries, RN, will be the project manager. Sally has authority to make changes in budget, time, scope, and performance by up to 10 percent. Any larger change requires approval from the clinic operating board.
Summary Milestones
• The project will commence on January 1.• A system to identify approved pharmacogenomic drugs will be available on
February 15.• The system will go live on March 15.
Stakeholder Influences
The following stakeholders will influence the project:• Clinicians will strive to provide the best care for their patients.• Patients will need to understand the benefits of this new system.• Clinic staff will need training and support tools.• Health plans should be a partner in this project as part of the supply chain.• Pharmaceutical firms should provide clinical information on the efficacy of
certain pharmacogenomic drugs.
EXHIBIT 5.6Project
Charter for VVH Precision
Medicine Project
(continued)
Chapter 5: Project Management 107
Project Scope and Work Breakdown
Once a project has been chartered, the detailed work of planning can begin.
ToolsAt this point, the project manager should consider acquiring two important tools. The first is the lowest of low tech, the humble three-ring binder. All projects need a continuous record of progress reports, team meetings, approved changes, and so on. A complex project requires many binders, and they will prove invaluable to the project manager. The classic organization of the bind-ers is by date, so the first pages should be the project charter. Of course, if the organization has an effective imaging and document management system, this can substitute for the binders.
The second tool is project management software. Although many options are available, the market leader is Microsoft Project, which is used for the
Functional Organizations and Their Participation
• Clinic management staff will lead.• Compcare (electronic health record vendor) will perform software
modifications.• VVH information technology (IT) department will support.• VVH main pharmacy department will support.
Organizational, Environmental, and External Assumptions and Constraints
• Success depends on appropriate substitution of pharmacogenomics for more traditional therapeutic approaches.
• Patients need to understand the benefits of this change.• Health plans need to continue to fund this project over a number of years.• IT modifications need to be approved rapidly by the VVH central IT
department.
Financial Business Case—Return on Investment
The project budget is $161,000 for personnel. Software modifications are included in the master VVH contract and, therefore, have no direct cost impact on this project. If the 4 percent increase in pharmacogenomic drug use is achieved, the two-year revenue increase should be approximately $175,000.
Project Sponsor with Approval Signature
Dr. Jim Anderson, Clinic President
James Anderson, MD
EXHIBIT 5.6Project Charter for VVH Precision Medicine Project (continued from previous page)
Healthcare Operat ions Management108
examples throughout the remainder of the chapter. Microsoft Project is part of the Microsoft Office suite and may already be on many computers in the organization. If not, a demonstration copy can be downloaded from Microsoft.
The companion website for this book provides addi-tional explanation related to the use of Project, along with detailed illustrations of the software’s use for the Riverview Clinic precision medicine drug project.
Project management software is not essential for small projects, but it is helpful and almost required for any project that lasts longer than six months and involves a large team of individuals. Although the Riverview Clinic pharma-cogenomic drug project is relatively small, Project software is used to manage it to provide an illustration of the program’s applicability.
ScopeThe project scope determines what activities fall within the parameters of a project—a good scope document is specific about these boundaries. The start-ing point for developing the detailed scope document is the project charter. To provide the level of detail needed for this document, the project manager revisits many of the same stakeholders who contributed to the charter to acquire specific inputs and requirements. A simple methodology is to interview stakeholders and ask them to list the three most important outcomes of the project, which can be combined into project objectives. The objectives must be specific, achievable, measurable, and comprehensible to stakeholders. In addition, they should be stated in terms of time-limited “deliverables.” For example, the objective “Improve the quality of care to patients with diabetes” is a poor one, whereas “Improve the rate of foot examinations for patients with diabetes by 25 percent in one year” is a much better objective because it states a specific, measurable goal that makes sense to stakeholders and is likely achievable.
The scope document also provides detailed requirements and descrip-tions of expected outcomes, and it often specifies what types of outcomes are not being sought. For example, the Riverview Clinic project scope document might state that the project does not include the use of pharmacogenomic drugs that are still undergoing clinical trials.
The types of deliverables should be specified in the project scope as well, such as implementation of a new process, installation of a new piece of equipment, or presentation of a report. The organization of the project’s per-sonnel is also clarified in the scope document. It names the project manager and team members and defines their relationships in terms of their roles in the overall organization.
An initial evaluation of potential risks to the project should be presented in the scope document. As with other details, the schedule length and milestones
On the web at ache.org/books/OpsManagement3
Chapter 5: Project Management 109
should be more detailed in the scope statement than in the charter. As discussed in the next section, the final schedule is developed on the basis of the work breakdown structure. Finally, the scope document should include methods for monitoring progress and making changes where necessary, including the formal approvals required.
The individual assigned to create the project scope document must avoid expanding the scope of the project beyond its original intent: “While we are at it, we might as well ____.” These add-ons, sometimes called “gold plating,” tend to be some of the most dangerous occurrences in the world of project management because they can result in projects going over budget, not being completed on time, and not meeting performance goals.
Work Breakdown StructureThe second major component of the scope document is the work breakdown structure (WBS), considered the engine of the project because it determines how the project’s goals are to be achieved. The WBS lists the tasks that need to be accomplished, including an estimate of the resources required (e.g., staff time, services, equipment). For complex projects, the WBS is a hierarchy of major tasks, subtasks, and work packages (subdivisions of the work contained in a subtask). Exhibit 5.7 demonstrates this framework graphically.
The size of each task should be planned carefully. A task should not be so small that its monitoring consumes a disproportionate share of the task itself. Similarly, an overly large task cannot be effectively monitored and should be divided into subtasks and then work packages. The task should be described
Work breakdown structure (WBS)A list of the tasks that need to be accomplished, their relationship to each other, and the resources required for a project to meet its goals.
Project
3ksaT2ksaT1ksaT
1.3ksatbuS1.2ksatbuS1.1ksatbuS Subtask 1.3Subtask 1.2
Work package 1.1.1
Work package 1.1.2
Work package 1.1.3
Work package 2.1.1 Work package 2.1.2
Note: This type of diagram can be generated in Microsoft Word and other Microsoft Office products by using the commands Insert → Smart Art → Hierarchy. WBS = work breakdown structure.
EXHIBIT 5.7General Format for WBS
Healthcare Operat ions Management110
in enough detail that the individual responsible, the cost, and the duration can be identified.
A reasonable guideline in terms of duration is that a task should be one to three weeks in length to be effectively monitored. Some tasks are particu-larly critical to the success of the project. These tasks should be identified as milestones. The completion of milestones provides a convenient shorthand method to communicate overall project progress to stakeholders.
The WBS can be developed by the project team itself or with the help of outside experts who have executed similar projects. At this point in the project, the WBS is the best estimate of how the project will be executed. Of course, WBSs are almost always inaccurate in some way, so the formal control and change procedures described in this section are essential to successful project management.
After the WBS has been constructed, the resources required and esti-mated time for each element must be refined. Estimating the time a task will require is an art and is best exercised by a team of individuals. Any previous experiences and data can be helpful in this phase. One group process that has proved useful is the program evaluation and review technique (PERT) for time estimation. Team members individually estimate the time a task will take at its best, worst, and most likely progression. After averaging the team’s responses for each of the times, the final PERT time estimate is computed as
Estimated task timeBest (4 Most likely) Worst
6=
+ × +
After a number of meetings, the Riverview Clinic team has determined that the pharmacogenomic drug project includes three major tasks, each with two subtasks, that needed to be accomplished to meet the goals of the project. The subtasks are as follows:
• Develop a clinical strategy that maintains quality care with the increased use of pharmacogenomics.
• Develop a system to inform clinicians of approved pharmacogenomics.• Update systems to ensure that timely patient medication lists are
available to clinicians.• Develop and deploy a staff education plan.• Develop a system to monitor performance.• Develop and begin to distribute patient education materials.
The WBS for Riverview Clinic’s project is displayed in exhibit 5.8. The actions listed in the bottom tier represent the higher-level tasks for this project. For a project of this scope to proceed effectively, many more subtasks, perhaps
Chapter 5: Project Management 111
50 to 100, are generally required; we are limiting this example to higher-level tasks to illustrate the principles of project management.
It is important to note that the time estimate for each task is the total time needed to accomplish a task, not the calendar time it will take—a three-day task can be accomplished in three days by one person or in one day by three people.
The next step is to determine what resources are needed to accomplish these tasks. Riverview Clinic has decided that this project will be accomplished by four existing employees and the purchase of consulting time from VVH’s IT supplier. The individuals involved are
• Tom Simpson, clinic administrator;• Dr. Betsey Thompson, oncologist;• Sally Humphries, RN, nursing supervisor;• Cindy Tang, billing manager; and• Bill Onku, IT vendor support consultant.
The Project software provides a convenient window in which to docu-ment these individuals’ participation and their cost per hour. The program also provides higher levels of detail, such as the hours an individual can devote to the project and actual calendar days that they are available. When asking clinicians to contribute to a project, the project manager should consider the revenue per hour generated by these individuals, as opposed to their salaries and benefits, because most organizations lose this revenue if the clinician has a busy practice. Exhibit 5.9 shows the Project window for the Riverview Clinic staff who will work on the pharmacogenomic drug project.
Pharmacogenomic drug project
Management and
administrationgniniarTsmetsyS
Develop clinical
strategy (10 days)
Develop monitoring
system (27 days)
Identify approved
pharmacogenomic drugs (22 days)
Supply current patient
medication list (33 days)
Train staff (17 days)
Provide patient education
(9 days)
Note: WBS = work breakdown structure.
EXHIBIT 5.8WBS for River-view Pharmaco-genomic Drug Project
Healthcare Operat ions Management112
Team members should be clear about their accountability for each task. A functional responsibility chart, such as the RASIC framework (which stands for responsible, approval, support, informed, and consult) is helpful; the Riverview project RASIC is displayed in exhibit 5.10. The RASIC diagram is a matrix of team members and tasks from the WBS. For each task, one individual is responsible (R) for ensuring that the task is completed. Other team members may need to approve (A) the completion of the task. Additional team members may work on the task as well, so they are considered support (S). Assigning to a team member the obligation to inform (I) other members helps a team communicate effectively. Finally, some team members need to be consulted (C) as a task is implemented.
RASICA chart delineating all project team members’ roles for each task in a project. The acronym comes from the members’ roles: responsible, approval, support, informed, consult.
EXHIBIT 5.9Resources for the Riverview
Clinic Pharma-cogenomic Drug
Project
WBS Task
Clinic Board of Directors
Lead MDBetsey
Thompson
Adminis-trator
Tom
Simpson
Project Manager
Sally
Humphries
Billing Lead
Cindy Tang
IT LeadBill Onku
Develop clinical
strategyA R C C I I
System to
identify approved
pharmacogenomics
A R S R S
Updated
medication listsR I S I S
Patient education A S S R I I
Staff education A R C S I
Monitoring system A C R C S C
Note: R = responsible; A = approval; S = support; I = inform; C = consult. WBS = work breakdown structure.
EXHIBIT 5.10RASIC for the
Riverview Pharma-
cogenomic Drug Project
Chapter 5: Project Management 113
SchedulingNetwork Diagrams and Gantt ChartsBecause the WBS does not specify a sequence of activities, the next step is to schedule each task to complete the total project. First, the logical order of the tasks must be determined. For example, the Riverview Clinic project team has determined that the system to identify appropriate pharmacogenomic drugs must be developed before the training of staff and education of patients can begin. Other constraints must also be considered in the schedule, including required start or completion dates and resource availability.
Two tools are used to visually display the schedule. The first is a net-work diagram that connects each task in precedence order. This is essentially a process map (chapter 6) in which the process is performed only once; the main difference is that network diagrams do not display paths that return to the beginning (as happens frequently in process maps). A practical way to develop an initial network diagram is to place each task on a sticky note and arrange, and rearrange, the notes on a set of flip charts until they meet the logical and date-imposed constraints. The tasks can then be entered into a project man-agement software system.
Exhibit 5.11 is the network diagram developed by the team for the Riverview Clinic pharmacogenomic drug project. This schedule can be entered into Project to generate a similar diagram. Another common scheduling tool is the Gantt chart, which lists each task on the left side of the page with a bar indicating the start and end times. The Gantt chart for the Riverview Clinic project, generated by Project, is shown in exhibit 5.12. Each bar indicates the duration of the task, and the small arrows connecting the bars indicate the predecessor–successor relationship of the tasks.
Network diagramA scheduling tool that connects tasks in order of precedence.
Gantt chartA scheduling tool that lists project tasks, with bars indicating start and end dates for each task.
Develop clinical strategy
Patient education
Staff education
Updated medication
lists
Monitoringsystem
Start
Implement
System to identify
approved pharma-
cogenomics
EXHIBIT 5.11Network Diagram for Riverview Pharma-cogenomic Drug Project
Healthcare Operat ions Management114
The next step is to assign resources to each task. Exhibit 5.13 shows how the resources are assigned for each day in the project. Care must be taken when assigning resources, as no person works 100 percent of the time. If any single individual is allocated at more than 80 percent in any period, the schedule may need to be adjusted to reduce this allocation. Adjusting the schedule to accommodate this constraint is known as “resource leveling.”
A fi nal review of this initial schedule is undertaken to assess how many tasks are being performed in parallel (simultaneously). A project with few
EXHIBIT 5.12Riverview
Pharma-cogenomic Drug
Project Gantt Chart
EXHIBIT 5.13Riverview
Pharma-cogenomic Drug
Project Tasks with Resources
Assigned
Chapter 5: Project Management 115
parallel tasks takes longer to complete than does one with more total tasks of which many are parallel. Another consideration may be date constraints. Examples include a task that cannot begin until a certain date because of staff availability and a task that must be completed by a certain date to meet an externally imposed deadline (e.g., new Medicare billing policy). The Project software provides tools to set these constraints in the schedule.
Slack and the Critical PathTo optimize a schedule, the project manager must pay attention to slack in the schedule and to the critical path. A task that takes three days but does not need to be completed for five days is said to have two days of slack. The criti-cal path is the longest sequence of tasks with no slack, or the shortest possible completion time of the project.
Slack is determined by the early finish and late finish dates. The early finish date is the earliest date that a task could possibly be completed, as determined by the early finish dates of predecessor tasks. The late finish date is the latest date that a task can be completed without delaying the finish of the project; it is based on the late start and late finish dates of successor tasks. The difference between early finish and late finish dates equals the amount of slack. For critical path tasks (which have no slack), the early finish and late fin-ish dates are identical. Tasks with slack can start later based on the amount of slack they have available. In other words, if (1) a task takes three days, (2) the early finish date is day 18 (based on its predecessors), and (3) the late finish date is day 30 (based on it successors), the slack for this task is 12 days; this task could start as late as day 27 without affecting the completion date of the project. The critical path, which determines the duration of a project, is the connected course through a project of critical tasks.
Calculating slack and the critical path can be complex and time consum-ing. Fortunately, Project performs these functions automatically. However, in some cases (e.g., a basic clinical research project), estimating the duration of tasks is difficult. If a project includes many tasks with high variability in their expected durations, the PERT estimating system should be used. Note that, although PERT employs probabilistic task times to estimate slack and criti-cal paths and is good for time estimation prior to the start of the project, the critical path method is better suited for project management once a project has begun. Having a range of start dates for a task is not particularly useful—what is really important is knowing when a task should have started and whether the project is ahead of or behind schedule.
Although Project provides a PERT scheduling function, the use of PERT is infrequent in healthcare and beyond the scope of this book. Exhibit 5.14 displays a Gantt chart for the Riverview Clinic pharmacogenomics project, with both slack and critical path calculated.
Healthcare Operat ions Management116
Schedule Compression Say the president of Riverview Clinic has been notifi ed by one health plan that if the Riverview pharmacogenomic drug project is implemented by March 1, the clinic will receive a $20,000 bonus. He asks the project manager to consider speeding up, or “crashing,” the project.
The term project crashing has negative associations, as the thought of a computer crashing stirs up dire images. However, a crashed project is simply one that has had its schedule compressed. Schedule compression of a project requires reducing the length of the critical path and can be achieved by using any of the following techniques (PMI 2013, 181):
• Shortening the duration of work on a task on the critical path• Changing a task constraint to allow for more scheduling fl exibility• Breaking a critical task into smaller tasks that can be worked on
simultaneously by different resources• Revising task dependencies to allow more scheduling fl exibility• Performing tasks in parallel as opposed to a linear sequence
(fast-tracking)• Setting lead time between dependent tasks where applicable• Scheduling overtime • Bringing in additional staff• Paying for expedited delivery of needed supplies• Assigning additional resources to work on critical path tasks• Lowering performance goals (not recommended without strong
stakeholder agreement)
The scope, time, duration, and performance relationships need to be considered in a crashed project. A crashed project has a high risk of costing more than the original schedule predicted, so the formal change procedure, discussed in the next section, should be used.
EXHIBIT 5.14Gantt Chart
for Riverview Clinic Pharma-
cogenomic Drug Project
with Slack and Critical Path
CalculatedSlack foreach task
Criticalpath
Chapter 5: Project Management 117
Project Control
Project management would be straightforward if every project’s schedule and costs occurred according to the initial project plan. However, because this is almost never the case, an effective project monitoring and change control system must be in place throughout the life of a project.
Monitoring ProgressThe first important monitoring element is a system to measure schedule comple-tion, cost, and expected performance against the initial plan. Microsoft Project provides a number of tools to assist the project manager in this area. After the plan’s initial scope document, WBS, staffing, and budget have been determined, they are saved as the “baseline plan.” Any changes during the project can be compared to this baseline.
On a disciplined time basis (e.g., once per week), the project manager needs to receive a prog-ress report from each task manager—the individual designated as responsible on the RASIC chart (see exhibit 5.10 and the companion website for examples)—regarding schedule completion and cost.
Change ControlThe project manager should hold a status meeting at least once a month, and preferably more frequently. At this meeting, the project team should review the actual status of the project in terms of task completion, expenses, person-nel utilization, and progress toward expected project outcomes. The majority of time spent in these meetings should be devoted to problem solving, not reporting.
Once deviations are detected, their source and causes must be deter-mined by the team. For major or complex deviations, diagnostic tools such as fishbone diagrams (chapter 6) can be used. Three courses of action are now available: Ignore the deviation if it is small, take corrective action to remedy the problem, or modify the plan by using the formal change procedure developed in the project charter and scope document.
One major cause of deviations is an event that occurs outside the proj-ect. The environment constantly changes during a project’s execution, and modifications of the project’s scope or performance level may be necessary. For example, the application of a new clinical breakthrough may take priority over projects that improve support systems, or a competitor may initiate a new service that requires a response.
A formal change mechanism is a key tool used by high-performing project managers. Resistance to communicating a schedule or cost problem to
On the web at ache.org/books/OpsManagement3
Healthcare Operat ions Management118
project sponsors and stakeholders is part of human nature. However, the con-sequences of this inaction can be significant, if not fatal, to large projects. The change process forces all parties involved in a project to subject themselves to disciplined analysis of options and creates disincentives for scope creep. Changes to the initial plan should be documented in writing and formally approved by the project sponsor as appropriate. They should then be added to the project records (three-ring binders or equivalent).
The Riverview Clinic project charter (and subsequent scope document) states that changes in plan that constitute less than 10 percent of the total affected resource can be made by project manager Sally Humphries. There-fore, she is authorized to adjust the schedule by up to 4.9 days, the cost by up to $6,100, and the performance goal by 0.4 percent. For deviations greater
than these amounts, Sally needs the clinic board to review and sign off on the adjustment. The compan-ion website contains project change documentation and a sign-off template.
CommunicationsA formal communications plan should be developed as part of scope creation. Communications to both internal and external stakeholders are critical to the success of a project. Many types of communications media can be used, from simple oral briefings to e-mails to formal reports. One approach used by many organizations today is to establish a web-based intranet that contains detailed information on the project, combined with a periodical e-mail update sent to stakeholders with a summary progress report and links back to the intranet site for more detailed information. A sophisticated communications plan is fine-tuned to meet stakeholder needs and interests and communicates only those issues of interest to each stakeholder. As part of the communications strategy, feedback from stakeholders should always be solicited, as changes in the project plan may affect one or more stakeholders in ways unknown to the project manager.
The project update communications should contain information gath-ered from quantitative reports. At a minimum, these communications should provide progress against baseline on schedule, cost, scope, and expected per-formance. Any changes to project baseline or the approval process should be noted, as should those issues that need resolution or are being resolved. The expected completion date is always of interest to all stakeholders and should be a prominent part of any project plan communication.
Risk ManagementComprehensive prospective risk management is another element of successful projects. A risk is an event that, if realized, causes the project to experience a
On the web at ache.org/books/OpsManagement3
Risk managementWithin a project, the identification of possible events that, if realized, will affect the execution of the project and a plan to mitigate these events.
Chapter 5: Project Management 119
substantial deviation from the planned schedule, cost, scope, or performance. Like many other aspects of project management, developing a risk manage-ment plan at the beginning of a project—and updating it continuously as the project progresses—takes discipline.
The most direct way to develop a risk management plan is to begin with the WBS. Each task in the WBS should be assessed for risks, both known and unknown. Risks can occur for each task in its performance, duration, or cost. If a project includes 50 tasks, it has 150 potential risks.
A number of techniques can be used to identify risks; the most straight-forward is a brainstorming exercise by the project team. (Some of the tools found in chapter 6, such as mind mapping, root-cause analysis, and force field analysis, can also be used in risk assessment.) Another useful technique is to interview stakeholders to identify risks to the project as viewed from their perspective. The organization’s strategic plan is also a resource, especially if it contains a strengths, weaknesses, opportunities, and threats analysis (frequently referred to as a SWOT analysis). The weaknesses and threats sections may contain clues as to potential risks to a project task.
Once risks have been identified for each task in the WBS structure, the project team should assign a risk probability to each. Those risks with the highest probability, or likelihood, of occurring during the project should be analyzed in depth and a risk management strategy devised. The failure mode and effects analysis method (chapter 6) can also be used for a more rigorous risk analysis.
For tasks that are critical to project execution or that carry high risk, a quantitative analysis can be conducted. Assuming that data can be collected for similar tasks in multiple circumstances, probability distributions can be created and used for simulation and modeling. An example of the applicability of this technique is in remodeling space in an older building. If an organization were to review a number of recent remodeling projects, it might determine that the average cost per square foot of remodeled space is $200 with a normal distribu-tion (think bell curve; distribution is discussed in more detail later in the book). This information may be used as the basis of a Monte Carlo simulation or as part of a decision tree (chapter 6). The results of these simulations provide the project manager with quantitative boundaries on the possible risks associated with the task and project and are useful in constructing mitigation strategies.
Tasks with the following characteristics may be high risk and thus should be considered carefully:
• Long duration• Highly variable estimates of duration• Dependence on external organizations• Requirement of a unique resource (e.g., a physician who is on call)• Likely to be affected by external government or payer policies
Healthcare Operat ions Management120
The management strategy for each identified risk should have three components. First, risk avoidance initiatives should be identified. Avoiding an adverse event is always preferable to dealing with its consequences. An example of a risk avoidance strategy is to provide mentoring to a young team member who has responsibility for key tasks in the project plan.
The second element of the risk management strategy is to develop a mitigation plan. One example of a mitigation response is to bring additional people and financial resources to a task. Another is to call on the project spon-sor to help break an organizational logjam.
Third, a project team may decide to transfer the risk to an insurance entity. This strategy is common in construction projects through the use of bonding for contractors.
All identified risks and their management plans should be outlined in a risk register, a listing of each task, identified risks, and prevention and mitiga-tion plans. This register should be updated throughout the life of the project.
The Riverview Clinic project team has identified three serious risks, which are listed in exhibit 5.15 with their mitigation plans.
Quality Management, Procurement, the Project Management Office, and Project ClosureQuality ManagementThe majority of the focus in this chapter has been on managing the scope, cost, and schedule of a project. The performance, or quality, of an operational project is the fourth key element in successful project management. In general, quality can be defined as meeting specified performance levels with minimal variation and waste.
Mitigation plan A set of tasks intended to reduce or eliminate the effect of risk in a project.
Risk Mitigation Plan
Pharmacogenomic drug use is not as effective as traditional methods.
Assistance will be sought from• VVH hospital pharmacy• Pharmaceutical firms• Health plans
Computer systems do not work.
• IT vendor has specialists on call who will be flown to Riverview Clinic.
• Assistance will be sought from VVH IT department.
Software modifications are more expensive than budgeted.
• Contingency funding has been earmarked in clinic budget.
EXHIBIT 5.15Risk Mitigation
Plan for the Riverview
Clinic Pharma-cogenomic Drug
Project
Chapter 5: Project Management 121
The fundamental tools for accomplishing these goals are described in chapters 6 and 9. More advanced techniques for reducing variation in outcomes can be found in chapter 9 (quality and Six Sigma), and chapter 10 discusses tools for waste reduction (Lean).
Throughout the life of a project, the project team should monitor the expected quality of the final product. Individual tasks that are part of a qual-ity management function within a project should be created in the WBS. For example, one task in the Riverview Clinic pharmacogenomics project is to develop a monitoring system. This system will not only track the use of pharmacogenomic drugs but also ascertain whether their use results in any negative clinical effects.
ProcurementMany projects depend on outside vendors and contractors, so a procurement system integrated with the organization’s project management system is essen-tial. The organization’s purchasing or procurement department can be helpful in this process as well. Procurement staff have developed templates for many of the processes described in the following paragraphs. They also have knowledge of the latest legal constraints an organization may face. However, the most useful attribute of the procurement department may be the frequency with which it executes the purchasing cycle. By performing this task frequently, its staff have developed expertise in the process and are aware of common pitfalls to avoid.
ContractingOnce an organization has decided to contract with a vendor for a portion of a project, three basic types of contracting are available. The fixed-price contract is an agreement that features a lump sum payment for the performance of speci-fied tasks. Fixed-price contracts sometimes contain incentives for early delivery.
Cost-reimbursement contracts call for payment to be made to the vendor on the basis of the vendor’s direct and indirect costs of delivering the service for a specified task. Clearly documenting in advance how the vendor will cal-culate its costs is important.
The most open-ended type of contract is known as a time-and-materials contract. Here, the task itself may be difficult to define, and the contractor is reimbursed for her actual time, materials, and overhead. A time-and-materials contract is commonly used for remodeling an older building, where the con-tractor is not certain of what she will find in the walls. Great caution and monitoring are needed when an organization uses this type of contracting.
Any contract should contain a statement of work (SOW). The SOW contains a detailed scope statement, including WBS, for the work to be per-formed by the contractor. It also includes expected quantity and quality levels,
Statement of work (SOW)A detailed set of tasks, expected outcomes, dates, and costs of a project undertaken by an external contractor.
Healthcare Operat ions Management122
performance data, task durations, work locations, and other details used to monitor the work of the contractor.
Selecting a VendorOnce a preliminary SOW has been developed, the organization solicits propos-als and selects a vendor. A useful first step is to issue a request for information (RFI) to as many possible vendors as the project team can identify. The RFIs generate responses from vendors about their products and experience with similar organizations. On the basis of these responses, the number of feasible vendors can be reduced to a manageable set for consideration.
A more formal request for proposal (RFP) can then be issued to the remaining vendors under consideration for the task. The RFP asks for a detailed proposal, or bid. The following criteria should be applied in the process of reviewing RFPs and awarding the contract:
• Does the vendor clearly understand the organization’s requirements?• What is the vendor’s total cost estimate for completing the task?• Does the vendor have the capability and correct technical approach to
deliver the requested service?• Does the vendor have a management approach to monitor successful
execution of the SOW?• Can the vendor provide maintenance or meet future requirements and
changes?• Does the vendor provide references from clients that are similar to the
contracting organization?• Does the vendor assert intellectual or proprietary property rights in the
products it supplies?
Project Management OfficeMany types of organizations outside the healthcare industry (e.g., architecture, consulting) are primarily project oriented. Such organizations often have a centralized project management office (PMO) to oversee the work of their staff. Because healthcare delivery organizations are primarily operational, the majority do not use this structure.
However, departments in large hospitals and clinics, such as IT and quality, have begun to use a centralized project office approach. In addition, some organizations have designated and trained project leaders in Six Sigma or Lean techniques. These project leaders are assigned from a central PMO.
PMOs provide a single structure through which to monitor progress on all projects in an organization and reallocate resources as needed when
Chapter 5: Project Management 123
projects encounter problems. They also serve as a resource for the training and development of project managers. PMOs support the project manager in many ways, including but not limited to the following (PMI 2013):
• Managing shared resources across all projects administered by the PMO• Identifying and developing project management methodology, best
practices, and standards• Coaching, mentoring, training, and oversight• Developing and managing project policies, procedures, templates, and
other shared documentation• Monitoring compliance with project management standards, policies,
procedures, and templates via project audits• Coordinating communications across all projects
Another useful function of a PMO is that it maintains an information system that can provide reports to project stakeholders and senior management. The contents of this information system may include the following:
• Progress reports on individual projects (schedule, cost, performance)• Risk management (tasks with high risks and their current status)• Performance failures and remediation steps• A log of lessons learned
Project ClosureA successful project should have an organized closure process, which includes a formal stakeholder presentation and approval process. In addition, the proj-ect sponsor should sign off at project completion to signify that performance levels have been achieved and all deliverables have been received. During the closeout process, special attention should be paid to project staff, who will be interested in knowing their next assignment. A disciplined handoff of staff from one project to the next allows successful completion of the closure process.
All documents related to the project should be indexed and stored. This process can be helpful if outside vendors have participated in the project and a contract dispute arises in the future. Historical documents can also provide a good starting point for the next version of a project.
The project team should conduct a final session to identify lessons learned—good and bad—in the execution of the project. These lessons should be included in the project documentation and shared with other project man-agers in the organization.
Healthcare Operat ions Management124
Agile Project Management
In some situations, knowledge of the tasks necessary for project success is not available as the project is chartered and scheduled. In these cases, agile project management often works better than other methods. Agile project management is adaptive, in contrast to the predictive style of formal project management.
Characteristics of agile project management include the following:
• Customer satisfaction is achieved by rapid, continuous delivery of services or new processes.
• Newly prototyped services or processes are delivered frequently (weekly rather than monthly).
• The effectiveness and ease of use of these prototypes are the principal measures of progress.
• The project team can easily incorporate late changes.• The project team and customer interact informally and frequently.• The project team and customer are colocated.• The project team is cross-functional across the organization.• Continuous attention is given to technical excellence and good design
in the new services or processes.• The project team regularly adapts to changing circumstances.
Exhibit 5.16 illustrates agile project management.Agile project management is best used for “mysteries,” to which there
are no known answers (e.g., finding the best treatment for an emerging disease),
Build prototype service or process.
Collaborate with customers to define and refine
requirements.
Does it meet expected
performance?Implement—go
to scale.
No
Yes
Charter project.Determine expected time
frame, cost, and performance.
EXHIBIT 5.16Agile Project
Management
Chapter 5: Project Management 125
as opposed to “puzzles,” to which the solution is known but complex (e.g., building a new clinic).
Innovation Centers
Mergers and consolidations of providers and health plans have increased since the enactment of the Affordable Care Act in 2010. A key strategic objective of the law was to encourage the development of systems of care, as they have demonstrated superior cost-effectiveness and quality. However, as these systems have proliferated, their leaders have realized that growth in size also frequently means growth in bureaucracy and the consequent loss of rapid innovation.
To remedy this dilemma, many systems have created innovation centers. The Commonwealth Fund conducted a survey in 2014–2015 of innovation centers in 31 healthcare systems. The authors of this study reviewed charac-teristics of these centers and determined that, in general, the innovation cen-ters were “‘places that are working to discover, develop, test, and/or spread new models of care delivery’—in hospitals, clinics, and patients’ homes. The innovations they test may be internally developed or adopted from elsewhere” (Commonwealth Fund 2015). These centers focused on a variety of topics, as displayed in exhibit 5.17.
Care coordinationDisease-specific outcomes
AccessPatient engagement
Workflow efficienciesPopulation health
Clinical decision supportIntraprofessional communication
UtilizationHome-based care
WellnessPatient safety
DevicesCommunity-based services
Price transparencyOther responses
0% 50%
30% 35%
55% 61% 61%
65% 65%
68% 71%
74% 77% 77%
84% 87% 87%
90%
100%
Percentage of Innovation Center Survey Respondents
EXHIBIT 5.17Focus of Innovation Centers
Source: Commonwealth Fund (2015). Used with permission.
Note: Percentages based on 31 innovation center respondents. “Other responses” include spending reductions, the uninsured, helping seniors age in place, teaching/education, data mining, and data analysis.
Healthcare Operat ions Management126
Disruptive InnovationMcLaughlin and Militello (2015) conducted a review of disruptive innovation literature and noted the following:
Clayton Christensen [(Christensen, Grossman, and Hwang 2008)] introduced the con-
cept of creative disruption to the healthcare industry. His basic theory of disruption
is a process by which complicated, expensive products and services are transformed
into simple, affordable ones. Disruptive solutions emerge almost always through
new companies or totally independent business units that create new, value-added
processes. However, significant disruption has arrived in the healthcare sector in a
way not imagined by Christensen. It came from the Affordable Care Act (ACA). . . .
Under Christensen’s idea of disruptive solutions, it seems that managers should
be at the forefront of knitting together new concepts of cost containment, quality,
and exceptional service to transform their organization into providers of simple and
affordable solutions within the context of ACA principles.
The review identified three key steps for creating effective disruptive innovations in healthcare (McLaughlin and Militello 2015):
1. Test the business model against the needs of the customer.2. Pilot test one or two ideas to create or counter disruption. 3. Look across disciplines and organizational boundaries for ideas and
encouragement.
All of the project management tools and approaches contained in this chap-ter are being used by organizations to undergird the work of innovation centers.
The Project Manager and Project Team
The project manager’s role is pivotal to the success of any project, as he must select, develop, and nurture high-functioning team members, among other critical activities. The project manager’s skills also include running effective meetings and facilitating optimal dialogue during these meetings.
Team SkillsA project manager can take on multiple roles in a project. In many smaller healthcare projects, the project manager is the person who actually accomplishes several of the project tasks. In larger or more complex projects, the project manager’s job is solely to lead and manage the individuals performing the tasks. Slack (2005) provides a useful matrix to determine what role a project manager should assume in projects of varying sizes (exhibit 5.18).
Chapter 5: Project Management 127
Team Structure and AuthorityTeam members may be selected and the project structure determined by the project manager, but in many cases they are outlined by the project sponsor and other members of senior management. Formally documenting the team makeup and how team members are assigned in the project charter and scope is impor-tant in clarifying team roles for both team members and project stakeholders.
A number of key issues must be addressed as the project team is formed. The most important is the project manager’s level of authority to make deci-sions. Can the project manager commit resources, or must he ask senior man-agers or department heads each time a new resource is needed? Is the budget controlled by the project manager, or does a central financial authority control it? Is administrative support available to the team, or do the project team members need to perform these tasks themselves?
Finally, care should be taken to avoid overscheduling team members. All members must have the availability to work on the project as expected.
Team MeetingsA weekly or biweekly project team meeting is highly recommended to keep a project on schedule. At this meeting, the project’s progress can be monitored and discussed and actions initiated to resolve deviations and problems.
All good team meetings include a comprehensive agenda and a complete set of minutes. Minutes should be action oriented (e.g., “The schedule slippage for task 17 will be resolved by assigning additional resources from the temporary pool”). In addition, the individual accountable for following through on the issue should be identified. If the meeting’s deliberations and actions are confidential, everyone on the team should be aware of the policy and adhere to it uniformly.
The decision-making process should be clear and understood by all team members. In some situations, all major decisions are made by the project manager. In others, team members may have veto power if they represent a major department that is expected to commit resources. Some major decisions may require review and approval by individuals external to the project team. The use of data and analytical techniques is strongly encouraged as part of the decision-making process.
Variable Small Project Medium Project Large Project
Effort range 40–400 hours 400–2,400 hours 2,400+ hours
Duration 1 week–3 months 3–6 months 6 months–2 years
Project leader role
“Doer” with some help
Manage and “do some”
Manage
Source: Slack (2005). Used with permission.
EXHIBIT 5.18Project Manager’s Role Based on Effort and Duration of a Project
Healthcare Operat ions Management128
Team members need to take responsibility for the success of the team. They can demonstrate this acceptance by following through on commitments, contributing to discussions, actively listening, and giving and being receptive to feedback. Everyone on a team should feel that she has a voice, and the project manager needs to lead the meeting in such a way as to balance the “air time” among team members. This approach requires occasionally interrupting—politely and artfully—the wordy team member and summarizing her point; it also means calling on the silent team member to solicit input.
At the end of a meeting, one useful activity is to evaluate the meeting itself. The project manager and team can spend a few minutes reviewing ques-tions such as the following:
• Did we accomplish our purpose?• Did we take steps to maintain our gains?• Did we document actions, results, and ideas?• Did we work together successfully?• Did we share our results with others?• Did we recognize everyone’s contribution and celebrate our
achievements?
Leadership SkillsAlthough this book is not primarily concerned with leadership, clearly the project manager must be able to lead a project forward. Effective project leadership requires the following skills:
• The ability to think critically using complex information• The strategic capability to take a long-term view of the organization• The ability to gain and maintain a systems view of the organization and
its environment (discussed in chapter 1)• The ability to create and lead change• The capacity to understand oneself to permit positive interactions,
conflict resolution, and effective communication• The ability to mentor and develop employees into high-performing
team members• The ability to develop a performance-based culture
Additional resources on leadership are available from Health Administration Press at www.ache.org/pubs/topic.cfm#Leadership.
Chapter 5: Project Management 129
Conclusion
This chapter provides a basic introduction to the science and discipline of project management. The field is finding a home in healthcare IT departments and has a history in construction projects. Successful healthcare organizations of the future will use this rigorous methodology to make significant changes and improvements throughout their operations.
Discussion Questions
1. Who should be included as members of the project team, key stakeholders, and project sponsors for a clinical project in a physician’s office? In a hospital? Support your choices.
2. Identify five common risks in healthcare clinical projects, and develop contingency responses for each.
Exercises
1. Download the project charter and project schedule from the companion website, and perform the following activities: a. Complete the missing portions of the
charter. b. Develop a risk assessment and mitigation plan. c. Add tasks to the schedule for those areas that require more
specificity. d. Apply resources to each task, determine the critical path, and devise
a method to crash the project to reduce its total duration by 20 percent.
2. Review the Institute for Healthcare Improvement website (see especially www.ihi.org/knowledge/Pages/ImprovementStories/default.aspx), and select one of the quality improvement projects described. Although you will not know all the details of the organization that executed this project, create a charter document for your chosen project.
3. For the project identified in exercise 2, create a feasible WBS and project schedule. Enter the schedule into Microsoft Project.
On the web at ache.org/books/OpsManagement3
Healthcare Operat ions Management130
References
Azzopardi, S. 2016. “The Evolution of Project Management.” Accessed September 13. www.projectsmart.co.uk/evolution-of-project-management.php.
Carden, L., and T. Egan. 2008. “Does Our Literature Support Sectors Newer to Project Management? The Search for Quality Publications Relevant to Nontraditional Industries.” Project Management Journal 39 (3): 6–27.
Christensen, C. M., J. H. Grossman, and J. Hwang. 2008. The Innovator’s Prescription: A Disruptive Solution for Health Care. New York: McGraw-Hill.
Commonwealth Fund. 2015. “Findings from a Survey of Health Care Delivery Innova-tion Centers.” Published April 28. www.commonwealthfund.org/publications/chartbooks/2015/apr/survey-of-health-care-delivery-innovation-centers.
Gapenski, L. C., and K. L. Reiter. 2016. Health Care Finance: An Introduction to Accounting and Financial Management. Chicago: Health Administration Press.
Lister Hill National Center for Biomedical Communications, US National Library of Medicine, National Institutes of Health. 2016. “Help Me Understand Genetics: Precision Medicine.” Published August 16. https://ghr.nlm.nih.gov/handbook/precisionmedicine.pdf.
McLaughlin, D. B., and J. Militello. 2015. “Thinking Beyond the Affordable Care Act.” Journal of Healthcare Management 60 (3): 160–63.
Project Management Institute (PMI). 2015. “Connections: Project Management Institute 2014 Annual Report.” Accessed August 1, 2016. www.pmi.org/~/media/PDF/Publications/pmi-annual-report-2014.ashx.
———. 2013. 2013 Guide to the Project Management Body of Knowledge: PMBOK® Guide—Fifth Edition. Newton Square, PA: PMI.
Quora. 2016. “How Many People Have PMP Certification in the U.S.?” Accessed August 1. www.quora.com/How-many-people-have-PMP-certification-in-the-U-S.
Slack, M. P. 2005. Personal communication, August 15.Thomson, L. 2013. “HealthCare.gov Diagnosis: The Government Broke Every Rule of
Project Management.” Forbes. Published December 3. www.forbes.com/sites/lorenthompson/2013/12/03/healthcare-gov-diagnosis-the-government-broke-every-rule-of-project-management/#2e03a6aa3d44.
van den Akker-van Marle, M. E., D. Gurwitz, S. B. Detmar, C. M. Enzing, M. M. Hopkins, E. Gutierrez de Mesa, and D. Ibarreta. 2006. “Cost-Effectiveness of Pharmacogenom-ics in Clinical Practice: A Case Study of Thiopurine Methyltransferase Genotyping in Acute Lymphoblastic Leukemia in Europe.” Pharmacogenomics 7 (5): 783–92.
Further Reading
Bonabeau, E., N. Bodick, and R. W. Armstrong. 2008. “A More Rational Approach to New-Product Development.” Harvard Business Review 86 (3): 96–102.
Chapter 5: Project Management 131
Chesbrough, H. W., and A. R. Garman. 2009. “How Open Innovation Can Help You Cope in Lean Times.” Harvard Business Review 87 (12): 68–76.
Curlee, W., and R. L. Gordon. 2010. Complexity Theory and Project Management. Hobo-ken, NJ: Wiley.
Kendrick, T. 2010. The Project Management Tool Kit: 100 Tips and Techniques for Getting the Job Done Right, 2nd edition. New York: AMACOM American Management Association.
Meredith, J. R., and S. J. Mantel. 2009. Project Management: A Managerial Approach, 7th edition. Hoboken, NJ: Wiley.
Meskendahl, S. 2010. “The Influence of Business Strategy on Project Portfolio Manage-ment and Its Success—A Conceptual Framework.” International Journal of Project Management 28 (8): 807–17.
Taylor, H. 2006. “Risk Management and Problem Resolution Strategies for IT Projects: Prescription and Practice.” Project Management Journal 37 (5): 49–63.
Thamhain, H. J., and D. L. Wilemon. 1975. “Conflict Management in Project Life Cycles.” Sloan Management Review 16 (3): 31.
Wheelwright, S. C., and K. B. Clark. 1992. “Creating Project Plans to Focus Product Development.” Harvard Business Review 70 (2): 70–82.
Wills, K. R. 2010. Essential Project Management Skills. Boca Raton, FL: Taylor & Francis.Young, T. L. 2010. Successful Project Management, 3rd edition. London: Kogan Page.
PART
IIIPERFORMANCE IMPROVEMENT
TOOLS, TECHNIQUES, AND PROGRAMS
CHAPTER
135
TOOLS FOR PROBLEM SOLVING AND DECISION MAKING
Operations Management in Action
At Allegheny General Hospital in Pitts-burgh, the organization’s two intensive care units had been averaging about 5.5 infections per 1,000 patient days, most of them bloodstream infections from catheters. That infection rate was a bit higher than the Pittsburgh average but a bit lower than the national average, says Dr. Richard Shannon, chair of medicine at Allegheny General.
Over the prior 12 months, 37 patients, already some of the sickest people in the hospital, had a total of 49 infections. Of those patients, 51 percent died. Shannon and the staff in the two units—doctors, residents, and nurses—applied the Toyota Production System root-cause analysis process, investigat-ing each new infection immediately.
Their main conclusion was that femoral intravenous lines, inserted into an artery near the groin, had a particu-larly high rate of infection. The team made an all-out effort to replace these lines with less risky ones in the arm or near the collarbone. Shannon, who oversaw the two units, gave the direc-tive to keep femoral lines to an absolute minimum. The result was a 90 percent decrease in the number of infections after just 90 days of using the new procedures.
Source: Adapted from Wysocki (2004).
6OVE RVI EW
This chapter introduces the basic tools associated with problem solv-
ing and decision making. Much of the work of healthcare profession-
als is just that—making decisions and solving problems—and in an
ever-changing landscape, that work must be accomplished well and
quickly. A structured approach can enable problem solving and deci-
sion making that is efficient and effective.
Major topics in this chapter include the following:
• The decision-making process, with a focus on framing the
problem or issue
• Mapping techniques, including mind mapping, process mapping,
activity mapping, and service blueprinting
• Problem identification tools, including root-cause analysis (RCA),
failure mode and effects analysis (FMEA), and the theory of
constraints (TOC)
• Analytical tools, such as optimization using linear programming
and decision analysis
• Force field analysis to address implementation issues
This chapter helps readers gain a basic understanding of vari-
ous problem-solving tools and techniques, enabling them to
• frame questions or problems,
• analyze a problem and various solutions to it, and
• implement one or more of those solutions.
The tools and techniques outlined in this chapter should pro-
vide a basis for tackling difficult, complicated problems.
Healthcare Operat ions Management136
Decision-Making Framework
A structured, rational approach to problem solving and decision making includes the following steps:
1. Identify and frame the issue or problem.2. Generate or determine possible courses of action, and evaluate those
alternatives.3. Choose and implement the best solution or alternative.4. Review and reflect on the previous steps and outcomes.
Decision Traps: The Ten Barriers to Brilliant Decision-Making and How to Overcome Them (Russo and Schoemaker 1989) outlines these steps (exhibit 6.1) and the barriers encountered in decision making (exhibit 6.2).
Framing
Typical amount of time: 5%
Recommended amount of time: 20%
Structuring the question. This means defining what must be decided and determining in a preliminary way what criteria would cause you to prefer one option over another. In framing, good decision makers think about the viewpoint from which they and others will look at the issue and decide which aspects they consider important and which they do not. Thus, they inevitably simplify the world.
Gathering intelligence
Typical amountof time: 45%
Recommended amount of time: 35%
Seeking both the knowable facts and the reasonable estimates of “unknowables” that you will need to make the decision. Good decision makers manage intelligence gathering with deliberate effort to avoid such failings as overconfidence in what they currently believe and the tendency to seek information that confirms their biases.
Coming to conclusions
Typical amountof time: 40%
Recommended amount of time: 25%
Sound framing and good intelligence don’t guarantee a wise decision. People cannot consistently make good decisions using seat-of-the-pants judgment alone, even with excellent data in front of them. A systematic approach forces you to examine many aspects and often leads to better decisions than hours of unorganized thinking would.
Learning from feedback
Typical amount of time: 10%
Recommended amount of time: 20%
Everyone needs to establish a system for learning from the results of past decisions. This usually means keeping track of what you expected would happen, systematically guarding against self-serving explanations, then making sure you review the lessons your feedback has produced the next time a similar decision comes along.
Source: Russo and Schoemaker (1989).
EXHIBIT 6.1Decision
Elements and Activities
Chapter 6: Tools for Problem Solving and Decis ion Making 137
The plan-do-check-act process for continuous improvement (discussed in depth in chapter 10), the define-measure-analyze-improve-control process of Six Sigma (chapter 9), and process improvement (chapter 11) all follow the same basic steps as presented in this chapter for the decision-making process. The tools and techniques found in this book can be used not only in the process of decision making but also to help in gathering the right information to make optimal decisions and learn from those decisions. Often, the learning step in the decision-making process is neglected, but it should not be. It is important
Framing the QuestionPlunging in—Beginning to gather information and reach conclusions without first tak-ing a few minutes to think about the crux of the issue you’re facing.
Frame blindness—Setting out to solve the wrong problem because you have created a mental framework for your decision with little thought, which causes you to overlook the best options or lose sight of important objectives.
Lack of frame control—Failing to consciously define the problem in more ways than one or being unduly influenced by the frames of others.
Gathering IntelligenceOverconfidence in your judgment—Failing to correct key factual information because you are too sure of your assumptions and opinions.
Shortsighted shortcuts—Relying inappropriately on “rules of thumb,” such as implicitly trusting the most readily available information or anchoring too much on convenient facts.
Coming to ConclusionsShooting from the hip—Believing you can keep straight in your head all the informa-tion you’ve discovered, and therefore you “wing it” rather than follow a systematic procedure.
Group failure—Assuming that with many smart people involved, good choices will fol-low automatically, and therefore you fail to manage the group decision process.
Learning/Failing to Learn from FeedbackFooling yourself about feedback—Failing to interpret the evidence from past outcomes for what it really says, either because you’re protecting your ego or because you are tricked by hindsight.
Not keeping track—Assuming that experience will make its lessons available automati-cally, and therefore you fail to keep systematic records to track results of your decisions and fail to analyze these results in ways that will reveal their true lessons.
Failure to audit your decision process—You fail to create an organized approach to understanding your own decision making, so you remain constantly exposed to all the aforementioned mistakes.
EXHIBIT 6.2The Ten Barriers to Brilliant Decision Making and the Key Elements into Which They Fall
Source: Russo and Schoemaker (1989).
Healthcare Operat ions Management138
to evaluate and analyze both the decision made and the process(es) used to reach the decision to ensure learning and enable continuous improvement.
FramingThe frame of a problem or decision encompasses the assumptions, attitudes, and preconceived limits that an individual or a team brings to the analyses. These assumptions can stifle the ability to solve the problem by reducing or eliminating creativity and causing the decision maker(s) to overlook possibili-ties. Alternatively, these assumptions can aid in problem solving by eliminating wildly improbable paths. That said, they usually hinder a team’s ability to find the best solution or even a possible solution.
Millions of dollars and working hours are wasted in finding solutions to the wrong problems. An ill-defined problem or mistaken premise can eliminate promising solutions before they are even considered. People tend to identify convenient problems and find solutions that are familiar to them rather than looking more deeply for problems that are meaningful to solve.
People also have a tendency to want to do something; quick and decisive action is seen as necessary in today’s rapidly changing environment. Leaping to the solutions before taking the time to properly frame the problem usually results in suboptimal solutions.
However, framing the problem can be difficult, as it requires an under-standing of the issue at hand. If the problem is well understood, the solution is more likely to be obvious; therefore, when framing a problem, it is important to approach it in an expansive way by soliciting many different viewpoints and considering many possible scenarios, causes, and solutions. The tools outlined in this chapter are designed to help with this process.
Mapping TechniquesMind MappingTony Buzan is credited with developing the mind-mapping technique (Buzan 1991; Buzan and Buzan 1994). Mind mapping develops thoughts and ideas in a nonlinear fashion and typically uses pictures or phrases to organize and further develop those thoughts. In this structured brainstorming technique, ideas are organized on a “map” and the connections between them are made explicit. Mind mapping can be an effective technique for problem solving because thinking linearly is not necessary. Making connections that are not obvious or linear can lead to innovative solutions.
Mind mapping starts with the issue to be addressed placed in the cen-ter of the map. Ideas on causes, solutions, and so on radiate from the central theme. Questions in the form of who, what, where, why, when, and how are often helpful for problem solving. Exhibit 6.3 illustrates a mind map related to high accounts receivables.
Mind mappingA nonlinear technique used to develop thoughts and ideas by placing pictures or phrases on a map to show logical connections.
Chapter 6: Tools for Problem Solving and Decis ion Making 139
Process MappingA process map, or flowchart, is a graphic depiction of a process showing inputs, outputs, and steps in the process. Depending on the purpose of the map, it can be high level or detailed. Exhibit 6.4 shows a high-level process map for
Process mapA graphic depiction of a process showing the sequence of events, including tasks, decisions, and other activities from inputs to outputs. A process map is a type of flowchart.
Dataentryerror
Doctorcoding
Incorrectcoding
Incorrectinformation
Documentationproblems
Insufficientinformation
Electronicmedicalrecords
Slowbilling
Slowpayment
or nopayment
Claimdenied
Procedurenot medically
necessary
Complicatedsystem
Newcomputersystems
FundingMissingrevenue
Privateinsurance
Identifyand fix
systematicproblems
Medicare/Medicaid
Noinsurance
Type ofinsurance
Highaccounts
receivable
Procedurenot
covered
EXHIBIT 6.3Mind Map: High Accounts Receivables
Note: Diagram created in Inspiration by Inspiration Software, Inc.
Healthcare Operat ions Management140
Vincent Valley Hospital and Health System’s (VVH) Riverview Clinic, and exhibit 6.5 shows a more detailed map of the check-in process at the clinic.
Process maps offer a clear picture of what activities are carried out as part of the process, where they occur, and how they are performed. Typically, process maps are used to understand and optimize a process. The process is commonly charted from the viewpoint of the material, information, or cus-tomer being processed (often the patient in healthcare) or the staff member carrying out the work. Process mapping is one of the seven basic quality tools (see chart below) and an integral part of most improvement initiatives (e.g., Six Sigma, Lean, balanced scorecard, RCA, FMEA).
The steps for creating a process map or flowchart are as follows:
1. Assemble and train the team. The team should consist of people from all areas and levels in the process of interest to ensure that the real process is captured.
2. Determine the boundaries of the process (where it starts and ends) and the level of detail desired. The level of detail desired, or needed, depends on the question or problem the team is addressing.
3. Brainstorm the major process tasks and subtasks. List them, and then arrange them in order. (Sticky notes are often helpful here.)
4. Create a formal chart. Once an initial flowchart has been generated, the chart can be formally drawn using the standard symbols of
Physician examand consultation
Visitcomplete
Wait
Patientarrives
Patientcheck-in
Wait
Move toexamining
room
Nurse doespreliminary
exam
Wait
EXHIBIT 6.4Riverview Clinic
High-Level Process Map
Seven Fundamental Quality Tools
• Check sheet• Pareto diagram• Histogram• Scatterplot
• Process map• Cause-and-effect diagram• Run chart or control chart
Chapter 6: Tools for Problem Solving and Decis ion Making 141
process mapping (exhibit 6.6). (This formal graphic can be completed most efficiently using software such as Microsoft Visio.) When first developing a flowchart, the important point is to obtain an accurate picture of the process rather than worrying about using the correct symbols.
5. Make corrections. The formal flowchart should be checked for accuracy by all relevant personnel. Often, inaccuracies are found in the flowchart and must be corrected in this step.
6. Determine any need for additional information. Depending on the purpose of the flowchart, data may need to be collected or information added at this stage. Often, data specifically related to process performance are collected and added to the flowchart.
Measures of Process PerformanceMeasures of process performance include throughput time, cycle time, and percentage of value-added time (chapter 10). Another important measure of process, subprocess, task, or resource performance is capacity utilization. Capacity is the maximum possible amount of output (goods or services) that a process or resource can produce or transform. Capacity measures can be based on outputs or on the availability of inputs. For example, if a hospital’s food service department or vendor can provide, at most, 1,000 meals in one day, the food service has a capacity of 1,000 meals/day. If all magnetic resonance images (MRIs) take one hour to perform, the MRI machine has a capacity of 24 MRIs/day. The choice of appropriate capacity measure varies with the situation.
Ideally, demand and capacity are perfectly matched. If demand is greater than capacity, some customers will not be served. If capacity is greater than demand, resources will be underutilized. In reality, perfectly matching demand
Capacity utilizationThe percentage of time that a resource (worker, equipment, space, etc.) or process is actually busy producing or transforming output.
HIPAAforms
HIPAAon file ?
No
YesSame
Wait
Move toexamining
room
Patientarrives
Line?
ChangedNew
Medicalinformation
Insuranceinformation
No
Yes
Existing Infor-mation
Patienttype
Wait
EXHIBIT 6.5Riverview Clinic Detailed Process Map: Patient Check-In
Healthcare Operat ions Management142
and capacity can be difficult because of fluctuations in demand. In a manufactur-ing environment, inventory can be used to compensate for demand fluctuations. In a service environment, this type of trade-off is not possible; therefore, excess capacity or a flexible workforce is often required to meet demand fluctuations. Advanced-access scheduling (chapters 10 and 12) is one way for healthcare operations to more closely match capacity to demand.
Capacity utilization is defined as the percentage of time that a resource (worker, equipment, space, etc.) or process is actually busy producing or trans-forming output. If the hospital’s food service provides 800 meals/day, the capacity utilization is 80 percent. If the MRI machine operates 18 hours/day, the capacity utilization is 75 percent [(18 ÷ 24) × 100]. Generally, higher-capacity utilization is better, but caution must be used in this evaluation. If the hospital’s food service has a goal of 95 percent capacity utilization, it can meet that goal by producing 950 meals/day, even if only 800 meals/day are actually consumed and 150 meals are discarded. Obviously, this solution would not result in the effective use of resources, but food service would have met its goal.
Typically, the more costly the resource, the greater is the importance of maximizing capacity utilization. For example, in a hospital emergency
An oval is used toshow inputs/outputs
to the process or start/end of the process.
A block arrow is used to show a transport.
Feedbackloop
A D-shape is used to show a delay.
An arrow shows the direction of flow of
the process.
Atriangle
shows inventory. For services,
it can represent customer waiting.
End
Adiamond
is used to show those points in the
process where a choice can be made or alternative
paths can be followed.
Arectangleis used to
show a task oractivity.
EXHIBIT 6.6Standard
Flowchart Symbols
Chapter 6: Tools for Problem Solving and Decis ion Making 143
department, the most costly resource is often the physician. In this case, with other resources (e.g., nurses, housekeeping staff, clerical staff) being less expensive, maximizing the utilization of the physicians is more important than maximizing the utilization of the other resources. In fact, underutilizing less expensive resources in an effort to maximize the utilization of more expensive resources is more economical. Simulation (chapter 11) can be used to help determine the most effective use of various types of resources.
Cross-Functional Process MapsThe cross-functional process map, or “swim lane” map, is a specialized pro-cess map that follows the flow of a process through the various departments of the organization. The swim lanes indicated by the dashed lines between departments show the work being completed by a particular department or individual in the process. The swim lane chart is useful for viewing the number of times an item is handed off between departments and how many times the process is creating duplication and rework.
Exhibit 6.7 is an example of a presurgery holding room (the area where patients are staged, vital signs are taken, consent forms are completed, surgery type is confirmed, etc., just prior to entering surgery) for a Veterans Admin-istration hospital.
Service BlueprintingService blueprinting (Shostack 1984) is a special form of process mapping (as is value stream mapping, covered in chapter 10). Service blueprinting begins by mapping the process from the point of view of the customer. The typical purpose of a service blueprint is to identify points where the service might fail to satisfy the customer and then redesign or add controls to the system to reduce or eliminate the possibility of failure. The service blueprint separates onstage actions (those visible to the customer) from backstage actions and support processes (those not visible to the customer). A service blueprint specifies the line of interaction, where the customer and service provider come together, and the line of visibility, or what the customer sees or experiences—the tangible evidence that influences perceptions of the quality of service (exhibit 6.8).
Problem Identification ToolsRoot-Cause AnalysisRoot-cause analysis (RCA) is a generic term used to describe structured, step-by-step techniques for problem solving. It aims to determine and correct the ultimate cause(s) of a problem, not just the visible symptoms, to ensure that the problem does not recur. Specifically, RCA consists of determining what happened, why it happened, and what can be done to prevent it from recurring.
Cross-functional process mapA map that follows the flow of a process through the various departments of the organization using dashed lines to show the work being completed by a particular department or individual in the process. Also called swim lane process map.
Service blueprintingA style of process mapping that separates actions into onstage (visible to the customer) and backstage (not visible to the customer) activities.
Root-cause analysis (RCA)A generic term describing structured, step-by-step techniques for problem solving.
Healthcare Operat ions Management144
Sna
psho
t of a
Por
tion
of a
Det
aile
d P
roce
ss M
ap fo
r a
Hol
ding
Roo
m
OR
NA
retu
rns
to c
ontr
ol d
esk
to re
trie
ve a
noth
er s
lip a
nd p
atie
nt
Yes
No
Ret
riev
e IV
tubi
ng
from
SP
D c
art
Is IV
tubi
ngav
aila
ble
onth
e ca
rt?
Clea
n to
p of
IV c
art
Was
the
cart
re
stoc
ked
by
phar
mac
y w
ith
salin
e an
d he
pari
n sy
ring
es?
Mul
tipl
e IV
bag
sar
e op
ened
and
spik
ed w
ith
IVtu
bing
SPD
is c
alle
dto
rest
ock
Oth
er it
ems
in IV
car
t are
re
stoc
ked
by a
nest
hesi
a te
chni
cian
(see
item
und
er
anes
thes
ia h
eadi
ng in
ho
ldin
g ro
om)
Syri
nges
retr
ieve
d fr
om
Pixi
s an
d ca
rt re
stoc
ked
OR NurseHolding Room NurseOR Nursing Assistant
Yes
No
Yes
No
Is p
atie
nt
read
y fo
r tr
ansp
ort?
Assi
sts
pati
ent
onto
the
litte
r or
whe
elch
air
Is th
e pa
tien
t sc
hedu
led
for
open
-hea
rt
surg
ery?
OR
NA
veri
fies
the
pati
ent’
s na
me
and
SSN
wit
h sl
ip
Oth
er s
urge
ry p
atie
nts
are
shav
ed/p
repp
ed
in th
e O
R
Pati
ent i
s sh
aved
by
OR
NA
befo
re
tran
spor
tW
aits
for p
atie
nt
Wha
t det
erm
ines
ho
w lo
ng th
e N
A w
aits
for t
he p
atie
nt?
OR
NA
wai
ts fo
r pa
tien
t to
be re
ady
to tr
ansp
ort
EXH
IBIT
6.7
Swim
Lan
e Pr
oces
s M
ap
Not
e: IV
= in
trav
enou
s lin
e; N
A =
nur
sing
ass
ista
nt; O
R =
ope
rati
ng ro
om; S
PD =
ste
rile
pro
cess
ing
depa
rtm
ent;
SS
N =
Soc
ial S
ecur
ity
num
ber.
Chapter 6: Tools for Problem Solving and Decis ion Making 145
The Joint Commission (2013) requires all accredited organizations to conduct an RCA of any sentinel event (an unexpected occurrence involving death or serious physical or psychological injury, or the risk thereof) and pro-vides tools to help an organization conduct that analysis. Not only are these tools useful for resolving sentinel events and adhering to Joint Commission requirements but they also provide a framework for any RCA. A variety of commercial software is available for conducting RCAs.
Although an RCA can be conducted in many different ways, its basis is always in asking why something happened, again and again, until the ultimate cause is found. Typically, some element of the system or process, rather than human error, is found to be the ultimate cause. The five whys technique and cause-and-effect diagram are examples of tools used in RCA.
Five Whys TechniqueThe five whys technique is a simple yet powerful tool. It consists of asking why the condition occurred, noting the answer, and then asking why for each answer over and over (five times is a good guide) until the root causes are identi-fied. Often, the reason for a problem is only a symptom of the real cause. This technique can help eliminate the focus on symptoms, discover the cause, and point the way to eliminating it and ensuring that the problem does not occur again. The following list demonstrates how the five whys technique progresses:
1. A patient received the wrong medication. – Why?
2. The doctor prescribed the wrong medication. – Why?
3. Relevant information was missing from the patient’s chart. – Why?
Five whys techniqueA technique that uses a series of logical questions to find the root cause of a problem.
Customeractions
Line of interactionCustomer givesprescription to
clerk
Customerreceivesmedicine
Pharmacistgives medicine
to clerk
Pharmacistfills
prescription
Clerkentersdata
Clerk gives prescription to
pharmacist
Clerkretrievesmedicine
Clerk givesmedicine to
customer
Onstageactions
Backstageactions
Line of visibility
EXHIBIT 6.8Service Blueprint
Healthcare Operat ions Management146
4. The patient’s most recent lab test results were not entered into the chart. – Why?
5. The lab technician sent the results, but they were in transit and the patient’s record was not updated.
The root cause here is the time lag between the test and data entry. Rather than simply concluding that the doctor made a mistake, find the root cause to help determine different possible solutions to the problem. The system may now be changed to increase the speed with which lab results are recorded or, at least, to allow a note to be documented on the chart that lab tests have been ordered but not yet recorded.
Cause-and-Effect DiagramUsing only the five whys technique for an RCA can be limiting because of the assumption that an effect is the result of a single cause at each level of why. Often, a set of causes is related to an effect. A cause-and-effect diagram can overcome this limitation.
Typically, a team uses a cause-and-effect diagram to investigate and eliminate a problem. The problem should be stated or framed as clearly as possible, including who is involved and where and when the problem occurs, to ensure that everyone on the team is attempting to solve the same problem.
One of the seven basic quality tools, the cause-and-effect diagram is used to explore and display all of the potential causes of a problem. This type of graphic is sometimes called an Ishikawa diagram (after its inventor, Kaoru Ishikawa [1985]) or a fishbone diagram (because it looks like the skeleton of a fish).
The problem or outcome of interest is the “head” of the fish. The rest of the diagram consists of a horizontal line leading to the problem statement and several branches, or “fishbones,” vertical to the main line. The branches represent different categories of causes. The categories chosen may vary accord-ing to the problem, but some categories are commonly used (exhibit 6.9).
Fishbone diagramA graphical technique used to display the relationship between the potential causes of a problem and the effect created by the problem. Sometimes called Ishikawa diagram.
Service (Four Ps) Manufacturing (Six Ms)
Policies Machines
Procedures Methods
People Materials
Plant/technology Measurements
Mother Nature (environment)
Manpower (people)
EXHIBIT 6.9Typical Cause-
and-Effect Diagram
Categories
Chapter 6: Tools for Problem Solving and Decis ion Making 147
Possible causes are attached to the appropriate branches. Each possible cause is examined to determine if a deeper cause lies behind it (stage C in exhibit 6.10); subcauses are attached as additional bones. In the final diagram, causes are arranged according to relationships and distance from the effect. This arrangement can help identify areas to focus on and allow comparison of the relative importance of different causes.
Cause-and-effect diagrams can also be drawn as tree diagrams. From a single outcome, or trunk, branches extend to represent major categories of inputs or causes that create that single outcome. These large branches then lead to smaller and smaller branches of causes all the way out to twigs at the
Old inner-city building
Lack of treatmentrooms
Elevatorsbroken
Wheelchairsunavailable
Transport arrives late
Process takestoo long
Excessive paperwork
Unexpected patients
Wrongpatients
Staff not available
Corridorblocked
Sick
Late
Disorganized files
Bureaucracy
Original appointment missed
Incorrect referrals
Lack of technology
Poor scheduling
Poor maintenance
HIPAAregulations
Waiting time
Waiting time
Methods
Machines
Mother Nature(environment)
Mother Nature(environment)
Waiting time
Methods
Machines Manpower(people)
Manpower(people)
(C)
(A)
(B)
EXHIBIT 6.10Cause-and-Effect Example
Healthcare Operat ions Management148
ends. A process-type cause-and-effect diagram (exhibit 6.11) can be used to investigate causes of problems at each step in a process. A process RCA is similar to FMEA (as discussed in detail later) but is less quantitative in nature.
An example from VVH illustrates the cause-and-effect diagramming process. The hospital has identified excessive waiting time as a problem, and a team is assembled to address the issue. The problem is placed in the head of the fish, as shown in exhibit 6.10, stage A. Next, branches are drawn off the large arrow representing the main categories of potential causes. Typical categories are shown in exhibit 6.10, stage B, but the categories selected should suit the particular situation. Then, all of the possible causes inside each main category are identified. Each cause should be thoroughly explored to identify the causes of causes. This process continues, branching off into more and more causes of causes, until every possible cause has been identified (stage C in exhibit 6.10).
Much of the value gained from building a cause-and-effect diagram comes from undertaking the exercise itself with a team of people. A common and deeper understanding of the problem develops, enabling ideas to emerge for further investigation.
Once the cause-and-effect diagram is complete, an assessment of the possible causes and their relative importance should be undertaken. Obvious, easily fixable causes can be dealt with quickly. Additional data may be needed to assess the more complex possible causes and solutions. A Pareto analysis (chapter 7) of the various causes is often used to separate the vital few from the trivial many.
Building a cause-and-effect diagram is not necessarily a onetime exercise. The diagram can be used as a working document and updated as more data are collected and various solutions are tried.
Order inwrong place
Wrong test
Wronginformation
Phone busy
Undecipherablehandwriting
Dispatcherbusy
No forms
Technician unavailable
Pager doesnot work
Long timeto obtain
test results
Dispatchersends to
technician
Secretarycalls
dispatcher
Doctororders test
EXHIBIT 6.11Process-Type
Cause-and-Effect Diagram
Chapter 6: Tools for Problem Solving and Decis ion Making 149
Failure Mode and Effects AnalysisThe failure mode and effects analysis (FMEA) process was developed by the US military in the late 1940s, originally aimed at equipment failure. More recently, FMEA has been adopted by many service industries, including health-care, to evaluate process failure. Hospitals accredited by The Joint Commission are required to conduct at least one FMEA or similar proactive analysis annually (JCR and JCI 2010). Whereas RCA is used to examine the underlying causes of a particular event or failure, FMEA is used to identify the ways in which a process (or piece of equipment) might potentially fail, and its goal is to elimi-nate or reduce the severity of such a potential failure. By proactively looking at the potential causes of failure, risk of failure is either eliminated or reduced.
A typical FMEA consists of the following steps:
1. Identify the process to be analyzed. Typically, this process is the highest priority for the organization.
2. Assemble and train the team. Processes usually cross functional boundaries; therefore, the analysis should be performed by a team of relevant personnel. No one person or functional area has the knowledge needed to perform the analysis.
3. Develop a detailed process flowchart, including all steps in the process.4. Identify each step or function in the process.5. Identify potential failures (or failure modes) at each step in the process.
Note that more than one failure may potentially occur at each step.6. Determine the worst potential consequence (or effect) of each possible failure.7. Identify the cause(s) (contributory factor) of each potential failure. An
RCA can be helpful in this step. Note that each potential failure may have more than one cause.
8. Identify any failure “controls” that are present. A control reduces the likelihood that causes or failures will occur, reduces the severity of an effect, or enables the occurrence of a cause or failure to be detected before it leads to the adverse effect.
9. Rate the severity of each effect (on a scale of 1 to 10, with 10 being the most severe). This rating should reflect the impact of any controls that reduce the severity of the effect.
10. Rate the likelihood (occurrence score) that each cause will occur (on a scale of 1 to 10, with 10 being certain to occur). As with step 9, this rating should reflect the impact of any controls that reduce the likelihood of occurrence.
11. Rate the effectiveness of each control (on a scale of 1 to 10, with 1 being an error-free detection system).
12. Multiply the three ratings by one another to obtain the risk priority number (RPN) for each cause or contributory factor.
Failure mode and effects analysis (FMEA)A technique developed by the US military to identify the ways in which a process (or piece of equipment) might fail and to determine how best to mitigate those risks.
Healthcare Operat ions Management150
13. Use the RPNs to prioritize problems for corrective action. All causes that result in an effect with a severity of 10 should be high on the priority list, regardless of RPN.
14. Develop an improvement plan to address the targeted causes (who, when, how assessed, etc.).
Exhibit 6.12 is an example of an FMEA for patient falls from the Institute for Healthcare Improvement (IHI). IHI provides an online interactive tool for FMEA and shares many real-world examples that can be used as a basis for FMEAs in other organizations (IHI 2016). The National Center for Patient Safety (2015) of the US Department of Veterans Affairs has developed a less complex FMEA process, which rates only the severity and probability of occur-rence and uses the resulting number to prioritize problem areas.
Theory of ConstraintsThe theory of constraints (TOC) was first described in the business novel The Goal (Goldratt and Cox 1986). The TOC maintains that every organization is subject to at least one constraint that limits its movement toward or achieve-ment of its goal. For many organizations, the goal is to make money now as well as in the future. Some healthcare organizations may have a different, but still identifiable, goal. Eliminating or alleviating the constraint can enable the organization to move toward its goal. Constraints can be physical (e.g., the capacity of a machine) or nonphysical (e.g., an organizational procedure).
Five steps are involved in the TOC:
1. Identify the constraint or bottleneck. What is the limiting factor stopping the system or process from achieving the goal?
2. Exploit the constraint. Determine how to get the maximum performance out of the constraint without major system changes or capital improvements.
3. Subordinate everything else to the constraint. Other nonbottleneck resources (or steps in the process) should be synchronized to match the output of the constraint. Idleness at a nonbottleneck resource costs nothing, and nonbottlenecks should never produce more than can be consumed by the bottleneck resource. For example, if the operating room is a bottleneck and it has an adjacent or associated surgical ward, a traditional view might encourage filling the ward. However, nothing would be gained—and operational losses would be incurred—by putting more patients on the ward than the operating room can serve. Thus, the TOC solution is to lower ward occupancy to match the operating room’s throughput, even if resources (heating, lighting, fixed staff costs, etc.) seem to be wasted.
4. Elevate the constraint. Take some action (expend capital, hire more people, etc.) to increase the capacity of the constraining resource until
Theory of constraints (TOC)The idea that every organization and process is subject to at least one constraint that limits its movement toward or achievement of its goal.
Chapter 6: Tools for Problem Solving and Decis ion Making 151
it is no longer the constraint. Some other factor will become the new constraint.
5. Repeat the process for the new constraint.
The process must be reapplied, perhaps several times. Many constraints are of an organization’s own making, through entrenched rules, policies, and
Source: IHI (2005). This material was accessed from the Institute for Healthcare Improvement’s website, IHI.org. www.ihi.org/ihi/workspace/tools/fmea/ViewTool.aspx?ToolId=1248.
EXHIBIT 6.12Patient Falls FMEA
Healthcare Operat ions Management152
procedures that have developed over time. Avoid allowing inertia to become one of those constraints.
The TOC and Operations MeasurementThe TOC defines three operational measurements for organizations:
1. Throughput—the rate at which the system generates money, in the form of selling price minus cost of raw materials (labor costs are part of operating expense rather than throughput).
2. Inventory—the amount of money the system has invested in products or services it will sell; inventory includes the products on hand as well as buildings, land, and equipment.
3. Operating expense—the amount of money the system spends turning inventory into throughput, including what is typically called overhead.
The following four measurements are then used to identify results for the organization:
Net profit = Throughput − Operating expenseReturn on investment = (Throughput − Operating expense) ÷ InventoryProductivity = Throughput ÷ Operating expenseTurnover = Throughput ÷ Inventory
These measurements can help employees make local, or frontline, deci-sions. A decision that results in increasing throughput, decreasing inventory, or decreasing operating expense generally is a good decision for the organization.
The TOC has been applied in healthcare at both a macro and micro level to analyze and improve systems. De Mast and colleagues (2011) developed a model that demonstrated a 37 percent increase of patients through a system, accounting for more than $300,000 in increased revenue. In a CT (computed tomography) scanning department, the model was deployed to help improve the utilization of the scanning room, which was identified as the constraint in the process. The model raised utilization of the bottleneck from 88 percent to more than 93 percent.
Stratton and Knight (2010) used the TOC to help improve patient flow. The results of their study show a nearly 25 percent reduction in overall length of stay—from 8.6 days to 6.3. In this instance, patient length of stay was reduced because the hospital was able to keep the constraint working on critical items by managing time effectively. Because the TOC focused on the entire hospital system, the researchers were able to demonstrate the theory in practice—and improve systems—by working on the discharge process. For a surgical suite or an emergency department to serve more patients, it would need to accommodate each patient with an available hospital room. This model
Chapter 6: Tools for Problem Solving and Decis ion Making 153
helped free up rooms so that patients could move through the system more quickly than before the TOC was applied.
Another way to manage constraints in a system is to accept that a bot-tleneck will always exist and to determine where it should be. Designing the system so that the bottleneck can be managed or controlled is a powerful way to deal with it.
Analytical ToolsOptimizationOptimization, or mathematical programming, is a technique used to deter-mine the ideal allocation of limited resources given a desired goal. In other words, of all possible resource allocations—people, money, or equipment—the goal or objective is to find the allocation(s) that maximizes or minimizes some numerical quantity, such as profit or cost.
Optimization problems are classified as linear or nonlinear depending on whether the problem is linear with respect to the variables. In many cases, it is not practically possible to determine an exact solution for optimization problems; a variety of software packages offer algorithms to find good solutions.
Optimization models have three basic elements:
• An objective function—the quantity that needs to be minimized or maximized
• The controllable inputs or decision variables that affect the value of the objective function
• Constraints that limit the values that the decision variables can take on
A solution in which all of the constraints are satisfied is called a feasible solution. Most algorithms used to solve these types of problems begin by find-ing feasible solutions, and then they attempt to improve on those solutions until a maximum or minimum is found.
Healthcare organizations need to maintain financial viability while work-ing within various constraints on their resources. Optimization techniques can help these organizations make the best allocation decisions. An example of how linear programming can be used in a healthcare organization using Microsoft Excel Solver follows.
Linear Programming ExampleVVH wants to determine the optimal case mix for diagnosis-related groups (DRGs) that will maximize profits. Limited resources (e.g., space, qualified employees) are available to serve patients classified in the various DRGs, and minimum levels of service (number of cases) must be achieved for each DRG (exhibit 6.13).
OptimizationA technique used to determine the ideal allocation of limited resources (e.g., people, money, equipment) given a desired goal. Also called mathematical programming.
Healthcare Operat ions Management154
Exhibit 6.14 shows that the respiratory DRG (DRGr) requires 7 hours of diagnostic services, 1 intensive care unit (ICU) bed day, 5 routine bed days, and 50 hours of nursing care. The profit for DRGr is $400, and the minimum service level is 15 cases.
RespiratoryCoronay Surgery
Birth/Delivery
Alcohol/ Drug Abuse Available
Resources
Diagnostic services (hours)
7 10 2 1 325
ICU bed days 1 2.5 0.5 0 55
Routine bed days 5 7 2 7 420
Nursing care (hours)
50 88 27 50 3,800
Margin $400.00 $2,500.00 $300.00 $50.00
Minimum cases 15 10 20 10
Note: DRG = diagnosis-related group; ICU = intensive care unit.
EXHIBIT 6.13DRG Linear
Programming Problem Data
Note: DRG = diagnosis-related group; ICU = intensive care unit.
EXHIBIT 6.14Excel Solver
Setup for DRG Linear
Programming Problem
Chapter 6: Tools for Problem Solving and Decis ion Making 155
The goal is to maximize profit, and the objective function is
($400 × DRGr) + ($2,500 × DRGcs) + ($300 × DRGbd) + ($50 × DRGada),
where the other DRGs are classified as coronary surgery (DRGcs), birth/delivery (DRGbd), and alcohol/drug abuse (DRGada).
Diagnostic services:
(7 × DRGr) + (10 × DRGcs) + (2 × DRGbd) + (1 × DRGada) ≤ 325 (1)
ICU bed days:
(1 × DRGr) + (2.5 × DRGcs) + (0.5 × DRGbd) ≤ 55 (2)
Routine bed days:
(5 × DRGr) + (7 × DRGcs) + (2 × DRGbd) + (7 × DRGada) ≤ 420 (3)
Nursing care:
(50 × DRGr) + (88 × DRGcs) + (27 × DRGbd) + (50 × DRGada) ≤ 3,800 (4)
Respiratory minimum case level:
DRGr ≥ 15 (5)
Coronary surgery minimum case level:
DRGcs ≥ 10 (6)
Birth/delivery minimum case level:
DRGbd ≥ 20 (7)
Alcohol/drug abuse minimum case level:
DRGada ≥ 10 (8)
Exhibit 6.15 shows the Excel Solver setup of this problem. As previously shown in exhibit 6.14, Solver finds that the hospital should provide service for 15 DRGr cases, 12 DRGcs cases, 20 DRGbd cases, and 29 DRGada cases. The total profit at the optimal case mix is
Healthcare Operat ions Management156
(15 × $400) + (12 × $2,500) + (20 × $300) + (29 × $50) = $43,450.
Information relating to the resource constraints is found in the computer solution (exhibit 6.15). The amounts reported as slack, or surplus, provide a measure of resource utilization. All available ICU bed days and hours of nursing care will be used. However, 17 routine bed days and almost 31 hours of diagnostic services will be unused. VVH may want to consider eliminating some hours of diagnostic services. Constraints 5 through 8 relate to the mini-mum service level for each DRG category. Slack values represent services that should be provided in excess of a minimum level. Only the minimum levels for birth/delivery and respiratory care should be provided. However, 2 additional coronary surgery and 19 alcohol/drug abuse cases should be taken.
Sensitivity AnalysisSensitivity analysis (exhibit 6.16) examines the impact of varying the assump-tions, or input variables, on the output of a model. In the exhibit, a sensitivity analysis has been conducted to analyze the allocation and utilization of resources (diagnostic service hours, ICU bed days, routine bed days, nursing care) in relation to the objective function (total profit). Shadow prices (the Lagrange multiplier in exhibit 6.16) show the dollar effect on total profit of adding or deleting one unit of the resource. This analysis allows the organization to
Sensitivity analysisA tool that examines the impact of independently changing input variables to see their effect on the output of a model.
Target Cell (Max)
Cell Name Original Value Final Value
00.052,3latoT nigraM9$I$ $ 43,454.00$
Adjustable Cells
Cell Name Original Value Final Value
$B$13 Optimal Cases Respiratory 1 15$C$13 Optimal Cases Coronary Surgery 1 12$D$13 Optimal Cases Birth/Delivery 1 20$E$13 Optimal Cases Alcohol/Drug Abuse 1 29.08
Constraints
Cell Name Cell Value Formula Status Slack
$I$4 Diagnostic Services (hours) Total 294.08 $I$4 ^=$G$4 Not Binding 30.925$I$55latoT syaD deB UCI5$I$ ^=$G$5 Binding 0
$I$6 Routine Bed Days Total 402.56 $I$6 ^=$G$6 Not Binding 17.44$I$7 Nursing Care (hours) Total 3800 $I$7 ^=$G$7 Binding 0$E$13 Optimal Cases Alcohol/Drug Abuse 29.08 $E$13
^
=$E$11 Not Binding 19.08$D$13 Optimal Cases Birth/Delivery 20 $D$13
^
=$D$11 Binding 0$C$13 Optimal Cases Coronary Surgery 12 $C$13
^
=$C$11 Not Binding 2$B$13 Optimal Cases Respiratory 15 $B$13
^
=$B$11 Binding 0
Note: DRG = diagnosis-related group.
EXHIBIT 6.15Excel Solver Solution for DRG Linear
Programming Problem
Chapter 6: Tools for Problem Solving and Decis ion Making 157
weigh the relative benefits of adding resources. In this example, adding one ICU bed day would increase total profit by $964.80, and adding one hour of nursing care would increase total profit by $1. If the cost of either of these options is less than the additional profit, the hospital should increase those resources. Because slack is present in routine bed days and diagnostic services, adding more of either of these resources would not change the total profit; these resources are already in excess.
Shadow price information is also presented for the DRG minimum service level requirements (the reduced gradient in exhibit 6.16). The shadow price is −$614.80 for DRGr; total profit will decrease by $614.80 for each case taken above the minimum level required in the DRGr category. The DRGr category has a higher profit ($400) than the DRGada and, without this analy-sis, the hospital might have mistakenly tried to serve more DGRr cases, to the detriment of DRGada cases.
Optimization analysis also allows organizations to run what-if analyses. For example, if a hospital wants to investigate the possibility of increasing beds in its ICU, perhaps by decreasing routine beds, it could use optimization to analyze the available choices.
Decision AnalysisDecision analysis is a process for examining and evaluating decisions in a struc-tured manner. A decision tree is a graphic representation of the order of events in a decision-making process. This structured process enables an organization to evaluate the risks and rewards of choosing a particular course of action.
Decision analysisA structured process for examining and evaluating decisions.
Decision treeA graphical representation of the order of future and current events for how decisions are made.
Adjustable Cells
FinalValue
ReducedGradientCell Name
$B$13 Optimal Cases Respiratory 15 –614.8$C$13 Optimal Cases Coronary Surgery 12 0$D$13 Optimal Cases Birth/Delivery 20 –209.4$E$13 Optimal Cases Alcohol/Drug Abuse 29.08 0
Constraints
FinalValue
LagrangeMultiplierCell Name
$I$4 Diagnostic Services (hours) Total 294.08 08.46955latoTsyaDdeBUCI5$I$
$I$6 Routine Bed Days Total 402.56 0$I$7 Nursing Care (hours) Total 3800 1
EXHIBIT 6.16Sensitivity Analysis for DRG Linear Programming Problem
Note: DRG = diagnosis-related group; ICU = intensive care unit.
Healthcare Operat ions Management158
In the construction of a decision tree, events are linked from left to right in the order in which they would occur. Three types of events, represented by nodes, can take place: decision or choice events (squares), chance events (circles), and outcomes (triangles). Probabilities of chance events occurring and benefits or costs for event choices and outcomes are associated with each branch extending from a node. The result is a tree structure with branches for each event extending to the right.
A simple example helps illustrate this process. A health maintenance organization (HMO) is considering the economic benefits of a preventive influenza vaccination program. If the program is not offered, the estimated cost to the HMO if a flu outbreak occurs is $8 million with a probability of occurrence of 0.4 (40 percent) and $12 million with a probability of 0.6 (60 percent). The program is estimated to cost $7 million, and the probability of a flu outbreak occurring is 0.7 (70 percent). If a flu outbreak does occur and the HMO offers the program afterward, it will still cost the organization $7 million, but the resulting costs to the HMO would be reduced to $4 million with a probability of 0.4 (40 percent) or $6 million with a probability of 0.6 (60 percent). What should the HMO decide? The decision tree for the HMO vaccination program is shown in exhibit 6.17.
HMO vaccination
decision
Program
Program
No program No program
Flu outbreak
No flu outbreak
Flu outbreak
No flu outbreak A
B
C
D
Note: The tree diagrams in exhibits 6.17 through 6.21 were drawn with the help of PrecisionTree, a software product of Palisade Corp., Ithaca, NY: www.palisade.com.
EXHIBIT 6.17HMO
Vaccination Program
Decision Tree 1
Chapter 6: Tools for Problem Solving and Decis ion Making 159
The probability estimates for each chance node, benefits (in this case, costs) of each decision branch, and outcomes at the end of each branch are added to the tree (exhibit 6.18).
The value of a node can be calculated once the values for all subsequent nodes are found. The value of a decision node is the largest value of any branch out of that node. The assumption is that the decision that maximizes the benefits will be made. The value of a chance node is the expected value of the branches out of that node. Working from right to left, the value of all nodes in the tree can be calculated. The expected value of chance node 6 is [0.6 × (–12)] + [0.4 × (–8)] = –10.4. The expected value of chance node 5 is [0.6 × (–6)] + [0.4 × (–4)] = –5.2. The expected value of the secondary vaccination program is –7 + (–5.2) = –12.2, and the expected value of not implementing the secondary vaccination program is –10.4. Therefore, at decision node 4, the choice would be to not implement the secondary vaccination program.
At chance node 3 (no initial vaccination program), the expected value is [0.7 × (–10.4)] + (0.3 × 0) = –7.28. The expected value at chance node 2 is 0.7 × 0 + 0.3 × 0 = 0, and the expected value of the initial vaccination program branch is –7 + 0 = –7. Therefore, at decision node 1, the choice would be to implement the initial vaccination program at a cost of $7 million, as choosing not to implement the initial vaccination program is expected to cost $7.28 million (exhibit 6.19).
HMO vaccination
decision
Program
Program
No program No program
Flu outbreak
Flu outbreak
A
B
C
D
–$12,000,000
–$8,000,000
$0
30.0%
30.0%
70.0%
70.0%
60.0%
60.0%
40.0%
40.0%
$0
$0
1
2
3
4
5
6
$0
$0
$0
–$4,000,000
–$6,000,000
–$7,000,000
–$7,000,000
EXHIBIT 6.18HMO Vaccination Program Decision Tree 2
Healthcare Operat ions Management160
A risk analysis on this decision-making process can then be conducted (exhibit 6.20). Choosing to implement the vaccination program results in a cost of $7 million with a probability of 1. Choosing not to implement the initial vaccination program results in a cost of $12 million with a probability of 0.42, $8 million with a probability of 0.28, and no cost with a probability of 0.3. Choosing not to implement the vaccination program would be less costly 30 percent of the time, but 70 percent of the time, implementing it would be less costly.
A sensitivity analysis might also be conducted to determine the impact of changing some or all of the parameters in the analysis. For example, if the risk of a flu outbreak were 0.6 rather than 0.7 (and all other parameters stayed the same), the optimal decision would be to not offer either vaccination program 1, in the original HMO decision, or vaccination program 2, initiated after the flu outbreak later in the decision tree (exhibit 6.21).
HMO vaccination
decision
Program
Program
No program No program
Flu outbreak
No flu outbreak
Flu outbreak
No flu outbreak A
B
C
D
Choose this path because expected
costs of $10.4 million are less than $12.2
million.
–7
Vaccination program #1–7
Vaccination program #2–10.4
Flu–7
Flu–7.28
Choose this path because expected costs of $7 million are less than $7.28
million.
0
70.0%
30.0%
70.0%
30.0%
0 0
–7
–6
60.0%
–4
40.0%
60.0%
40.0%
–12
Costs–10.4
Costs–12.2
–8
0
0
0
EXHIBIT 6.19HMO
Vaccination Program
Decision Tree 3
Chapter 6: Tools for Problem Solving and Decis ion Making 161
For this example, dollars were used to represent costs (or benefits), but any type of score can be used. In the medical field, decision trees are often used to decide among a variety of treatment options and cost models for medical applications (Freitas 2011; Ribas et al. 2011).
Decision trees can be powerful aids to evaluating and choosing the optimal course of action. However, care must be taken when using them. Possible out-comes and the probabilities and benefits associated with them are only estimates, and these estimates may differ greatly from reality. Also, when using expected value (or expected utility) to choose the optimum path, the underlying assump-tion is that the decision will be made over and over. On average, the expected payout is received, but in each individual situation, different amounts are received.
Initial Vaccination Program No Initial Vaccination Program
Number X P X P
1 –7 1 –12 0.42
2 –8 0.28
3 0 0.30
Note: X = cost in millions of dollars; P = probability.
EXHIBIT 6.20Risk Analysis for HMO Vaccination Program Decision
HMO vaccination
decision
Program
Program
No program No program
Flu outbreak
No flu outbreak
Flu outbreak
No flu outbreak A
B
C
D
–7
Vaccinationprogram #1
–7 Vaccinationprogram #2
–10.4
Flu–7
Flu–6.24
0
60.0%
40.0%
60.0%
40.0%
0 0
–7
–6
60.0%
–4
40.0%
60.0%
40.0%
–12
Costs–10.4
Costs–12.2
–8
0
0
0
EXHIBIT 6.21Decision Analysis Sensitivity to Change in Risk of Flu Outbreak
Healthcare Operat ions Management162
Implementation: Force Field Analysis
Derived from the work of Kurt Lewin (1951), force field analysis is a technique for evaluating all of the various forces for and against a proposed change. It can be used to decide if a proposed change can be successfully implemented. Alternatively, if a decision to change has already been made, force field analysis can be used to develop strategies that enable the change to be implemented successfully.
In any situation, driving forces help to achieve a change, and restraining forces work against the change. Force field analysis identifies these forces and assigns relative scores to each. Exhibit 6.22 lists typical forces that should be considered. If the total score of the restraining forces is greater than the total score of the driving forces, the change may be doomed to failure. Force field analysis is typically used to determine how to strengthen or add driving forces or weaken the restraining forces to enable successful implementation of a change.
Application of Force Field Analysis at Vincent Valley Hospital and Health SystemPatients at VVH have expressed a belief that they are insufficiently involved in and informed about their care. After analyzing this problem, hospital staff expect to solve (or lessen) it by moving the location of shift change handovers from the nurses’ station to the patients’ bedsides. A force field analysis has been conducted and is illustrated in exhibit 6.23.
Although the restraining forces are greater than the driving forces in this example, the decision is made to implement the change in handover procedures. To improve the project’s chances for success, a protocol is developed for the actual procedure, making explicit the following guidelines:
• Develop and disseminate the protocol (new driving force +2).• Exchange confidential information at the nurses’ station, not at the
bedside handover (decrease fear of disclosure by 2).• Follow solution-based directives developed and incorporated into the
protocol to address delayed handovers (decrease problems associated with late arrivals by 2).
Force field analysisA graphical technique that demonstrates all the forces for and against making a key change.
Available resourcesCostsVested interestsRegulationsOrganizational structuresPresent or past practices
Institutional policies or norms
Personal or group attitudes and needs
Social or organizational norms and values
EXHIBIT 6.22Common Forces
to Consider in Force Field
Analysis
Chapter 6: Tools for Problem Solving and Decis ion Making 163
These changes thus increase the driving forces by 2, to 21, and decrease the restraining forces by 4, to 17. The change has been successfully imple-mented; more important, patients now feel involved in their care and the number of complaints is reduced.
Conclusion
The tools and techniques outlined in this chapter are intended to help orga-nizations along the path of continuous improvement. The choice of tool and when to use that tool depends on the problem to be solved; in many situations, several tools from this and other chapters should be used to ensure that the best possible solution is found.
Discussion Questions
1. Answer the following questions quickly for a fun illustration of some of the ten decision traps (Russo and Schoemaker 1989):
secroF gniniartseRsecroF gnivirD
Plan:Change to
bedside shifthandover
Critical incidentson the increase
Staff knowledgeable inchange management
Increase in dischargeagainst medical advice
Complaints from patientsand doctors increasing
Care given is predominantlybiomedical in orientation
Ritualism andtradition
Fear that this maylead to more work
Fear of increasedaccountability
Problems associatedwith late arrivals
Possible disclosure ofconfidential information
4
5
3
3
4
Total: 19
4
4
3
5
5
Total: 21
EXHIBIT 6.23Force Field Analysis
Healthcare Operat ions Management164
• Can a person living in Milwaukee, Wisconsin, be buried west of the Mississippi River?
• If you had only one match and entered a room with a lamp, an oil heater, and some kindling wood, which would you light first?
• How many animals of each species did Moses take along on the ark?• If a doctor gave you three pills and said to take one every half hour,
how long would they last?• If you have two US coins totaling 55 cents and one of the coins is
not a nickel, what are the two coins?What decision traps did you fall into when answering these questions?
2. Discuss a problem your organization solved or a suboptimal decision the organization made because the frame was incorrect.
Exercises
1. For the HMO vaccination program example provided in the chapter, reanalyze the situation assuming that the probability of a flu outbreak is 65 percent and the cost of the vaccination program is $8 million. What is your decision under these new conditions?
2. In the DRG case-mix problem, VVH determined that it could convert 15 of its routine beds to ICU beds for a cost of $2,000. What should VVH do, and why?
3. The high cost of medical care and insurance is a growing societal problem. Develop a mind map of this issue. (Advanced: Use Inspiration software.)
4. Individually or in teams, develop a map of a healthcare process or system with which you are familiar. Make sure that your process map has a start and an endpoint, all inputs and outputs are defined, and all key process steps are included. Explain your map to the rest of the class—this step may help you determine if anything is missing. (Advanced: Use Microsoft Visio.)
5. Choose a service offered by a healthcare organization, and create a service blueprint of it. You may have to imagine some of the systems and services that take place backstage if you are unfamiliar with them.
6. Think of a problem in your healthcare organization. Perform an RCA of the identified problem using the five whys technique and a fishbone diagram.
7. Pick one solution to the problem identified in exercise 6, and conduct a force field analysis of it.
Chapter 6: Tools for Problem Solving and Decis ion Making 165
References
Buzan, T. 1991. Use Both Sides of Your Brain. New York: Plume.Buzan, T., and B. Buzan. 1994. The Mind Map Book: How to Use Radiant Thinking to
Maximize Your Brain’s Untapped Potential. New York: Dutton.De Mast, J., B. Kemper, R. J. M. M. Does, M. Mandjes, and Y. van der Bijl. 2011. “Process
Improvement in Healthcare: Overall Resource Efficiency.” Quality and Reliability Engineering International. Published April 1. http://onlinelibrary.wiley.com/doi/10.1002/qre.1198/abstract.
Freitas, A. 2011. “Building Cost-Sensitive Decision Trees for Medical Applications.” AI Communications 24 (3): 285–87.
Goldratt, E. M., and J. Cox. 1986. The Goal: A Process of Ongoing Improvement. New York: North River Press.
Institute for Healthcare Improvement (IHI). 2016. “Failure Modes and Effects Analysis Tool.” Accessed August 24. http://app.ihi.org/Workspace/tools/fmea/.
———. 2005. “Workspace Tools.” Cambridge, MA: IHI.Ishikawa, K. 1985. What Is Total Quality Control? Translated by D. J. Lu. Englewood Cliffs,
NJ: Prentice-Hall.Joint Commission Resources (JCR) and Joint Commission International (JCI). 2010. Failure
Mode and Effects Analysis in Health Care: Proactive Risk Reduction, 3rd edition. Oakbrook Terrace, IL: JCR.
Lewin, K. 1951. Field Theory in Social Science: Selected Theoretical Papers, edited by D. Cartwright. New York: Harper.
National Center for Patient Safety, US Department of Veterans Affairs. 2015. “Healthcare Failure Mode and Effect Analysis (HFMEA).” Updated June 3. www.patientsafety.va.gov/professionals/onthejob/hfmea.asp.
Ribas, V. J., J. C. Lopez, J. C. Ruiz-Rodriguez, A. Ruiz-Sanmartin, J. Rello, and A. Vellido. 2011. “On the Use of Decision Trees for ICU Outcome Prediction in Sepsis Patients Treated with Statins.” Proceedings of the IEEE Symposium on Computational Intel-ligence and Data Mining, CIDM 2011. IEEE Symposium Series on Computational Intelligence 2011, April 11–15, Paris. Accessed August 24, 2016. http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=5949439&abstractAccess=no&userType=inst.
Russo, J. E., and P. J. H. Schoemaker. 1989. Decision Traps: The Ten Barriers to Brilliant Decision-Making and How to Overcome Them. New York: Doubleday.
Shostack, G. L. 1984. “Designing Services That Deliver.” Harvard Business Review 62 (1): 133–39.
Stratton, R., and A. Knight. 2010. “Managing Patient Flow Using Time Buffers.” Journal of Manufacturing Technology Management 21 (4): 484–98.
Wysocki, B., Jr. 2004. “To Fix Health Care, Hospitals Take Tips from Factory Floor.” Wall Street Journal, April 9, A1–A5.
CHAPTER
167
STATISTICAL THINKING AND STATISTICAL PROBLEM SOLVING
Operations Management in Action
In February 2016, the World Health Organization (WHO) and the Centers for Disease Control and Prevention (CDC) announced that the Zika virus was growing rapidly throughout the world (Botelho 2016). According to the CDC (2016a), the Zika virus is contracted and spread to humans by a particular species of mosquito. A Zika infection is typically not fatal, but pregnant women can transmit the virus to their unborn children, potentially resulting in birth defects (CDC 2016b). WHO estimated that Zika infections would reach 3 mil-lion to 4 million cases in the Americas over a 12-month period from 2016 to 2017 (Botelho 2016).
This estimate by WHO was devel-oped in the wake of intense criticism of the agency over its mishandling of the Ebola virus in Africa, starting with its inaccurate initial estimates of the number of cases expected (Botelho 2016). Ebola killed thousands of peo-ple across a portion of the continent despite humanitarian efforts to help control the outbreak. An internal WHO document reveals that its officials did not pay attention to evidence related to the rise of the virus, likely perpetuat-ing the deadly situation (Sanchez 2014).
7OVE RVIEW: STATISTICAL TH I NKI NG I N H EALTHC AR E
What Is Statistical Thinking?As defined by Joseph Juran, statistical thinking is the collection, orga-
nization, analysis, interpretation, and presentation of data (Juran and
De Feo 2010). In most business systems in the healthcare industry,
statistical thinking is lacking. Knowledge-based management and
improvement require that decisions be based on facts rather than
on feelings or intuition. Collecting the right data and analyzing them
correctly enable fact-based decision making.
Variance is present in all systems. The ability of leadership to
understand and control variance distinguishes high-performing sys-
tems from poorly run systems. Hospitals often ignore variance, and the
erratic behavior that underlies it continues until, as exemplified in this
text, major issues emerge that require system redesign (DeLia 2007).
The importance of understanding statistical concepts in devel-
oping high-performing healthcare systems cannot be overstated. Deliv-
ering high-quality healthcare in a sustained manner depends on under-
standing and controlling variance. The irony of this relationship is that
many clinical quality and safety rules and regulations are designed and
driven by the understanding of variance, while the supporting business
systems are often designed to meet regulatory agency requirements and
not to manage the variance in the system. The good news is that this situ-
ation provides the opportunity to make massive changes to both system
and financial performance simply by understanding data and metrics.
Metrics and Key Process IndicatorsThe terms metrics and key process indicators (KPIs) have become
increasingly prevalent in discourse about healthcare operations man-
agement in recent years. Hospitals and healthcare organizations are
constantly searching to find effective metrics that indicate the health
of their overall system. For healthcare systems, the term metrics can
Healthcare Operat ions Management168
The original estimates for the Ebola virus in Africa were demon-strated to be grossly underestimated (Melt-zer et al. 2014). Initial estimates of virus outbreaks help coun-tries and governments allocate resources to manage the situa-tion. Once officials determined that the Ebola estimates were low, the many groups involved in the response to the crisis had to react quickly to gain control of it.
While we can-not say for certain, we may reasonably assume that the mod-els created by WHO resulted in a wide range of estimates as to the incidence of Zika. To avoid the same level of criticism it encountered fol-lowing its estimates of Ebola, WHO likely published projec-tions that were much higher than predicted by most estimation models. If so, this sce-nario is a perfect dem-onstration of how bias and situation affect the interpretation of statistical models.
OVE RVI EW (Continued)
be difficult to grasp because it includes clinical metrics of safety
and quality as well as business system metrics. Each area has its
own challenges that hinder the collection of data.
To design effective and efficient systems or improve existing sys-
tems, knowledge of the system itself, including both inputs to the system
and the desired output, is needed. The goal of data collection is to obtain
valid information to enhance understanding toward improving the system
being studied. Decisions made or solutions implemented on the basis of
invalid data are doomed to failure. Ensuring that the data obtained are
valid is an important part of any study, and often the most problematic.
What constitutes valid data in a healthcare system can vary depending
on which system area—clinical or business—is being analyzed.
Clinical systems are appropriately designed around
patient safety and quality outcomes. Procedures and processes
are designed and tested under rigorous statistical guidelines.
The outcome of these statistical procedures is that the delivery of
care improves over time and patient safety and quality increase.
Business systems, on the other hand, usually develop over time to
meet the needs of the market and the technological needs of the
institution, with the net effect being that the business system must
flex to meet the ever-changing demands of the healthcare industry.
Thus, although these two areas are integral to any healthcare
system, their processes for data gathering may conflict, making
the collection of appropriate data for analysis challenging.
Efforts to change clinical processes require perfect data.
Clinical studies follow rigorous data collection procedures and
include control groups and test groups, and results are compared
under the most demanding statistical procedures. However, to
change business systems, the data only need to be “good enough.”
The data should point to the major problems and provide a basis
for how to change the system. The goal in business systems is
continuous improvement, not perfection. The needs of both types
of healthcare professionals—clinicians and business developers—
working in the same industry lead to arguments about data and
data integrity, which often result in little or no analysis.
Statistical thinking requires that our decisions be driven
by data and not by individual preferences. However, ultimately,
the goal is continuous improvement, and the opposite of progress
is doing nothing at all. The focus of this chapter is on providing a
solid understanding of data collection, measurement, and analysis.
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 169
Foundations of Data Analysis
To become an effective practitioner of continuous improvement and an effective business analyst, one must become adept at problem solving and data analy-sis, including knowledge of the fundamental issues related to data collection, basic probability, and statistical analysis. For those readers with little or no background in statistics or probability, this chapter provides an introduction to the basic concepts used in fundamental problem solving, many of which are integral to the continuous improvement philosophy of quality. For readers who wish to gain a greater understanding of statistics and probability, the book’s companion website has in-depth coverage of many statistical concepts and techniques. For purposes of this chapter, we discuss the following topics:
• Graphic tools for data presentation and analysis• Mathematical forms of data description• Probability, including basic and conditional probability• Confidence intervals and hypothesis testing• Linear regression
After reading this chapter, readers should understand the fundamental tools of problem solving and quality.
Where to Start?The critical error often made during problem solving is failing to understand what data are needed to solve the problem or how the data will be acquired.
A helpful first step is to establish why the data are needed and what they will be used for. Another useful consideration is whether the patterns of the past will be repeated in the future. If you have reason to believe that the future will look different from the past, data from the past will not help you answer the question and other, nonquantitative methods of problem solving should be used. This is the logic phase of the data collection process, where the focus is on ensuring that the right question is being asked and that the question can actually be answered.
Graphic Tools
A core technique of problem solving is to consider the data and problem visu-ally prior to studying the data analytically. This section discusses graphic data illustration techniques, including mapping, check sheets, histograms, Pareto charts, dot plots, and scatter plots.
On the web at ache.org/books/OpsManagement3
Healthcare Operat ions Management170
MappingMind mapping is a versatile graphic approach to data illustration because it can also be employed before the actual problem solving begins by enabling the collection of valid data. A mind map helps frame the problem or question in an attempt to avoid the commission of logic errors throughout the problem-solving process (exhibit 7.1; see also chapter 6).
Check SheetsAn essential tool used in problem solving is the check sheet. Check sheets are custom-designed forms that allow users to collect data on problems and defects. The form has checkbox items that describe typical problems in the system. When an employee uses a check sheet, he selects the appropriate box every time an error occurs. This type of tool is designed to collect data in real time as it is being created. The gathering of check sheet data is necessary prior to conducting analysis.
Although effective check sheets may be simple to develop and compre-hend, they are difficult tools to execute well. The most effective check sheets have the following characteristics:
• They are simple to use.• The data points reflect a consistent level of analysis.• They include just a few boxes for the user to check.• Data are collected as they occur.
CAUSES OF EMERGENCY ROOM
DELAYS
Technology
Admission does not classify patients
Room Scheduling
Rooms are assigned in random fashion
Need to manually enter in data
20 different service lines
No standardization to get a consult
Unclear admissions process
Not trained on admissions process
Personnel
Processes
Ancillary Services
EXHIBIT 7.1Causes of
Emergency Room Delays
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 171
When executed correctly, a check sheet allows a data analyst to access current data that can be used to demonstrate the state of a problem. However, many issues are encountered with the use of check sheets, mostly related to the process of collecting the data. Many people complete check sheets incor-rectly because the sheets are not clear; some staff fail to fill them out as the data are created.
Data Visualization TechniquesOnce valid data are collected, they need to be analyzed to answer the original question or make a decision. The data must be examined not only to deter-mine their general characteristics but also to look for interesting or unusual patterns. Subsequent sections of this chapter cover numeric tools that can be employed for this purpose.
The human mind is powerful and has the ability to discern patterns in data, which can then be validated through numeric methods. Visual representa-tions of the data aid in both answering questions and convincing others of the accuracy of those answers. This section taps insights on graphic analysis tools from Tufte (1997, 1990, 1983), which provide guidance on visually presenting data. The first step in data analysis is always to graph the data.
Histograms and Pareto DiagramsHistograms and Pareto diagrams are two of the seven basic quality tools intro-duced in chapter 6 and discussed in detail in chapter 8. A histogram (exhibit 7.2) is used to summarize discrete or continuous data. These graphs can be useful for investigating or illustrating important characteristics of the data, such as their overall shape, symmetry, location, and spread and the outliers, clusters, and gaps that emerge. Worth noting, however, is that for some distributions, a particular choice of bin width (interval in which frequency of data points is
HistogramA graph summarizing discrete or continuous data. Histograms visually display how much variation exists in the data.
0
2
4
6
8
10
12
14
Freq
uenc
y
LOS (days)
1–2 3–4 5–6 7–8 9–10 11–12 13–14 15–16 17–18
EXHIBIT 7.2Histogram of Hospital Length of Stay (LOS)
Healthcare Operat ions Management172
measured) can distort the features of a data set. For an example of this problem, see the Old Faithful His-togram applet linked from the companion website.
To construct a histogram, the data are divided or grouped into classes. For each group, a rectangle is created with its base equal to the range of values in the group and its area proportional to the number of observations falling into the group. If the ranges are the same length, the height of the histogram is also proportional to the number of observations falling into that group.
In this way, histograms allow an analyst to see the shape of the distri-bution of the data. She can quickly see if data points follow patterns of tight variation or wide variation or simply if some data points might be considered outliers (extremes; discussed later in the chapter) to the overall data set.
The histogram in exhibit 7.2, which depicts an example of hospital length of stay (LOS), demonstrates a distribution that is skewed to the right, revealing that the majority of inpatients stay between one and two days.
Pareto diagrams are a type of frequency diagram, which indicates the number of times a particular item occurs in a situation. The Pareto principle, or the 80/20 rule, dictates that 80 percent of costs, defects, or other types of issues are attributable to 20 percent of the items being measured. In exhibit 7.3, a hospital collected data related to a high percentage of late starts for surgeries. In the first diagram, 80 percent of all issues are related to just two issues: missing equipment at the start of surgery and late-arriving patients. Using another Pareto diagram to dissect the reasons for missing equipment
Pareto diagramA rank-ordered frequency chart that indicates the number of times a particular item occurs in a situation.
40
80
70
60
50
30
20
10
0
40.0%
90.0%
80.0%
100.0%
70.0%
60.0%
50.0%
30.0%
20.0%
10.0%
60.2%50
1712
31
80.7%
95.2%
98.8%
Mis
sing
equ
ipm
ent
Peop
lear
rive
d la
te
Prob
lem
wit
h se
tup
Ord
ers
not c
orre
ct
Tran
spor
t
Reason for Delays
40
50
30
20
10
0
40.0%
90.0%
80.0%
100.0%
70.0%
60.0%
50.0%
30.0%
20.0%
10.0%
64.0%
84.0%
92.0% 98.0%
Poor
lead
tim
e
Not
cle
an
Expi
red
Oth
er
Bro
ken
Reason for Missing Equipment
32
10
43
1
EXHIBIT 7.3Causes for
Delays in Surgery
On the web at ache.org/books/OpsManagement3
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 173
demonstrated that 64 percent of those cases had insufficient lead time to clean, prepare, and load the surgery cart to arrive in time for the surgery. This analysis allowed a problem-solving team to focus its efforts on improving those few activities that made an immediate impact on the situation.
Dot PlotsA dot plot (exhibit 7.4) is similar to a histogram; rather than showing points on a graph connected by a line, a dot plot represents frequency of occurrence by a dot. Dot plots are useful for displaying small data sets with positive values because they are quick and easy to construct by hand.
Scatter PlotsScatter plots are another of the seven basic quality tools. A scatter plot graphi-cally displays the relationship between a pair of variables and can offer an initial indication of whether two variables are related, how strongly they are related, and the direction of the relationship. For example, is a relationship present between hospital LOS and a patient’s weight? Does LOS increase (decrease) as weight increases? How strong is the relationship between LOS and weight? A scatter plot can help to answer these questions. Regression—the statistical tool related to scatter plots that gives more detailed, numeric answers to these questions—is discussed later in this chapter.
To construct a scatter plot related to the aforementioned questions, data on LOS and patient weight from the population of interest are collected. Typi-cally, the cause, or independent variable, is on the horizontal (x) axis and the effect, or dependent variable, is on the vertical (y) axis. Each pair of variables is plotted on this graph.
Scatter plots are useful tools for determining what variables in the system need to be controlled to obtain desired outputs. Much like a Pareto diagram, a scatter plot helps narrow the number of variables an analyst needs to consider in solving the problem. A typical scatter plot is shown in exhibit 7.5.
Dot plotA chart in which frequency is represented by a dot. Useful for displaying small data sets with positive values.
Scatter plotA graph displaying two variables that indicates whether they are related, how strongly they are related, and the direction of the relationship.
Days
181512963
EXHIBIT 7.4Dot Plot of Hospital Length of Stay
Healthcare Operat ions Management174
Exhibit 7.5 shows that the consumption of one to two glasses of wine per day has a positive effect on reducing vascular disease. This relationship is negative because increased wine consumption leads to a reduction in disease.
Mathematical Descriptions
When describing or summarizing data, the three characteristics of interest for any analyst are central tendency, spread or variation, and the probability distribution. In this section, the following simple data set is used to illustrate some of these measures: 3, 6, 8, 3, 5.
Measures of Central TendencyThe three common measures of central tendency are mean, median, and mode.
MeanThe mean is the arithmetic average of the population:
Population mean= µ=ΣxN
,
140
160
120
80
100
60
40
20
0
543210 6
Wine Consumption (dl/person/day)
SMR
for C
ereb
rova
scul
ar D
isea
se
Source: Reprinted from Truelsen and Grønbæk (1999).
Note: SMR = standardized mortality ratio.
EXHIBIT 7.5Scatter Plot
Between Wine Consumption and Vascular
Disease
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 175
where x = individual values and N = number of values in the population.
The population mean can be estimated from a sample:
Σ= =x
xn
Sample mean ,
where n = number of values in the sample. For our simple data set,
=+ + + +
=x3 6 8 3 5
55.
Referring to the histogram in exhibit 7.6, if the data shape looks like a bell curve, the mean is the point in the middle, or the average of all data.
MedianThe median is the middle value of the sample or population. If the data are arranged into an array (an ordered data set),
↑
3,3,5,6,8
5 is the middle value or median.
ModeThe mode is the most frequently occurring value. In the previous example, the value 3 occurs more often (two times) than any other value, so 3 is the mode.
20
25
15
10
5
0
50%40%30%20%10%0% 60%
Practice-Level SQUID Value
Freq
uenc
y (n
umbe
r of p
ract
ices
)
EXHIBIT 7.6Histogram of Summary Quality Index (SQUID)
Source: Reprinted from Truelsen and Grønbæk (1999).
Healthcare Operat ions Management176
Measures of VariabilitySeveral measures are commonly used to summarize the variability of the data, including range, mean absolute deviation, variance, standard deviation, coef-ficient of variation, and outliers.
RangeA simple way to capture the variation or spread in the data is to determine the range—the difference between the high and low values. All of the information in the data is not being used with this measure, but it is simple to calculate, as shown here with our sample data set:
Range = xhigh − xlow = 8 – 3 = 5.
Mean Absolute DeviationAnother possible measure of the variability or spread in the data is the aver-age difference from the mean. However, for any data set this average equals zero, because the values above the mean always balance the values below the mean. One way to eliminate this problem is to determine the absolute value of the differences from the mean. This measure is called the mean absolute deviation (MAD) and is commonly used in forecasting to measure variability. For the sample data set,
x xn
MAD| | 2 1 3 2 0
585
1.6,Σ
=−
=+ + + +
= =
where n is the number of values in the sample.Because absolute values are difficult to work with mathematically, we
do not cover them in depth here.
VarianceThe average square difference from the mean—called the variance—provides another measure of the variability in data. Variance is a good measure of devia-tion from the mean in a population. However, for a sample, it can be proven that variance is a biased estimator and needs to be adjusted; rather than dividing the numerator by n, it must be divided by n – 1:
Population variance= 2 =(x μ)2
N=
4+1+9+ 4+05
=185= 3.6
Sample variance= s 2 =(x x)2
n 1=
4+1+9+ 4+05 1
=184= 4.5
VarianceA statistical term that indicates how much a measurement varies around the mean.
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 177
Standard DeviationCalculating the square root of the variance results in the units of this measure being the same as the units of the mean, median, and mode. This measure, the standard deviation, is the most commonly used measure of variability.
Population standard deviation= 2
=(x μ)2
N
=4+ 4+0+1+9
5=
185= 3.6 =1.9
Σ=
=−
=+ + + +
−= = =
s
x xn
Sample standard deviation
( )
4 4 0 1 95 1
184
4.5 2.1
2
2
Coefficient of VariationThe coefficient of variation (CV) indicates the amount of variation relative to the mean. The CV is computed by dividing the mean by the standard devia-tion. The larger the mean relative to the standard deviation, the less relative variation exists in the data.
CV =σµ= 5
or
= =sx
1.95
0.4.
OutliersOutliers are observations that are far from the mean or median in the data set. An outlier is an important discovery because it represents an opportunity for analysts to seek improvements in that area.
If the histogram data are reasonably bell shaped, we use Shewhart’s rule to determine if outliers are present in the data. Shewhart’s rule indicates that outliers are present if the data points are greater than the mean at a rate of ±3 × standard deviation.
Standard deviationA measurement of variation around the mean.
Coefficient of variation (CV)A measure of variation in the data relative to the measure of central tendency in the data.
Shewhart’s ruleAn outlier exists in bell-shaped data if a data point is greater than three standard deviations from the mean.
Healthcare Operat ions Management178
If the histogram data are skewed (not bell shaped), we use Tukey’s rule to determine if outliers are present in the data:
Q1 − 1.5 × IQR
or
Q3 + 1.5 × IQR,
where Q1 and Q3 represent the first and third quartiles of the data set and IQR is the interquartile range. IQR is computed by subtracting Q1 from Q3.
Probability
A common belief in healthcare systems is that events related to illness are not predictable. These types of events are more predictable than most people real-ize, and the laws of probability help explain the likelihood of events occurring. Many issues arise in healthcare systems because the impact of probability on the system is not understood. For example, not understanding the probability of increased admittance to the hospital could create a situation in which beds are not available to patients who need them.
Two types of models explain what is seen in the world: deterministic and probabilistic. In a deterministic model, the given inputs determine the output with certainty. For example, given a person’s date of birth and the current date, his age can be determined. The inputs determine the output:
Date of BirthCurrent Date Age Model Person’s Age
Age Life SpanModel
Person’sRemaining LifeSpan
In a probabilistic model, the given inputs provide only an estimate of the out-put. For example, given a person’s age, her remaining life span can only be estimated:
Date of BirthCurrent Date Age Model Person’s Age
Age Life SpanModel
Person’sRemaining LifeSpan
Determination of ProbabilitiesProbabilities can be determined (1) through observation or experimentation, (2) by applying theory or reason, or (3) subjectively through opinion making.
Observed ProbabilityObserved probability is a summary of the observations or experiments involved in determining probability and is referred to as empirical probability or relative
Tukey’s ruleAn outlier exists in a skewed data set if a data point is greater than Q1 − one step or Q3 + one step, where one step = 1.5 × IQR.
Observed probabilityThe number of times an event occurred divided by the total number of trials.
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 179
frequency. Observed probability is the relative frequency of an event—the number of times the event occurred divided by the total number of trials.
= =Prn
(A)Number of times A occured
Total number of observations, trials, or experiments,
where P is probability, A is the event, r is rate, and n is number of trials.Drug or protocol effectiveness is often determined in this manner:
= =Prn
(Drug is effective)Number of times patients cured
Total number of patients given the drug.
For business analysts, observed probability is the most commonly applied probability type because it gives an accurate representation of how the system or processes are functioning.
Theoretical ProbabilityThe second method of determining probability, the theoretical relative fre-quency of an event, is based on logic—it is the theoretical number of times an event will occur divided by the total number of possible outcomes:
= =Prn
(A)Number of times A could occur
Total number of possible outcomes.
Casino revenues are based on this theoretical determination of probabil-ity. If a card is randomly selected from a common deck of 52, the probability that it will be a spade is determined as follows:
= = =P(Card is a spade)Number of spades in the deck
Total number of cards in the deck1352
0.25.
Theoretical probability is often used by health insurance companies to predict the number of occurrences of disease and illness to set premium rates.
Properties of ProbabilitiesBounds on ProbabilityProbabilities are bounded, such that the least number of times an event could occur is zero; therefore, probabilities must always be greater than or equal to zero. An event that cannot occur has a probability of zero. The largest num-ber of times, t, an event could occur is equal to the total possible number of outcomes—t cannot be any larger; therefore, probabilities must always be less than or equal to 1:
0 ≤ P(A) ≤ 1.
Theoretical probabilityThe number of times an event will occur divided by the total number of possible outcomes.
Healthcare Operat ions Management180
The sum of the probabilities of all possible outcomes is 1. From this property, it follows that
P(A) + P(A′) = 1
and
1 – P(A) = P(A),
where A′ is “not A,” meaning A does not occur. This property can be useful when determining probabilities, as determining the probability of not A is often easier than finding the probability of A.
Multiplicative PropertyTwo events are independent if the outcome of one event does not affect the outcome of the other event. For two independent events, the probability of both A and B occurring, or the intersection (∩) of A and B, is the probability of A occurring multiplied by the probability of B occurring:
P(A and B occurring) = P(A ∩ B) = P(A) × P(B).
For example, when combining a coin toss with a die toss, we can deter-mine the probability of obtaining both heads and a three:
P (H ∩ 3) = P (H) × P (3) = 1
2 ×
1
6 =
1
12.
A tree diagram (exhibit 7.7) or a Venn diagram (exhibit 7.8) can be used to illustrate this property. (Note that decision trees, discussed in chapter 6, are different from the tree diagrams presented here. Decision trees follow a time pro-gression and analyses of the choices that can be made at particular points in time.)
The multiplicative property provides a way to test whether events are independent. If they are not independent,
P(A ∩ B) ≠ P(A) × P(B).
Additive PropertyFor two events, the probability of A or B occurring—the union (∪) of A with B—is the probability of A occurring plus the probability of B occurring minus the probability of both A and B occurring:
P(A or B occurring) = P(A ∪ B) = P(A) + P(B) + P(A ∩ B).
Building on the earlier example, when combining a coin toss with a die toss, we can determine the probability of obtaining heads or a three, but not both:
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 181
P H P H P P H( 3) ( ) (3) ( 3)12
16
112
612
212
112
712
.
∪ = + − ∩
= + − = + − =
A tree diagram (exhibit 7.9) or Venn diagram can be used to illustrate the additive property.
Coin Toss Die Toss Probability
1 1/12
2 1/12
H3 1/12
4 1/12
5 1/12
6 1/12Start
1 1/12
2 1/12
T
3 1/12
4 1/12
5 1/12
6 1/12
P(H) = 1/2 P(3) = 1/6
P(H 3) = 1/12
EXHIBIT 7.7Tree Diagram—Multiplicative Property
Toss heads Toss 3
1 515
EXHIBIT 7.8Venn Diagram—Multiplicative Property
Healthcare Operat ions Management182
Conditional ProbabilityConditional probability estimates how frequently events occur after a previ-ous event has taken place. For example, suppose a patient usually waits in the emergency department (ED) for fewer than 30 minutes before being moved into an examination room. However, on Friday nights, when the department is busy, the wait is longer; the probability of waiting 30 minutes or less is lower. This is the conditional probability of waiting less than 30 minutes given that the time frame of interest is a Friday night.
The conditional probability that A will occur given that B has occurred is as follows:
PP
P(A | B)
(A B)(B)
.=∩
Now suppose a study were conducted of 100 ED patients in which 50 patients were observed on a Friday night and 50 patients were observed at other
Coin Toss Die Toss Probability
1 1/12
2 1/12
H3 1/12
4 1/12
5 1/12
6 1/12Start
1 1/12
2 1/12
T
3 1/12
4 1/12
5 1/12
6 1/12
7/12P(H) = 1/2 P(3) = 1/6
P(H 3) = 1/12
P(H 3) = 7/12
EXHIBIT 7.9Tree Diagram—
Additive Property
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 183
times. On Friday night, 20 people waited less than 30 minutes, but 30 people waited longer than 30 minutes. At other times, 40 people waited less than 30 minutes, and only 10 people waited longer than 30 minutes. A contingency table (exhibit 7.10) summarizes this information.
Contingency tables are used to examine the relationships between qualitative or categorical variables by showing the frequency of one variable as a function of another variable. The column of the table in which an observation falls (e.g., either ≤30 minutes or >30 minutes) is contingent on (depends on) the row where the subject is placed (e.g., time of day).
For all patients in the earlier ED wait time example, the probability of waiting longer than 30 minutes is
P(Wait > 30 minutes) = Number of patients who wait > 30
Total number of patients
= 40
100
= 0.40.
Furthermore, the (conditional) probability of waiting more than 30 minutes given that the time frame is Friday night is
=
=
=
PP
P(A | Friday night)
(Wait > 30 minutes and Friday night)(Friday night)
Number of patients who wait > 30 minutes on a Friday nightNumber of patients on a Friday night
30500.60.
A tree diagram for this example is shown in exhibit 7.11.Note that P(A ∩ B) = P(A | B) × P(B) = P(B | A) × P(A), and if one
event has no effect on the other event—that is, the events are independent—then P(A | B) = P(A) and P(A ∩ B) = P(A) × P(B). In the coin and die toss example, the coin toss and die toss are independent events, so the probability of tossing a six is the same no matter the outcome of the coin toss. For the ED wait time example, if night and wait time are independent events, then the probability
Contingency tableA tool used to examine the relationships between qualitative or categorical variables.
≤30-Minute Wait >30-Minute Wait
Friday night 20 30 50
Other times 40 10 50
Total 60 40 100
EXHIBIT 7.10Contingency Table for Emergency Department Wait Times
Healthcare Operat ions Management184
of waiting less than 30 minutes on a Friday night is 0.5 × 0.6 = 0.30. But this contingency is not present; wait time and night are not independent—rather, they are related. From this simple study, one could not conclude that Friday night causes wait time.
Bayes’ theorem allows the use of new information to update the con-ditional probability of an event. It is expressed mathematically as follows:
| =∩
=| ×
=| ×
| × + | ′ × ′P
PP
P PP
P PP P P P
(A B)(A B)
(B)(B A) (A)
(B)(B A) (A)
(B A) (A) (B A ) (A ).
Bayes’ theorem is often used to evaluate the probability of a false-positive test result. If a test for a particular disease is performed on a patient, the pos-sibility exists that the test will return a positive result even if the patient does not have the disease. Bayes’ theorem allows the determination of the prob-ability that a person who tests positive for a disease actually has the disease. For example, if a tested patient has the disease, the test reports that finding with 99 percent accuracy, and if the patient does not have the disease, the test reports that finding with 95 percent accuracy. Now suppose that the incidence of the disease is rare—only 0.1 percent of the population has the disease. The following equation expresses the scenario mathematically:
P
P
No disease | Test positive
Test positiv
( ) =
ee | No disease (No disease)Test positi
( ) × PP vve | No disease No disease Test posi( ) × ( ) +P P ttive | Disease Disease
0.050 0.9990.
( ) × ( )×
P
0050 0.999 0.990 0.001× + ×= 0 981..
A tree diagram (exhibit 7.12) helps to illustrate this problem.As demonstrated by application of the above equation to exhibit 7.12,
the test results are positive 0.00099 + 0.04995 = 0.05094 of the time; 0.04995 of that time, the person does not have the disease. Therefore, the probability
Bayes’ theoremA formula used to revise the calculation of conditional probability as new information is obtained in the situation.
Night Wait Time ProbabilityConditionalProbability
0.2 0.3/0.5 = 0.6Friday
0.3Start
0.4Other
≤ 30 minutes
≤ 30 minutes
^
30 minutes
^
30 minutes 0.1
EXHIBIT 7.11Tree Diagram—
Emergency Department
Wait Time
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 185
that a person does not have the disease, although the test for the disease was positive, is
0.04995
0.05094 = 0.981, or 98.1 percent.
Conditional probability and Bayes’ theorem are often used in healthcare in clinical studies to test drug interactions. In addition, conditional probability is useful in predicting outcomes on the basis of demographics.
Confidence Intervals and Hypothesis TestingCentral Limit TheoremThe central limit theorem states that as the sample size from a population becomes sufficiently large, the sampling distribution of the mean approaches normality, no matter the distribution of the original variable. Additionally, the mean of the sampling distribution is equal to the mean of the population and the standard deviation of the sampling distribution of the mean approaches σ/ n , where σ is the standard deviation of the population and n is the sample size. If a sample is taken from any distribution, the mean of the sample will follow a normal distribution with mean = µ and standard deviation σ/ n , commonly called the standard error of the mean (sx or SE). This theorem is extremely valuable because data that follow the normal distribution have parameters that are easier to understand than those of data with non-normal distribution.
The central limit theorem can be used to determine a confidence inter-val (CI) for the true mean of the population. If the standard deviation of the population is known, the CI for the mean is
x za/2 x μ x + za/2 x
x za/2 nμ x + za/2 n
,
Central limit theoremA theory demonstrating that as the sample size from a population becomes sufficiently large, the sampling distribution of the means approaches normality, no matter the distribution of the original variable.
Confidence interval (CI)The probability that a population parameter falls between two values.
Patient Test Result Probability
0.00099
Has disease0.001 0.00001
Start0.94905
No disease0.999 0.04995
Positive0.990
Negative0.010
Negative0.950
Positive0.050
EXHIBIT 7.12Tree Diagram—Bayes’ Theorem Example
Healthcare Operat ions Management186
where za/2 is the z-value associated with an upper- or lower-tail probability of α. In other words, to obtain a 95 percent CI, the upper- and lower-tail probabilities must be 0.025 (2.5 percent in the upper tail and 2.5 percent in the lower tail, leaving 95 percent in the middle) and the associated z-value is 1.95 (2 is commonly used). Note that increasing the sample size tightens the confidence limits.
If the population standard deviation (σ) is unknown, the sample standard deviation (s) is used to estimate the standard error of the mean:
x za/2 x μ x + za/2 x
x za/2snμ x + za/2
sn
.
Small samples (generally, n < 30, where n is the sample size) do not fol-low a z-distribution; they follow a t-distribution. The t-distribution has greater probability in the tails of the distribution than a z-distribution has and varies according to the degrees of freedom, n – 1. Therefore, for small samples, the following equation is used:
x −ta/2×sn≤µ≤ x + ta/2×
sn
.
Returning to our ED wait time example, if the waiting time for a random sample of 16 patients were measured and their mean wait time found to be 10 minutes with a standard deviation of 2 minutes, a 95 percent CI for the true value of wait time would be
x ts
nx t
s
n− × ≤ ≤ + ×
− × ≤ ≤ + ×
αα μ
µ
//
. .
22
10 2 132
1610 2 13
22
1610 1 06 10 1 06
8 94 11 06.
− ≤ ≤ +
≤ ≤
. .
. .
μ
μ
Because we computed a 95 percent CI, in 19 out of 20 times, if a similar sample were taken, the CI obtained would include the true value of the mean wait time. To an analyst, this result indicates that under similar situations, the expectation is that the mean value falls between 8.94 and 11.06.
If a larger sample of 49 patients had been taken and their mean wait time were found to be 10 minutes with a standard deviation of 1 minute, a 95 percent CI for the true value of the mean would be
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 187
x−za/2×sn≤µ≤ x + za/2×
sn
10−2×149≤µ≤10+2×
149
10−0.3≤µ≤10+0.39.7≤µ≤10.3.
Hypothesis TestingIn the previous section, we demonstrated that a range of likely values for the population parameter of interest can be obtained by computing a CI. This interval may be used to determine whether claims about the value are correct by confirming or disproving that the CI captured the claimed value. In the wait time example, if an observer claimed that the wait time for most patients was eight minutes, the claim would be rejected on the basis of the information obtained. However, if the claim were made that the mean wait time was ten minutes, the study would support this claim. Hypothesis testing is a formal way of testing such claims and is closely related to CIs.
Hypothesis testing includes three components: a belief, called the null hypothesis; a competing belief, called the alternative hypothesis; and a decision rule for evaluating the beliefs.
In the wait time example, these components are expressed as follows:
Ho (belief): µ = 8 minutesHa (alternative belief): µ ≠ 8 minutesDecision rule: If t ≥ t*, reject the null hypothesis
Here, t = (x – µ)/σx, the number of standard errors away from the mean, and t* is the test statistic based on the desired confidence level and the degrees of freedom. If t is greater than t*, finding a sample mean that is dif-ferent from the true value of the mean is unlikely; therefore, the belief about the true value of the mean (Ho) would be rejected. In the wait time example, t* for a 95 percent CI with 15 degrees of freedom (sample size of 16) is 2.13. Therefore, t = (x – µ)/σx = (10 – 8)/0.5, and t ≥ t*. Under this condition, Ha would be rejected.
More typically, hypothesis testing is used to determine whether an effect exists. Suppose a pharmaceutical company wants to evaluate a new headache remedy by administering it to a collection of subjects. If the new headache remedy appears to relieve headaches, the company must be able to state with confidence that the effect was in fact due to the new remedy, not just chance. Most headaches eventually go away on their own, and some headaches (or
Hypothesis testingThe process of testing a statistical distribution parameter against that of another distribution parameter to assess if statistical differences exist in the data.
Healthcare Operat ions Management188
some people’s headaches) are difficult to relieve, so the company can make two kinds of mistakes: incorrectly concluding that the remedy works when in fact it does not, and failing to notice that an effective remedy works. The null hypothesis (Ho) is that the remedy does not relieve headaches; the alternative hypothesis (Ha) is that it does.
Type I and Type II ErrorsA type I, or α, error occurs if the company concludes that the remedy works when in fact it does not. A type II, or β, error occurs if the remedy is effective but the company concludes that it is not.
Hypothesis testing is similar to the process of determining guilt in the US criminal court system. In a trial, the assumption is that the defendant is innocent (the null hypothesis) until proven guilty (the alternative hypothesis); evidence is presented (data), and the jury decides whether the defendant is guilty or not guilty on the basis of proof (decision rule) that must convince the jury beyond a reasonable doubt (confidence level). A jury does not declare the defendant innocent, but rather not guilty.
If the defendant is in fact innocent but the jury decides that he is guilty, then it has sent an innocent person to jail (type I error). If a defendant is guilty but the jury finds him not guilty, a criminal is set free (type II error). In the US criminal court system, a type I error is considered more important than a type II error, so a type I error is protected against, to the detriment of a type II error. This assessment is analogous to hypothesis testing (Kenney 1988), as illustrated in exhibit 7.13.
Usually, the null hypothesis is that something is not present: that a treatment has no effect or that no difference exists between the effects of different treatments. The alternative hypothesis is that some effect is present: that a treatment has an effect or that a difference does exist in the effects of different treatments. Assuming the null hypothesis is true allows one to com-pute the probability that the test rejects the null hypothesis, given that it is true (type I error).
The decision rule is founded on the probability of obtaining a sample mean (or another statistic) given the hypothesized mean (or another statistic).
A comparison of waiting time at two different clinics, on two different days, or during two different periods would apply the following hypothesis test:
Type I (α) errorThe probability of rejecting the null hypothesis when it is true.
Type II (β) errorThe probability of accepting the null hypothesis when it is false.
Reality
Assessment or guess Innocent Guilty
Innocent — Type II error
Guilty Type I error —
EXHIBIT 7.13Type I and Type II Error—Court
System Example
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 189
Ho: µ1 – µ2
Ha: µ1 ≠ µ2
Decision rule: If t ≥ t*, reject Ho
(Note that t* is usually determined with statistical software using the Satherwaite approximation, because the two-sample test statistic does not exactly follow a t-distribution.) Exhibit 7.14 illustrates the errors associated with this example.
Equal Variance t-TestIf the wait time at two different clinics were of interest, wait time for a ran-dom sample of patients from each clinic might be measured. If wait time for a sample of 10 patients (simplified for explanatory purposes) from each clinic were measured and it was determined that clinic A had a mean wait time of 12 minutes, clinic B had a mean wait time of 10 minutes, and both had a standard deviation of 1.5 minutes, the standard deviations could be pooled and the distribution would follow a t-distribution with (n1 + n2 – 2) degrees of freedom. At a 95 percent confidence level,
t =x1−x2( )− µ1−µ2( )
s p1n1
+1n2
,
where
sn s n s
n n
t
1 12
1.5
12 10 0
1.51
101
10
20.67
2.99 * 2.10.
p1 1
22 2
2
1 2
( ) ( )=
− + −+ −
=
− −
+= = ≥ =
Therefore, this test would reject Ho, the belief that the mean wait time at the two clinics is the same.
Reality
Assessment or Guess
Wait times at the two clinics are the same ( µ
1 = µ
2)
Wait times at the two clinics are not the same ( µ
1 ≠ µ
2)
Wait times at the two clinics are the same ( µ
1 = µ
2)
— Type II error
Wait times at the two clinics are not the same ( µ
1 ≠ µ
2)
Type I error —
EXHIBIT 7.14Type I and Type II Error—Clinic Wait Time Example
Healthcare Operat ions Management190
Alternatively, a 95 percent CI for the difference in the two means could be found:
(x1−x2)± t*× s p1n1
+1n2
2−(2.10×0.67)≤µ1−µ2 ≤ 2+ (2.10×0.67)0.6≤µ1−µ2 ≤3.4.
Because the interval does not contain zero, the wait time for the two clinics is not the same.
Statistical software provides the p-value of this test. The p-value of a statistical significance test represents the probability of obtaining values of the test statistic that are equal to or greater than the observed test statistic. For the wait time example, the p-value is 0.015, meaning that Ho would be rejected with a confidence level of up to 98.5 percent, or that zero would not be con-tained in a 98.5 percent CI for the mean. Smaller p-values cause rejection of the null hypothesis.
Another type of test is the t-test, which can be used to examine the mean difference between paired samples; it is performed when the standard deviations of the means differ. See the companion website for more information on these types of t-tests.
ProportionsConsider an example in which staffing levels at two clinics are compared. In clinic A, the ratio of nurses to total staff is 12 nurses of 20 staff, and in clinic B, the ratio is 10 of 20. To determine if these proportions are different, we use the following test:
Ho: π1 – π2Ha: π1 ≠ π2Decision rule: If z ≥ z*, reject Ho
The proportion of nurses at clinic A is 12/20 = 0.60, and the proportion of nurses at clinic B is 10/20 = 0.50. The standard error of the difference in sample proportions is
( ) ( )−+
−p pn
p pn
1 1,
1 2
where
( ) ( )=
++
=+
=pn p n p
n n20 0.6 20 0.5
400.55.1 1 2 2
1 2
On the web at ache.org/books/OpsManagement3
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 191
At a 95 percent confidence level,
zp p
p pn
p pn
=− − −
−+
−=
−( ) ( )
( ) ( )
( . .1 2 1 2
1 2
1 1
0 60 0π π 550 0
0 55 0 4520
0 55 0 4520
0 100
)
( . )( . ) ( . )( . )
.
−
+
=..
. * .157
0 64 1 96.= < =t
Therefore, Ho cannot not be rejected, and no difference can be present in the proportion of nurses at each clinic.
A CI for a proportion can be found from the following:
p z p z pp− × ≤ ≤ + ×αα σ π σ// 22,
wherep p
n(1 )
.p̂σ =−
A 95 percent CI for the difference in the two proportions of nurses is
( )( ) ( )
. – ( . .
p p zp p
np p
n1 21 2
1 1
0 10 1 96 0 157
− ± × − + −
× )) – . ( . . )
. – .
≤ ≤ + ×
− ≤ ≤
π π
π π1 2
1 2
0 10 1 96 0 157
0 2 0 41.
Because the interval contains zero, the proportion of nurses at the two clinics cannot be different. The p-value for this test is 0.53; therefore, Ho is not rejected.
Practical Versus Statistical SignificanceDistinguishing between statistical significance and practical significance is important because some statistical differences have no impact on the business and other differences that are not statistically significant may have tremendous impact. Statistical significance is related to the ability of the test to reject the null hypothesis, whereas practical significance looks at whether the difference is large enough to be of value in a practical sense. If the sample size is large enough, statistical significance can be found for small differences when limited or no practical importance is associated with the finding.
For instance, in the clinic wait time example, if the mean wait time at clinic A were 10.1 minutes, the mean wait time at clinic B were 10.0 minutes, and the standard deviation for both were 1 minute, the difference would not be significant if the sample size at both clinics were 10. If, however, the sample size were 1,000, the difference would be statistically significant. The statistical results from Minitab (a statistical software package) for this example are shown
Statistical significanceThe differences in two parameters of two data sets are large enough to reject the null hypothesis using hypothesis testing.
Practical significanceThe differences in the parameters of two data sets are large enough to be meaningful for the person or organization studying the situation, whether or not they are statistically significant.
Healthcare Operat ions Management192
in exhibit 7.15. Tests for statistical significance should not be applied blindly—the practical significance of a difference of 0.1 minute is a judgment call.
Simple Linear Regression
Regression is a statistical tool used to model the association of a variable with one or more explanatory variables. The variables are typically metric, although categorical variables may also be analyzed using regression. The relationship(s) can be described using an equation.
Simple linear regression is the simplest type of regression. The equation representing the relationship between two variables is Y = βX + α + ε. Many readers will remember Y = mX + b from high school. In statistics, α represents the intercept
Simple linear regressionAn equation that relates two variables using a slope and an intercept in a linear fashion.
Two-Sample t-Test and CISample N Mean SD SEM1 10 10.10 1.00 0.322 10 10.00 1.00 0.32
Difference = μ1 − μ2
Estimate for difference: 0.10000095% CI for difference: (−0.839561, 1.039561)t-Test of difference = 0 (vs. not =): t-value = 0.22p-Value = 0.826, df = 18Both use pooled SD = 1.0000
Two-Sample t-Test and CI Sample N Mean SD SEM1 1,000 10.10 1.00 0.0322 1,000 10.00 1.00 0.032
Difference = μ1 − μ2
Estimate for difference: 0.10000095% CI for difference: (0.012295, 0.187705)t-Test of difference = 0 (vs. not =): t-value = 2.24p-Value = 0.025, df = 1,998Both use pooled SD = 1.0000
EXHIBIT 7.15Statistical
Significance of Differences—
Minitab Output for Clinic Wait Time Example
Note: CI = confidence interval; df = degrees of freedom; SD = standard deviation; SEM = standard error of the mean.
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 193
(the b from high school), β signifies the slope (the m from high school; in statistics m or µ represents the mean, so a different variable name is used), and ε is the error.
A simple example helps illustrate the concept of regression. Assume that the relationship between number of dependents and yearly healthcare expense is of interest and the data in exhibit 7.16 have been collected (for explanatory purposes only, as a larger data set would be needed for a true regression analysis).
First, to visually examine the nature of the relationship between the variables, a scatter plot of the data (exhibit 7.17) is produced. From the scat-ter plot, we see that a linear relationship exists—a line can be drawn that best represents the relationship between the two variables.
The most precise model is one that results in the smallest, or lowest, total absolute error. The best-fitting regression line minimizes sum of squared error.
The estimated lineŶ = 1.3(X) + 2.4 has the lowest squared error term for the data (exhibit 7.18).
InterpretationThe linear model presented in exhibit 7.18 is interpreted as follows. The slope of the line from the previous equation indicates that, with each additional
Number of Dependents Annual Healthcare Expense
0 $3,000
1 $2,000
2 $6,000
3 $7,000
4 $7,000
EXHIBIT 7.16Data for Regres-sion Example: Relationship Between Num-ber of Depen-dents and Yearly Health-care Expense
0
1
2
3
4
5
6
7
8
0 1 2 3 4 5Number of Dependents
Annu
al H
ealt
hcar
e Co
st ($
1,00
0)
EXHIBIT 7.17Scatter Plot—Number of Dependents Versus Annual Healthcare Costs
Healthcare Operat ions Management194
dependent, the annual cost of healthcare rises by $1,300 on average; the annual cost of healthcare for people with no dependents is $2,400, as seen by the location of the intercept. Where X = 0 (no data) and without additional infor-mation, the intercept is not a meaningful number.
Coefficient of Determination and Correlation CoefficientThe next question is, How good is the model? This measure of how well the model fits the data is called the coefficient of determination (r2). Note that this is not a statistical test, but rather a measure of the percentage of error explained by the model. The square root of this number is called the correlation coef-ficient (r). A negative correlation coefficient indicates a negative slope, and a positive correlation coefficient indicates a positive slope. The correlation coef-ficient is a measure of the linear relationship between two variables, with a value of 1 indicating perfect correlation and a value of 0 indicating no relationship. (Refer to exhibit 7.19 for sample scatter plots and their correlation coefficients.)
The coefficient of determination (r2) measures the percentage of variance explained in Y using the X variable (exhibit 7.19). Examining the regression
Coefficient of determinationThe measure of how well a model fits the data.
Correlation coefficientA measure of the linear relationship between two variables.
r = 0.00
(2)
X
Y
X
Y (1)
r = 0.05 r = 0.91
(3)
X
Y
r = 0.56
(6)
X
Y
r = 0.79
(5)
X
Y
r = 0.75
(4)
X
Y
EXHIBIT 7.19Examples of
Low and High r and r2 Plots
X Y Ŷ = 1(X) + 3 e2 Ŷ = 1.3(X) + 3 e2 Ŷ = 0(X) + 5 e2
0 3 3 0 2.4 0.36 5 4
1 2 4 4 3.7 2.89 5 9
2 6 5 1 5.0 1.00 5 1
3 7 6 1 6.3 0.49 5 4
4 7 7 0 7.6 0.36 5 4
∑ 6 5.10 22
EXHIBIT 7.18Sum of
Squared Errors Associated with
the Various Linear Models
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 195
output of the data from exhibit 7.19, shown in exhibit 7.20, r2 = 0.768, which can be interpreted as 77 percent of the variance in annual healthcare expense (Y variable), can be explained by the number of dependents (X variable). The “best” value a model can achieve is 100 percent of the variance explained. So, is 77 percent a “good” value? The answer depends on many factors. A number closer to 100 percent is ideal, but if your sample size is sufficiently large, finding a variable that explains 25 percent of the variance could be helpful.
Problems with Correlation CoefficientsThe coefficient of determination and the correlation coefficient are both mea-sures of the linear relationship between two variables. A scatter plot of the two variables should always be examined when initially evaluating the appropriate-ness of a model. Statistical techniques for judging the appropriateness of the model are discussed later in this chapter.
Does a low r2 mean that no relationship exists between two variables? No. Exhibit 7.21 illustrates two cases (1 and 2) in which r2 and r are both 0. In case 1, no relationship is evident; in case 2, a relationship is seen, just not a linear relationship. The relationship can be perfectly captured with the equation Y = α + β1X + β2X2, a curve or quadratic relationship (curve-type relationships are discussed later in the chapter). A low r2 may also mean that other variables needed to explain the outcome variable are “missing” from the model.
Does a reasonable or high r2 mean the model is a good fit to the data? No. Exhibit 7.21 illustrates several cases in which the model is not a good fit to the data. The r2 and r can be heavily influenced by outliers, as in cases 4 and 6. In case 5, a better model would be a curve. Always look at the scatter plot of the data.
0
1
2
3
4
5
6
7
8
9
0 1
Number of Dependents
Annu
al H
ealt
hcar
e Co
st ($
1,00
0)
10
2 3 4 5
Y = 5Y = X + 3
Y = 1.3X + 2.4
Y = 1.2X + 2
6
EXHIBIT 7.20Scatter Plot with Possible Relationship Lines
Healthcare Operat ions Management196
Does a high r 2 mean that useful predictions will be obtained with the model? No. Recall the previous discussion of practical and statistical signifi-cance. Finally, does a high r 2 mean that a causal relationship exists between the variables? No—correlation is not causation. The observed correlation between two variables might be due to the action of a third, unobserved variable. For example, Yule (1926) found a high positive correlation between yearly suicides and membership in the Church of England. However, membership in the Church of England did not cause suicides.
Statistical Measures of Model FitIf no linear relationship is present between the two variables, the slope of the best-fitting line will be 0. This idea undergirds the statistical tests for the “goodness of fit” of the model.
F-testThe F-test is a hypothesis test of whether all β values in the model Y = α + βX + ε are equal to 0. In the case of simple linear regression, there is only one β, and the test determines whether β is 0.
First, we express the hypotheses and decision rule as follows:
Ho: all β values = 0Ho: all β values ≠ 0
Decision rule: If F* ≥ F(1–α; 1; n–2), reject Ho
Next, we apply the F-test equation:
F* = Mean square regression
Mean square error =
MSR
MSE =
SSR/1
SSE/n – 2,
where MSR is mean square regression, MSE is mean square error, SSR is sum of squares regression, and SSE is sum of squares error. If the two variables are related, the regression line explains most of the variance and SSR is large compared to SSE. Therefore, large values of F* imply a relationship and the slope of the line is not equal to 0.
t-TestFor simple linear regression, the t-test gives the same answer as the F-test. The t-test is a hypothesis test of whether a particular β is 0.
SUMMARY OUTPUT
Regression StatisticsMultiple RR SquareAdjusted R SquareStandard ErrorObservations
0.87650.76820.69090.8790
5
InterceptY—$1000 Annual Health care Expense
Coefficients
–0.9545
0.5909
EXHIBIT 7.21Regression Output for Healthcare
Expense Example
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 197
H0: β = 0Ha: β ≠ 0Decision rule: If t* ≥ F(1–a; 1; n–2), reject Hα,
where t* = b/sb.
Alternatively, a CI for β would be
b – t(1–α; n–2)sb ≤ β ≤ b – t(1–α; n–2)sb.
If the interval contains 0, Ho can be rejected. Statistical software provides these tests for linear regression as well as for r and r 2.
Assumptions of Linear RegressionLinear regression is based on several principal assumptions:
• The dependent and independent variables are linearly related.• The errors associated with the model are not serially correlated.• The errors are normally distributed and have constant variance.
If these assumptions are violated, the resulting model will be misleading.Various plots (and statistical tests) can be used to detect such problems.
These plots are usually provided in the software and should be examined for evidence of violations of the assumptions of regression. A scatter plot of the observed versus predicted value should be symmetrically distributed around a diagonal line, and a scatter plot of residuals versus predicted value should be symmetrically distributed around a horizontal line. A normal probability plot of the residuals should fall closely around a diagonal line.
If evidence is seen that the assumptions of linear regression are being violated, a transformation of the dependent or independent variables may fix the problem. Alternatively, one or two extreme values may be the cause of assumption violations. Such values should be scrutinized closely: Are they genu-ine (i.e., not the result of data entry errors), are they explainable, are similar events likely to occur again in the future, and how influential are they in the model-fitting results? If the values are merely errors, or if they can be explained as unique events not likely to be repeated, removing them may not be neces-sary. In some cases, however, the extreme values in the data may provide the most useful information about values of some coefficients or provide the most realistic guide to the magnitudes of prediction errors.
TransformationsIf the variables are not linearly related or the assumptions of regression are violated, the variables can be transformed to possibly produce a better model. Transformations are applied to ensure that the model is accurate and reliable.
TransformationThe process of converting a variable by linear regression into a format that is more readily usable.
Healthcare Operat ions Management198
If a person were to jog to her doctor’s appointment, she would need to wait before having her blood pressure measured, especially if a high reading would result in a diagnosis of hypertension. Blood pressure values obtained immedi-ately after exercising are unsuitable for detecting hypertension; the reason for waiting is not to avoid the diagnosis of hypertension but to ensure that a high reading can be believed. The concept is similar with transformations.
Deciding which transformation is best is often an exercise in trial and error in which several transformations are tried to see which one provides the best model. Possible transformational functions include square root, square, cube, log, and inverse. Any data that are transformed need to be accounted for when interpreting the findings. For example, imagine that the original variable was measured in days but, to improve the model, an inverse transformation was applied. Here, the lower the value for this transformed variable (1/days), the higher the value of the original variable (days). If the dependent variable is binary (0/1), the assumptions of regression are violated. The logit transforma-tion of the variable, ln[p/(1 – p)], is used in this case.
Conclusion
An outline for analysis is shown in exhibit 7.22, with each item correspond-ing to the steps in the plan-do-check-act process for continuous improvement (chapter 9), the define-measure-analyze-improve-control process of Six Sigma (chapter 9), and the key elements of decision making (chapter 6).
PDCA DMAIC Key Element
1. Define the problem/question. Plan Define Frame
2. Determine what data will be needed to address the problem/question.
Plan Define Frame
3. Collect the data. Do Measure Gather
4. Graph the data. Do Analyze Gather
5. Analyze the data using the appro-priate tool.
Do Analyze Conclude
6. Fix the problem. Do Improve Conclude
7. Evaluate the effectiveness of the solution.
Check Control Learn
8. Start again. Plan Define Frame
Note: DMAIC = define, measure, analyze, improve, and control; PDCA = plan, do, check, and act.
EXHIBIT 7.22Outline for
Analysis
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 199
Which Technique to UseThe statistical tool or technique chosen to analyze the data depends on the type of data collected. The next chapter, on healthcare analytics, discusses techniques to gain insights from data using current technology. In addition, more traditional statistical tests are included in the supplemental section on the book’s companion website.
Discussion Questions
1. Discuss a situation from your personal experience in which a study had bad data. How were the data collected, and what were the reported problems with data collection?
2. John Allen Paulos (whose work can be found at http://abcnews.go. com/Technology/WhosCounting/) and Jordan Ellenberg (at www.slate.com/authors.jordan_ellenberg.html) both write on numbers, statistics, and probability. Read an article of interest to you and discuss.
3. How would you redesign a report you receive at work to make it more useful? Would a visual presentation of the data be helpful? How would you present the data?
4. Discuss the difference between correlation and causation.5. Discuss the difference between statistical significance and practical
significance.6. The balanced scorecard, Six Sigma, Lean, and simulation employ many
of the tools, techniques, and tests found in this chapter. Discuss how, where, and why a particular tool would be used for each approach.
Exercises
The following problems use data from three data sets available on the companion website. Each data set contains the raw data as well as reduced or reor-ganized data for ease of analysis.
1. Think of a question, a problem, or an issue in your organization, and design a study to address it. Be sure to discuss how you would address all aspects of data collection, including how you would collect the data. How could you make sure the data are representative of the actual situation?
2. Using the data in the file labeled HealthInsuranceCoverage.xls, compare insurance coverage in Minnesota to coverage in Texas.
On the web at ache.org/books/OpsManagement3
On the web at ache.org/books/OpsManagement3
Healthcare Operat ions Management200
a. Analyze the validity of the data.b. Produce a histogram to compare the two states. Do Minnesota and
Texas appear to have similar coverage types?c. Produce a Pareto chart for the two states. What does this chart
indicate?d. What is the probability that a resident of Minnesota or Texas will be
uninsured? Insured? Insured by Medicare or Medicaid?e. What is the 95 percent CI for the proportion of uninsured in Texas?
In Minnesota? What is the 99 percent CI?f. What is the 99 percent CI for the difference in the two propor tions?g. Set up and perform a hypothesis test to determine if the proportion
of uninsured differs at a 95 percent confidence level between the two states.
h. Comment on the statement, “Living in Texas causes more people to be uninsured.” What other information might be helpful to either validate or disprove the statement?
3. Use the data in the file labeled WorldHealth.xls to analyze worldwide life expectancy. Answer the following questions:a. Construct a histogram, dot plot, and normal probability plot of the
Central Intelligence Agency’s (CIA) totals for life expectancy at birth (years) for 2006. What do these graphs indicate? Is this random variable normally distributed?
b. Construct a graph of the CIA life expectancy data—total, male, and female. What do these graphs show?
c. Determine the mean, median, mode, range, variance, and standard deviation for the CIA life expectancy data—total, male, and female. What do these numbers show?
d. What is the 95 percent CI for mean life expectancy of males and females as reflected in the CIA data? The 99 percent CI?
e. What is the 99 percent CI for the difference in the two means?f. Set up and perform a hypothesis test to determine if life expectancy
for males and females differs at a 95 percent confidence level.g. Construct histogram and normal probability plots for the CIA
gross domestic product, television, and hospital bed data. Do these random variables appear normally distributed?
h. Perform three separate simple linear regression analyses for the CIA gross domestic product, television, and hospital bed data with the CIA life expectancy total. Interpret your results. (Note: Excel will not perform a regression analysis when data are missing. The Excel workbook titled “World Health Regression” has eliminated
Chapter 7: Stat ist ical Thinking and Stat ist ical Problem Solving 201
countries for which no data are available on life expectancy. You may need to sort the data and run the analysis only on complete data.)
i. Discuss the following statement: “World life expectancy could be increased if everyone in the world owned a television.”
j. Look at x-y scatter plots for each pair of variables in exercise 3h. Do the relationships appear to be linear? Would a transformation of the x variable improve the regression?
References
Botelho, G. 2016. “Zika Virus ‘Spreading Explosively,’ WHO Leader Says.” CNN. Updated February 20. www.cnn.com/2016/01/28/health/zika-virus-global-response/.
Centers for Disease Control and Prevention (CDC). 2016a. “About Zika.” Updated August 1. www.cdc.gov/zika/about/index.html.
———. 2016b. “Health Effects & Risks.” Updated August 9. www.cdc.gov/zika/pregnancy/ question-answers.html.
DeLia, D. 2007. “Hospital Capacity, Patient Flow, and Emergency Department Use in New Jersey.” Rutgers Center for State Health Policy. Published September. www.cshp.rutgers.edu/Downloads/7510.pdf.
Juran, J. M., and J. A. De Feo. 2010. Juran’s Quality Handbook: The Complete Guide to Performance Excellence, 6th edition. New York: McGraw-Hill Education.
Kenney, J. M. 1988. “Hypothesis Testing: Guilty or Innocent.” Quality Progress 21 (1): 55–57.Meltzer, M. I., C. Y. Atkins, S. Santibanez, B. Knust, B. W. Petersen, E. D. Ervin, S. T.
Nichol, I. K. Damon, and M. L. Washington. 2014. “Estimating the Future Number of Cases in the Ebola Epidemic—Liberia and Sierra Leone, 2014–2015.” Centers for Disease Control and Prevention. Published September 26. www.cdc.gov/mmwr/preview/mmwrhtml/su6303a1.htm.
Sanchez, R. 2014. “WHO to Review Ebola Response Amid Criticism of Its Efforts.” CNN. Updated October 19. www.cnn.com/2014/10/18/world/who-ebola-response/index.html.
Truelsen, T., and M. Grønbæk. 1999. “Wine Consumption and Cerebrovascular Disease Mortality in Spain.” Stroke 30 (1): 186–88.
Tufte, E. R. 1997. Visual Explanations: Images and Quantities, Evidence and Narrative. Cheshire, CT: Graphics Press.
———. 1990. Envisioning Information. Cheshire, CT: Graphics Press.———. 1983. The Visual Display of Quantitative Information. Cheshire, CT: Graphics Press.Yule, G. U. 1926. “Why Do We Sometimes Get Nonsense-Correlations Between Time-
Series?—A Study in Sampling and the Nature of Time-Series.” Journal of the Royal Statistical Society 89 (1): 1–63.
CHAPTER
203
HEALTHCARE ANALYTICS
Operations Management in Action
A major clinical research study was one of the earli-est and best uses of clinical big data and advanced analytics. In the mid-1990s, a number of new drugs came onto the market to control hypertension. Early studies showed these new drugs to be effective, but they were much more expensive than the existing therapies.
The National Institutes of Health (NIH) undertook a large study, known as ALLHAT, to evaluate these new pharmaceuticals in terms of their specific level of effectiveness. The study took eight years and cost $120 million to complete. The results, announced in 2002, were surprising: The investigators found that for approxi-mately 40 percent of the population, the older therapies worked as well as the new drugs. However, for the remaining 60 percent, the newer drugs appeared to be superior. The next important question was how to determine which of the newer drugs worked best for which patients. The NIH had neither the funding nor the time to complete this next study.
However, Kaiser Permanente researchers had been following the study and decided to embark on their own version of the phase II study using the organization’s electronic patient records. Using better data from the electronic health records (EHRs) than were available to the NIH, the researchers were able to match their patients to the most effective drugs within 18 months—at a cost of $200,000. The analytics gathered from the EHR allowed the researchers to finish the study at much lower cost and more quickly than the larger NIH study could have.
Source: Begley (2011).
What Is Analytics in Healthcare?
In 2007, Thomas Davenport and Jeanne Harris wrote their seminal book, Competing on Analytics: The New Science of Winning. This text demonstrates how companies from many different industries can use analytics to create value and improve organizational performance.
8OVE RVI EW
“Too much data and not enough information” has never
resonated more than in today’s healthcare environ-
ment. In response, the disciplines of analytics, big data,
and informatics have exploded and even become com-
monplace in hospital and health system operations.
Healthcare Operat ions Management204
Analytics, defined by one source as the “the systematic computational analysis of data or statistics” (Oxford Living Dictionaries 2016), has become particularly popular across the healthcare landscape for a number of reasons:
• More data than ever before are generated and available—particularly with the wide adoption of EHRs.
• The current regulatory environment requires the reporting of thousands of measures.
• Hospitals and health systems are facing increased pressure to improve clinical, operational, and financial results.
• Population health has become a competitive strategy, and analytics is crucial to shaping effective population health initiatives.
• Information technology and software are increasingly sophisticated, allowing analysis of data on a massive scale.
More DataDue to ever-increasing computing power and the advent of cloud storage, smartphones, and other technologies, more data and information are available today than ever before. This availability presents both challenges and oppor-tunities in data storage, security, and management. In large part because of the availability of funds—and new mandates—from the American Recovery and Reinvestment Act of 2009, most hospitals and clinics have installed EHR systems. The massive conversion from paper charts and records was difficult for many organizations to accomplish, but EHRs are finally stable enough to be used as a good data resource. Epic Systems Corporation and Cerner Cor-poration are the two largest software companies to have created platforms to store health records. The widespread use of these systems has given healthcare providers the capability to longitudinally collect data on patients, which offer healthcare systems comprehensive insights and, potentially, the capability to improve care decision making.
Regulatory EnvironmentThe effective use of analytics can help healthcare organizations manage mounting regulatory pressures. The Centers for Medicare & Medicaid Services requires every hospital to report approximately 1,700 quality measures for regulatory compliance (Blumenthal, Malphrus, and McGinnis 2015). The sheer number of data points that must be collected forces organizations to dedicate significant resources to collecting and managing the data. And this effort does not take into account the additional resources required to analyze and make decisions with the data.
Pressure to Produce ResultsIn Minnesota, a unique relationship was formed between Allina Health and Health Catalyst. Health Catalyst provides analytics services to Allina to assist
Chapter 8: Healthcare Analyt ics 205
in project management, continuous improvement, population health analysis, and financial analytics. The use of large-scale data allows the healthcare system to focus on achieving results through coordinated efforts. Data have been used to analyze a variety of system elements, including doctors’ efficiency, clinic efficiency, and overall system effectiveness.
Healthcare regulation and competition pressure have changed the mar-ketplace for health systems. Organizations need a systematic approach to reduc-ing costs and finding new market opportunities. The operations improvement tools discussed in many of the chapters of this book can be made more powerful with the use of advanced analytics.
Population HealthKindig and Stoddart (2003) define population health as “the health outcomes of a group of individuals, including the distribution of such outcomes within the group.” The increased use of EHRs gives health systems the ability to understand the costs and clinical trends related to the patients they serve. This capability allows the development of specific treatments for diseases and conditions, which leads to improved outcomes. One of the hallmarks of big data analysis in healthcare is the use of predictive models.
Winters-Miner (2014) identifies seven ways predictive analytics can improve healthcare:
• Improves diagnosis• Helps with preventive medicine and public health efforts• Provides answers to physicians for the treatment of individual patients• Provides employers and hospitals tools to predict insurance product costs• Allows smaller test cases to be used to prove models• Helps pharmaceutical companies meet the needs of the public for
medication• Potentially helps improve outcomes
Sophisticated TechnologyTechnology breakthroughs are enabling analysts to tackle increasingly complex problems. Analytics technology not only allows larger data sets to be used but also increases the speed in which analysis can be completed.
Introduction to Data Analytics
The goal of data (big and small) analytics is to obtain actionable insights that result in smarter decisions and better business outcomes. Many of the tools and statistical techniques from chapters 6 and 7 can be applied in an analytics environment.
Healthcare Operat ions Management206
The basic work of an analyst is to build a data framework through the following major goals:
• Gathering data—Data are the facts provided by databases.• Building information—Information is the layer on top of data that
helps make sense of the data. Without essential knowledge of the business situation, the information is likely not valuable.
• Gaining actionable insights—Actionable insights are those nuggets of knowledge from the information that affect the organization. The insights should enhance a leader’s ability to make improved decisions.
This framework for data analysis provides the background for the various forms of analytics.
Analytics can be described as taking place in three distinct phases:
• Descriptive analytics• Predictive analytics• Prescriptive analytics
Descriptive AnalyticsDescriptive analytics is the process of condensing large data sets into meaningful information that can assist in decision making. Descriptive statistics examine past performance and summarize data to discern trends and patterns to explain behavior. In healthcare, reporting mechanisms such as regulatory compliance, quality measures, and financial results commonly use descriptive analytics.
Descriptive analytics makes up the largest subset of the analytics field. One main feature of data visualization is making data consumable by people. The process of converting raw data is necessary because data alone are not typically usable to managers.
Examples of descriptive analytics outputs include the following:
• Business intelligence reports • Dashboards with key performance indicators (KPIs)• Descriptive statistics• Traditional data visualization techniques
Predictive AnalyticsPredictive analytics builds models on the basis of data that can help forecast the future in terms of probabilities. Models cannot perfectly predict the future but can provide insights for individuals to make effective decisions. Predictive analytics uses a variety of statistical techniques ranging from regression modeling to machine learning to data mining to make projections about future events.
Business intelligenceThe process of converting raw data through a variety of methods into information that can assist with decision making.
Chapter 8: Healthcare Analyt ics 207
In healthcare, the use of predictive models has become popular in disease management and population health. For example, some healthcare organiza-tions have begun to examine early indicators of diabetes to help prevent and lower costs associated with diabetes management (Barton 2016). This analytics activity is important as, according the Centers for Disease Control and Preven-tion (CDC 2009), more than 75 percent of total healthcare spending in the United States is related to chronic healthcare conditions.
At Hennepin County Medical Center (HCMC), the population health analysts discovered that individuals diagnosed with HIV also suffered from poor nutrition. A predictive model was constructed showing the positive impact of improved nutrition on healthcare costs. Today, HCMC distributes healthy food with HIV medications for many of the patients in this population and have found overall costs to be reduced.
In short, predictive models have become a common approach to help reduce overall costs, improve quality outcomes, and lower overall patient risk.
Predictive ToolsThree approaches are typically used for developing predictive models: regres-sions, decision trees, and neural networks.
RegressionsChapters 7 and 13 describe a number of regression-type approaches that can be used to predict future performance from historical data. Most analytical software (e.g., SAS, SPSS) packages include numerous regression tools.
Decision Trees Decision trees are a form of “supervised learning” tools. The decision tree algorithm first suggests a split of the databases into a series of “leaves,” whereby each data point is allocated to one leaf. If the analyst agrees with the computer’s selection of leaves, the computer then suggests a further subdivision of the leaves. This process continues until the analyst believes the full tree represents a good model of the data.
Although regressions may be more accurate in their predictive capability, decision trees are useful for explaining the predictions to nonanalysts. A ver-sion of the decision tree tool was used to create the Medicare diagnosis-related group (DRG) system in 1983. Exhibit 8.1 demonstrates the use of a decision tree to predict annual costs for Medicare patients.
Neural Networks Neural networks attempt to mimic the human brain in the following ways:
• Input units obtain the values of input variables and, if the analyst chooses, standardize those values.
Healthcare Operat ions Management208
Not
e: D
iagr
am c
reat
ed w
ith
SAS
Min
er. D
ata
are
sim
ulat
ed M
edic
are
diag
nosi
s an
d an
nual
inpa
tien
t pay
men
ts.
EXH
IBIT
8.1
Dec
isio
n Tr
ee E
xam
ple
Chapter 8: Healthcare Analyt ics 209
• Hidden units perform internal computations, providing the nonlinearity that makes neural networks powerful.
• Output units compute predicted values and compare those predicted values with the values of the target variables.
Units pass information to other units through connections. Connections are directional and indicate the flow of computation in the network.
Once a neural network is created, it can be applied to predict outputs on the basis of new inputs. A challenge in using neural networks is that they are sensitive to the initial data used to calibrate the network. In addition, because of the hidden computations, neural networks are difficult to diagnose and correct if they are not operating properly.
Prescriptive AnalyticsPrescriptive analytics provides decision makers with models that offer guidance in the form of recommendations. These models use a combination of predictive models, optimization, mathematical models, and other techniques to generate prescriptive solutions. Examples of prescriptive models include the following:
• Models for staffing that maximize quality outcomes and minimize costs• Models to maximize capacity in operating rooms• Strategic models that demonstrate efficient allocation of capital
investments• Risk models that minimize adverse health events
Healthcare problems are complex and multidimensional and can be difficult to model. In the modeling process, many assumptions are made in prescriptive models such as optimization. Decision makers can use prescrip-tive models in combination with their knowledge of the healthcare system to make effective decisions.
Data Visualization
Data visualization tools help decision makers extract value from raw big data. They enable users to quickly view, and make sense of, large amounts of data and to combine several data sources.
When dealing with most real-world data sets, the analyst can expect to spend up to 80 percent of her time finding, acquiring, loading, cleaning, and transforming data. Some of this process can be performed with automated tools, but almost any data cleaning involving two or more data sets requires some level of manual work.
Healthcare Operat ions Management210
Many forms of data visualization have been developed. Those discussed in this section include traditional charts and graphs and dashboards. These examples represent just a few of the common forms of visualization used today in hospitals and health systems.
Traditional Charts and GraphsBar GraphsBar graphs, or column graphs, help users visualize the scale of differences between categories. Exhibit 8.2 is a bar graph showing how much a hospital system is spending on purchasing by vendor type. This is a classic example of a traditional business intelligence report created in Microsoft Excel.
Line GraphsAnother traditional business intelligence report is a classic line graph. Line graphs are useful in examining data over time. Exhibit 8.3 is a line graph show-ing the number of cases of biological agents reported to the CDC from 1957 to 2012. The peak in the early 2000s represents the anthrax cases reported in the time frame following the 9/11 terrorist attacks on New York City and Washington, D.C., in 2001. As the exhibit demonstrates, line graphs reveal opportunities to explore trends and peaks in activity.
Map FunctionalityExhibit 8.4 is an example of a map of diabetes concentration by county in the United States created in Tableau. Tableau is a powerful data analysis and visu-alization software that allows a user to create pictures by inputting data. While
Dual Source Multi Source Sole SourceTotal 99863523.02 298821859.6 183731163.3
$-
$50,000,000
$100,000,000
$150,000,000
$200,000,000
$250,000,000
$300,000,000
$350,000,000
Vendor Arrangement
EXHIBIT 8.2Bar Graph
Showing Total Allocation by Vendor Type
Chapter 8: Healthcare Analyt ics 211
such mapping does not have any predictive capability, exhibit 8.4 demonstrates its effectiveness in showing, for example, where the highest concentrations of reported diabetes patients reside. These types of maps help decision makers understand the concentration of data in geographic locations.
19570
10
20
Num
ber
Year
Biologic terrorism-related cases
30
40
1962 1967 1972 1977 1982 1987 1992 1997 2002 2007 2012
EXHIBIT 8.3Line Graph Showing Number of Biological Agent Cases Reported, 1957–2012
Source: Adams et al. (2014).
EXHIBIT 8.4Interactive Map of Diabetes Prevalence by US County, 2004–2012
Source: Cook (2015). Used with permission.
Note: Higher numbers and darker grayscale indicate an increase in diabetes prevalence.
Healthcare Operat ions Management212
Histograms and Scatter PlotsScatter plots show the relationships between two variables, and histograms are graphical representations of the distribution of data. These visualization techniques are covered in more detail later in the chapter.
Dashboards1
A key purpose of an analytics department is to collect and define metrics and KPIs for executive and operational dashboards. While the techniques discussed here can be used across many different business intelligence gathering efforts, they are also useful for collecting and organizing business data into a format for effective dashboard design.
With the explosion of dashboard tools and technologies in the busi-ness intelligence market, many people have different understandings of what a dashboard, metric, and KPI consist of. In an effort to create a common vocabulary, we define a set of terms that form the basis of our discussion. Although the definitions provided in the following subsection might seem onerous and require a second reading to fully understand them, once grasped, these concepts avail you of a powerful set of tools for creating dashboards with effective and meaningful metrics and KPIs.
Metrics and Key Performance IndicatorsMetrics and KPIs are the building blocks of many dashboard visualizations, as these components are the most effective means of alerting users to their progress toward achieving their objectives. In addition to being the products of an organization’s goals and objectives, metrics and KPIs may arise from strategy maps (discussed in chapter 4).
The definitions that follow build from one concept to the next and help inform dashboard design. Take the time to understand each definition and the related concepts before moving on to the next definition.
MetricsThe term metric refers to a direct numerical measure that represents a piece of business data in relationship with one or more dimensions. One example is gross sales by week. The measure is dollars (gross sales), and the dimension is time (week). For any given measure, viewing the values across different hier-archies in a dimension may be helpful. For instance, a display of gross sales by day, week, and month shows the dollars (gross sales) measure along different hierarchies (day, week, and month) in the time dimension. The term grain refers to the association of a measure with a specific hierarchical level in a dimension.
Looking at a measure across more than one dimension, such as gross sales by territory and time, is called multidimensional analysis. Most dashboards do not leverage multidimensional analysis except in a limited and static way; more dynamic “slice and dice” tools are available in the business intelligence
Chapter 8: Healthcare Analyt ics 213
market. This qualification is important to note. Say you uncover a significant need for this type of analysis in the requirements gathering process. Know-ing that these robust tools exist, you have the option of supplementing your dashboards with some type of multidimensional analysis tool.
Key Performance IndicatorsA KPI is simply a metric that is tied to a target. Most often, a KPI represents the distance a metric is above or below a predetermined target. KPIs usually are shown as a ratio of actual to target and are designed to instantly let a busi-ness user know if he is on or off track without having to consciously focus on the metrics represented. For instance, an organization may decide that, to hit the quarterly sales target, it needs to sell $10,000 worth of syringes per week. The metric is syringe sales per week, and the target is $10,000. Using a percentage gauge visualization to represent this KPI, and assuming we had sold $8,000 in syringes by Wednesday, the user would instantly see that he is at 80 percent of the goal.
When selecting targets for KPIs, remember that a target is needed for each grain you want to view in a metric. Having a dashboard that displays a KPI for gross sales by day, week, and month, for example, requires that targets be identified for each associated grain.
Scorecards, Dashboards, and ReportsThe difference between a scorecard, a dashboard, and a report can be one of fine distinctions. Each of these tools can combine elements of the other, but at a high level they all target distinct and separate levels of the business decision-making process.
ScorecardsStarting at the highest, most strategic level of the business decision-making spectrum are scorecards. Scorecards are primarily used to help align operational execution with business strategy. The goal of a scorecard is to keep the business focused on a common strategic plan by monitoring real-world execution and mapping the results of that execution back to a specific strategy (see chapter 4). The primary measurement used in a scorecard is the KPI. These indicators are often a composite of several metrics or other KPIs that measure the orga-nization’s ability to execute a strategic objective. One example of a scorecard KPI is profitable sales growth, which combines several weighted measures, such as new customer acquisition, sales volume, and gross profitability, into one final score.
DashboardsA dashboard resides one level down from a scorecard in the business decision-making process, as it is less focused on a strategic objective and more tied to
Healthcare Operat ions Management214
operational goals. An operational goal may directly contribute to one or more high-level strategic objectives. In a dashboard, execution of the operational goal itself becomes the focus, not the high-level strategy.
The purpose of a dashboard is to provide the user with actionable busi-ness information in a format that is both intuitive and insightful. Dashboards leverage operational data primarily in the form of metrics and KPIs.
ReportsProbably the most prevalent business intelligence tool seen in business today is the traditional report. Reports can be simple and static in nature, such as a list of sales transactions for a given time period, or more sophisticated cross-tab reports with nested groupings, rolling summaries, and dynamic drill-through or linking. Reports are most appropriate when the user needs to look at raw data in an easy-to-read format.
When combined with scorecards and dashboards, reports allow users to analyze the specific data underlying their metrics and KPIs.
Gathering Key Performance Indicator and Metric Requirements for a DashboardTraditional business intelligence projects often take a bottom-up approach in determining requirements, where the focus is on the domain of data and the relationships that exist in those data. When collecting metrics and KPIs for your dashboard project, however, taking a top-down approach is preferred. A top-down approach starts with the business decisions that must be made first and then works down into the data needed to support those decisions. To take a top-down approach, you must involve the business users who will be utilizing these dashboards, as these are the only people who can determine the relevancy of specific business data to their decision-making process.
Data Mining for Discovery
The vast majority of work being performed in healthcare analytics today is in reporting (descriptive analytics), with some highly specialized work in both predictive and prescriptive analytics. Almost all of these tasks share a common characteristic: They entertain a specific hypothesis. Examples are as follows:
• I believe that patients of some doctors experience significantly longer lengths of stay than those of other doctors for the same DRG.
• I believe I can predict the amount of time a health plan will take to remit payment.
• I believe I can predict which patients will not fill their prescriptions on the basis of their zip code.
Chapter 8: Healthcare Analyt ics 215
However, another powerful approach—data mining—is being used in industries outside of healthcare. In this approach, data are explored without a specific hypothesis being established, relying only on a general sense that the data might reveal insights. Data mining is a subfield of computer science that uses algorithms to discover patterns of data interactions in large data sets. It uses artificial intelligence machine learning, classical statistics, and advanced database systems such as Hadoop. Examples of data mining tools are cluster-ing and text mining. Cognitive computing tools such as IBM’s Watson also support data mining.
ClusteringClustering places objects into groups, or clusters, suggested by the nature of the data. The objects in each cluster tend to be similar to each other in some sense, and objects in different clusters tend to be dissimilar. If obvious clusters or groupings are developed prior to the analysis, the clustering analysis can be performed by simply sorting the data.
The clustering methods perform disjoint cluster analysis on the basis of Euclidean distances computed from one or more quantitative variables and seeds that are generated and updated by the algorithm. The user can specify the clustering criterion used to measure the distance between data observations and seeds. The observations are divided into clusters so that every observation belongs to at most one cluster.
After clustering is performed, the characteristics of the clusters can be examined graphically using a clustering package in software such R or SAS statistical packages. Exhibit 8.5 is a cluster analysis of the same Medicare data used for the decision tree in exhibit 8.1. Note that beneficiaries with chronic conditions cluster together because of their high use of inpatient services.
Text MiningEHRs contain a significant amount of text, such as doctors’ and nurses’ notes. Therefore, a useful subset of data mining tools for healthcare providers is text miners. The case study that follows demonstrates the applicability of text min-ing to public health initiatives.
Case Example: Text Mining at the State FairThe authors undertook an engagement in 2015 to assist a local nonprofit, Health Fair 11, an annual event sponsored by a local television station in Minneapolis–St. Paul in conjunction with the Minnesota State Fair (for more information, visit www.kare11.com/news/health/healthfair-11/a-healthy-minnesota-state-fair-tradition/296307902). The initiative provides fairgoers with access to medical workers who check pulses, blood pressures, glucose levels, weight, and eyes and ears for potential health problems. Flu shots are also available, and advocacy groups are on hand to share health information
Healthcare Operat ions Management216
on topics from gluten-free diets to stroke prevention to memory loss. Vendor groups include nonprofit organizations, professional associations, and for-profit companies.
The managers of Health Fair 11 were interested to know if their opera-tional strategy needed adjustment. We surveyed a sample of 351 participants over six days. One of the key questions we asked was, “Why did you choose to get health screening at the state fair?” Our general hypothesis was that the reason fairgoers used the Health Fair 11 screening services was either low cost or convenience. In addition to the results from these two options on our data collection form, we collected text answers (comments written freehand on the form).
We then used SAS text miner Topic tools to cluster the text responses. Exhibit 8.6 is the clustered response. Much to our surprise, the word fun appeared frequently. This unexpected result allowed us to pursue this concept with the organization and its vendors. We came to understand that the fair-goers felt empowered and engaged in this screening, as they were in control and did not have to go through the many gatekeepers of the traditional health system. This finding has proved useful to Health Fair 11 and carries important implications for primary care and population health.
Plan_CVRG_TOT_NUM
CLUS5
CLUS3
SP_STRKETIA
SP_CNCR CLUS8
SP_DEPRESSN
SP_CHRNKIDN
SP_RA_OA SP_ISCHMCHTSP_CHF
SP_OSTEOPRS
SP_ALZHDMTA
SP_DIABETESSP_COPD
CLUS1
CLUS2
CLUS10
CLUS9
CLUS6
MEDREIMB_CAR
MEDREIMB_OP
BENRES_IP
PPPYMT_IP
SP_STATE_CODE
CLUS4 BENRES_CARBENE_DEATH_DT
BENE_COUNTY_CD
BENE_RACE_CD
BENRES_OP
PPPYMT_CAR
PPPYMT_OP
BENE_HI_CVRAGE_TOT_MONS
BENE_SMI_CVRAGE_TOT_MONS
BENE_BIRTH_DT
BENE_SEX_IDENT_CD
CLUS7
HMO_CVRAGE_TOT_MONS
Note: Data used in this exhibit are the same as those used for the decision tree in exhibit 8.1.
EXHIBIT 8.5Cluster Analysis
of Sample Medicare Data
Chapter 8: Healthcare Analyt ics 217
Cognitive Computing for Data MiningAs discussed earlier, a major challenge for the analyst is data preparation and deployment of the analytical tools in the most sophisticated software packages. To address this issue, a number of technology firms are developing cognitive computing systems to simplify this work. Cognitive computing systems are designed to mimic human thought and provide natural language interfaces. A leading example is IBM Watson Analytics. Users load data into the system, and Watson performs significant preprocessing to suggest interesting correlations for the analyst to examine.
Exhibit 8.7 shows the starting screen from Watson as it looks at the Medicare beneficiary data used in earlier examples. It immediately offers six questions for the analyst to pursue. It also provides a natural language inquiry interface to delve deeper into the data.
Watson is a sophisticated example of a supervised learning tool and will continue to evolve as its underlying artificial intelligence software improves.
Conclusion
Analytics has become increasingly prevalent in healthcare. Hospitals and health-care systems are using analytics as a means to gain insights into strategic, opera-tional, and clinical issues. Today, the technology enables healthcare analytics to
W3 – Why did you get screening here?
TopicNo. of documents
1 fun,+learn,fun-check,doctor,doc’s office 5
2 +screening,clinic,+check,office,health assessment 6
3 md,md’s office,fair,offer,sucha 1
4 +learn,+live,fun-check,doctor,doc’s office 3
5 time,fun-check,doctor,doc’s office,doc 2
6 fair,information,valuable-love,access,convenient 2
7 +check,work,industry,+thing,health 4
8 doctor,fun-check,+visit,test,doc’s office 2
9 sitting,down,cool,fan,fun-check 1
10 health assessment,keep,assessment,awareness,+build 2
11 random check,random,check,fun-check,doctor 1
12 doc’s office,doc,office,+screening,work 3
EXHIBIT 8.6Text Clustering Results from Health Fair 11 Survey
Healthcare Operat ions Management218
produce better visuals, build more sophisticated models, and analyze much more complex large data sets than at any time in the past. When executed correctly, data are converted into actionable insights that allow enhanced decision making.
Discussion Questions
1. Identify a healthcare operating issue that could benefit from each of the analytical techniques:a. Descriptiveb. Predictivec. Prescriptive
2. How could text mining be used to improve the care of patients with chronic disease?
3. Design a dashboard for each of the following care delivery types:a. Inpatient intensive care unitb. Outpatient imaging centerc. Dental officed. Home health agency
EXHIBIT 8.7 Opening Page
Screenshot from Watson
Analytics
Source: IBM Watson Analytics. Used with permission.
Chapter 8: Healthcare Analyt ics 219
Note
1. Portions of this section are adapted from BrightPoint Consulting (2016). Used with permission.
References
Adams, D. A., R. A. Jajosky, U. Ajani, J. Kriseman, P. Sharp, D. H. Onweh, A. W. Schley, W. J. Anderson, A. Grigoryan, A. E. Aranas, M. S. Wodajo, and J. P. Abellera. 2014. “Summary of Notifiable Diseases—United States, 2012.” Morbidity and Mortality Weekly Report. Published September 19. www.cdc.gov/mmwr/preview/mmwrhtml/mm6153a1.htm.
Barton, M. 2016. “Understanding Population Health Management: A Diabetes Example.” Health Catalyst. Accessed August 31. www.healthcatalyst.com/managing- diabetes-population-health-management.
Begley, S. 2011. “The Best Medicine.” Scientific American 305 (1): 50–55.Blumenthal, D., E. Malphrus, and J. M. McGinnis (ed.), Committee on Core Metrics for
Better Health at Lower Cost, Institute of Medicine. 2015. Vital Signs: Core Metrics for Health and Health Care Progress. Washington, DC: National Academies Press.
BrightPoint Consulting. 2016. “Dashboard Design: Key Performance Indicators and Metrics.” Accessed October 17. www.brightpointinc.com/download/key-performace-indicators/.
Centers for Disease Control and Prevention (CDC). 2009. “The Power of Prevention: Chronic Disease . . . the Public Health Challenge of the 21st Century.” Accessed August 31, 2016. www.cdc.gov/chronicdisease/pdf/2009-power-of-prevention.pdf.
Cook, L. 2015. “America’s Problem with Diabetes, in One Map.” US News & World Report. Published April 9. www.usnews.com/news/blogs/data-mine/2015/04/09/americas-problem-with-diabetes-in-one-map.
Davenport, T. H., and J. G. Harris. 2007. Competing on Analytics: The New Science of Win-ning. Boston: Harvard Business Review Press.
Kindig, D., and G. Stoddart. 2003. “What Is Population Health?” American Journal of Public Health 93 (3): 380–83.
Oxford Living Dictionaries. 2016. “Analytics.” Accessed December 30. https://en.oxforddictionaries.com/definition/analytics.
Winters-Miner, L. A. 2014. “Seven Ways Predictive Analytics Can Improve Health-care.” Elsevier Connect. Published October 6. www.elsevier.com/connect/seven-ways-predictive-analytics-can-improve-healthcare.
CHAPTER
221
QUALITY MANAGEMENT—FOCUS ON SIX SIGMA
Operations Management in Action
In 2015, HealthPartners, an integrated healthcare organization based in Minnesota, conducted a Six Sigma study to examine Clostridium difficile (C. diff) in Regions Hospital. The reduction of C. diff targeted several elements of the Institute for Healthcare Improvement’s Triple Aim.
The business case behind a C. diff proj-ect is that it can address all three elements of the Triple Aim: enhanced care, improved patient experience, and reduced healthcare costs. First, C. diff is the leading cause of antibiotic-associated diarrhea and a highly problematic healthcare-associated infection (Khanna and Pardi 2012). The reduction of C. diff will have a positive impact on patient care. Second, any infection causes pain and discomfort, directly affecting the patient’s experience with the healthcare visit. Finally, a single inpatient C. diff infection incurs a total average cost of more than $35,000 and results in an average increase in hospital length of stay by 2.8 to 5.5 days (Walsh 2012).
The Regions team followed a traditional Six Sigma approach to reduce the number of C. diff cases in its system. The team created a project charter, conducted voice of the customer (VOC) interviews, measured baseline performance, and applied several fundamental quality tools to exam-ine the problem. A cause-and-effect diagram with a five whys analysis led to the discovery of several root causes of C. diff infection at Regions Hospital.
9OVE RVI EW
Quality management became imperative for the manufac-
turing sector in the 1970s and 1980s; for service organiza-
tions in the 1980s and 1990s; and, finally, for the health-
care industry in the 1990s, culminating with the landmark
Institute of Medicine (IOM 1999) report To Err Is Human.
The report details alarming statistics on the number of
people harmed by the US healthcare system and recom-
mends major improvements in quality as related to patient
safety. In it, IOM recognizes the need for systemic changes
and calls for innovative solutions to ensure improvement
in the quality of healthcare.
Since that groundbreaking publication, IOM has
commissioned a committee to make recommendations
on achieving a more value- and science-driven healthcare
system. In the spirit of continuous improvement and with
a nod to the efficacy of Six Sigma, IOM’s resulting report,
Engineering a Learning System, stresses the need to “trans-
form the current healthcare system into one that learns
throughout the continuum of care” (Grossmann et al. 2011,
xviii). The report goes on to discuss the need for that learn-
ing system to adopt continuous improvement and quality
techniques to achieve that goal (Grossmann et al. 2011).
The healthcare industry is facing increasing pres-
sure not only to increase quality but also to reduce costs.
This chapter provides an introduction to quality manage-
ment tools and techniques that are being successfully used
by healthcare organizations. The major topics covered
include the following:
• Defining quality
• The costs of quality(continued)
Healthcare Operat ions Management222
The results of the study helped the team discover several poten-tial sources causing infection rates to rise.
Regions then imple-mented several action items, including standardization of hand sanitization procedures, consistent setup of equip-ment, and enhanced use of signage and visuals. The team conducted a failure mode and effects analysis to assess potential sources of future C. diff infections. Results related to the decrease in C. diff infec-tions showed a positive impact on total number of cases, tests ordered, and costs.
Defining Quality
Although most people agree that ensuring quality in healthcare is of the utmost importance, many disagree on what the term quality means. The supplying organization’s perspective includes performance (or design) quality and con-formance quality. Performance quality includes the features and attributes designed into the product or service. Conformance quality is concerned with how well the product or service conforms to desired goals or specifications.
Garvin (1987) defines eight dimensions of product quality from the customer’s perspective:
• Performance—operating characteristics• Features—supplements to the basic characteristics of the product• Reliability—the probability that the product will work over time• Conformance—product adherence to established standards• Durability—length of time that the product will continue to operate• Serviceability—ease of repair• Esthetics—beauty related to the look or feel of the product• Perceived value—ideas of the product’s worth
OVE RVI EW (Continued)
• The Six Sigma quality program
• Six Sigma tools and techniques (note these are
different from Six Sigma programs), including
the define-measure-analyze-improve-control
(DMAIC) process, the seven basic quality tools,
statistical process control (SPC), and process
capability
• Other quality tools and techniques, including
quality function deployment (QFD), Taguchi
methods, and poka-yoke.
After completing this chapter, readers
should have a basic understanding of quality, quality
programs, and quality tools, enabling application of
the tools and techniques to begin improving quality
in their organizations.
Chapter 9: Qual i ty Management—Focus on Six Sigma 223
Parasuraman, Zeithaml, and Berry (1988) define five dimensions of service quality as follows:
• Tangibles—physical facilities, equipment, and appearance of personnel• Reliability—ability to perform promised service dependably and
accurately• Responsiveness—willingness to help customers and provide prompt
service• Assurance—knowledge and courtesy of employees and their ability to
inspire trust and confidence• Empathy—care and individualized attention
From the healthcare perspective, most agree that the elements of quality relate to the patient. The 2001 IOM report Crossing the Quality Chasm outlines six dimensions of quality in healthcare: safe, effective, patient centered, timely, efficient, and equitable. In addition, the Quality Assurance Project (2003) found nine dimensions of quality in healthcare: technical performance, access to services, effectiveness of care, efficiency of service delivery, interpersonal relations, continuity of services, safety, physical infrastructure and comfort, and choice. Finally, the Triple Aim highlights patient care, population health, and cost as the critical components of healthcare. Obviously, as in general industry, quality and its various dimensions in healthcare may be viewed in many ways. For each hospital or health system, quality is a vital dimension of how it operates.
Cost of Quality
The costs of quality—or the costs of poor quality, according to Juran and De Feo (2010)—are the costs associated with providing a poor-quality product or service. Crosby (1979) notes that the cost of quality is “the expense of nonconformance—the cost of doing things wrong.”
Quality improvement initiatives and projects cannot be justified sim-ply because “everyone is doing it”; they must be considered on the basis of financial or societal benefits. Goldstein and Iossifova (2011) demonstrated that hospitals with significant financial resources were able to benefit greatly from the use of quality management practices. Fineberg (2012) notes that the potential annual excess cost from systemic waste in the US healthcare system is more than $765 billion, including $210 billion in unnecessary services, $130 billion in inefficiently delivered services, $190 billion in excess administrative costs, $105 billion in excessively high prices, $55 billion in missed opportunities
Cost of qualityThe costs associated with producing poor-quality goods and services, including tangible costs, such as scrap and rejects, and intangible costs, such as lost customer goodwill.
Healthcare Operat ions Management224
for disease prevention, and $75 billion in fraud. In all, these costs amount to approximately 30 percent of total health expenditures in the system. Problems such as poor quality of care, overtreatment, and administrative waste may account for as much as $1 trillion annually in costs that contribute nothing to the improvement of the health of the population.
According to Juran and De Feo (2010), the cost of quality is usually separated into four parts:
• External failure—costs associated with failure after the customer receives the product or service (e.g., sentinel event, incorrect billing)
• Internal failure—costs associated with failure before the customer receives the product or service (e.g., overtime for nurses because of treatment errors, reinserting an intravenous line several times)
• Appraisal—costs associated with inspecting and evaluating the quality of supplies or the final product or service (e.g., X-ray costs associated with ensuring that no surgical equipment was left inside patients, hiring a person to inspect supply cabinets to make sure the right equipment is in place)
• Prevention—costs incurred to eliminate or minimize appraisal and failure costs (e.g., Six Sigma training costs, automated equipment for laboratory testing)
Often, the costs associated with prevention are seen as expenses, whereas the other, less apparent costs of appraisal and failure are hidden in the system (Suver, Neumann, and Boles 1992). However, preventing quality problems is usually less costly than fixing quality failures. Striving for continuous improve-ment not only improves quality but also can enhance an organization’s financial situation.
While some companies use the costs of quality as a mechanism to catego-rize their overall quality costs, most apply the concept as a way of thinking about quality in the system. Particularly in healthcare, where providers are taught to save the patient at all costs—literally and figuratively—the actual costs related to that mentality can be extreme. For example, suppose a hospital experiences a sentinel event in which a scope was not cleaned prior to surgery. As a result of the problems encountered, one physician starts to clean his own scopes to make sure they are sterile prior to surgery. What are the costs associated with that doctor cleaning his own scopes?
While it may seem that the doctor is doing the right thing by cleaning the scope each time to ensure the quality of the surgery, he is considered an expensive resource whose time spent on an activity is not cost-efficient. How many surgeries does the hospital lose as a result of the doctor not being available?
Viewing the situation from another perspective, the doctor may not be qualified to clean scopes; a lower-cost technician has the appropriate training
Chapter 9: Qual i ty Management—Focus on Six Sigma 225
to perform this job. If no such technician is on staff because the hospital administrator says the budget has no room to hire one to perform this task, someone with higher-cost credentials must do it. These are all costs of poor quality.
Changing the employees’ mind-set to see the cost of poor quality can be a difficult undertaking, as staff are inclined to “do whatever it takes” to get the job done. Such workarounds often lead to lowered overall system quality, and changing the mind-set is essential if a continuous improvement program is to survive in healthcare.
The Six Sigma Quality Program
This book focuses on the Six Sigma methodology because of its popularity and demonstrated effectiveness, but other, equally valid programs for quality management and continuous improvement are available as well. ISO 9000 certification and the Baldrige criteria are two such programs; review chapter 2 to see how the various quality methodologies compare.
Six Sigma was developed in the 1980s at Motorola as the organization’s in-house quality improvement program. Since that time, the methodology has become the defining quality strategy for many organizations. General Electric adopted Six Sigma as a mechanism to gain strategic advantage through qual-ity, and the company is widely recognized as having experienced the greatest success with Six Sigma programs.
Critical to Six Sigma is its focus on strategy with an emphasis on elimi-nating defects through removal of variance in business systems. Six Sigma has been defined as a philosophy, a methodology, a set of tools, and a goal. The Six Sigma philosophy transforms the culture of the organization. Its methodology employs a project team–based approach to process improvement using the define-measure-analyze-improve-control (DMAIC) cycle. As a set of tools, Six Sigma is composed of quantitative and qualitative statistically based tools used to provide management with facts to allow improvement of an organization’s performance. Finally, Six Sigma as a mathematical term (6σ) signifies a goal of no more than 3.4 defects per million opportunities (DPMOs).
Six Sigma programs can take many forms, depending on the organiza-tion adopting them, but those that are successful share some common themes:
• Top management support for Six Sigma as a business strategy• Extensive change management training to pave the way for a new way
of conducting business• Team-based projects for improvement that directly affect the
organization’s strategic success and financial health
Healthcare Operat ions Management226
• Extensive training at all levels of the organization in the methodology and use of tools and techniques
• Emphasis on the DMAIC approach and use of quantitative measures of project success
Strategy and MeasurementThe success of Six Sigma programs hinges on the organization’s ability to use the program as a technique for achieving strategic goals. When properly executed, Six Sigma drives a series of projects to help propel the organization’s strategy forward. These projects are often internally focused, such as reducing overall patient length of stay, lowering the cost of inventory, and increasing the accuracy and quality of various procedures. Many of these initiatives originate from the strategic dashboard and balanced scorecard discussed in chapter 4. The balanced scorecard provides the measurement system used when selecting Six Sigma projects and assigning resources.
CultureSix Sigma, like all other successful change initiatives, requires and supports cultural change in the organization. The culture of the organization can be thought of as its personality, made up of the assumptions, values, norms, and beliefs of the whole of the organization’s members. It is demonstrated by how tasks are performed, how problems are solved, and how employees interact with one another and the outside world.
Leaders and employees both shape and are shaped by the culture of the organization. For any entity to achieve 6σ, or 3.4 DPMOs, the entire organiza-tion must adopt a mind-set of continuous improvement. This mind-set is the culture of the organization. Organizations that attempt to use the Six Sigma program but do not embrace the culture necessary to achieve its goals struggle in gaining long-term, sustainable results.
LeadershipTo lead change in any organization, top executives must present a sense of purpose for the organization. Truly supporting a major organizational initiative is more than expressing vocal support for it. The organization’s leaders must first provide human and financial resources to launch the pro-gram and then to “hardwire,” or embed, the gains into the culture of the organization made during the continuous improvement journey. Perhaps the most critical endeavor of top leadership is to set targets for department supervisors, such as the chief of surgery and the director of primary care, and then hold those leaders accountable for achieving those targets. Without a core set of metrics for the organization, any Six Sigma effort is limited in
Chapter 9: Qual i ty Management—Focus on Six Sigma 227
its ability to achieve success because the metrics provide the baseline from which comparison is made.
Organizational Infrastructure and TrainingSuccessful Six Sigma initiatives require a high level of proficiency in the appli-cation of the method’s qualitative and quantitative tools and techniques. To achieve this level of excellence, Six Sigma initiatives involve an extensive amount of training at all levels of the organization.
As shown in exhibit 9.1, the Six Sigma infrastructure is hierarchical. As with any other pyramid, the base provides a broad foundation for the structure. Without the involvement of all employees in the continuous improvement efforts, the structure eventually collapses. Moving up the pyramid, the technical and project management skills of the workers involved increase. As employees receive more training and become more proficient, they are designated as yellow belts, green belts, black belts, and master black belts. At the top of the pyramid is the deployment champion.
Yellow BeltsYellow belts are given training on basic quality management and problem-solving techniques. Training on the fundamental problem-solving techniques can last a half day or a full day. Yellow belts help collect, collate, and analyze data related to projects that affect their workflow.
Deploymentchampion
Master black belts
Black belts
Green belts
Yellow belts
All employees
EXHIBIT 9.1Six Sigma Infrastructure
Healthcare Operat ions Management228
Green BeltsGreen belts organize projects and solve problems at the front lines of the organization. In healthcare Six Sigma programs, the green belts help solve the immediate problems when patients receive care. Most green belts spend between 10 and 25 percent of their time running projects, which usually accounts for two to three projects per year. In highly functional Six Sigma systems, the green belts are effective project managers and tend to be influential people from various departments throughout the organization.
Green belts receive training covering quality management and control, problem solving, data analysis, group facilitation, and project management. To obtain certification, they typically must pass a written examination and successfully complete and defend a Six Sigma project. Green belts continue to perform their usual jobs in addition to Six Sigma projects. Many organizations have a goal of training all their employees to the green-belt level.
Black BeltsIf green belts are project managers, black belts are portfolio managers. The black belt oversees large organizational projects that are usually composed of many smaller green-belt projects. For example, a project to increase throughput in an operating room might consist of several smaller projects involving room turnover, staff scheduling, equipment location accuracy, and many other tasks. An organization’s black belts should focus on achieving significant improve-ment in measurements that the organization deems important.
Black belts have more Six Sigma project leadership experience than green belts have. They also are trained in higher-level statistical methods, and they mentor green belts. Black belts are dedicated full time to the Six Sigma efforts of the organization. Their primary responsibility is to ensure that the major projects selected for deployment are successful.
Master Black BeltsMaster black belts are qualified to train and mentor green belts and black belts and therefore are given extensive training in statistical methods as well as com-munication and teaching skills. Master black belts are often seen as the oracles of the Six Sigma program and treated as internal consultants who make sure all of the project management systems are progressing smoothly.
Deployment ChampionAt the top of the pyramid is the deployment champion, who is responsible for the progress of the Six Sigma program and making sure it hits the targets set in the strategic plan. The deployment champion serves a vital role as the liaison between top management, key process owners, and the various black and green belts in the organization. She coordinates all of the major initiatives,
Chapter 9: Qual i ty Management—Focus on Six Sigma 229
allocates resources, and manages expectations for the major projects in the organization.
The hierarchy framework serves several purposes:
• It provides the organization with in-house experts. • It enables everyone in the organization to speak the same language, to
understand exactly what Six Sigma and Six Sigma projects are all about. • It ensures that the organizational goals and objectives are met.• Using black belts for a limited amount of time and then returning them
to their usual positions in the organization helps seed the organization with Six Sigma disciples.
Define-Measure-Analyze-Improve-ControlDMAIC is the acronym for the five phases of a Six Sigma project: define, measure, analyze, improve, and control. The DMAIC framework, or improve-ment cycle (exhibit 9.2), is used almost universally to guide Six Sigma process improvement projects. DMAIC is based on the plan-do-check-act continuous improvement cycle developed by Shewhart and Deming (see chapter 2) but is much more specific.
Some observers have defined insanity as doing the same thing over and over again and expecting different results. Six Sigma uses this definition as a fundamental tenet of its philosophy. At its core, this definition assumes that if we do the same things over and over again, the system in which we do those things will produce the same results. The DMAIC process is designed
ANALYZE
MEASURE
DEFINECONTROL
IMPROVE
PLAN
CHECK
ACT
DO
EXHIBIT 9.2DMAIC Process
Healthcare Operat ions Management230
to help develop consistently repeatable processes that deliver value to the end customer—the patient in the healthcare system.
DefineIn the definition phase, the Six Sigma team chooses a project that is aligned with the strategic objectives of the business and the needs or requirements of the customers of the process. The problem to be solved (or process to be improved) is operationally defined in terms of measurable results. “Good” Six Sigma projects typically have the following attributes:
• The project will save or make money for the organization.• The desired process outcomes are measurable.• The problem is important to the business, has a clear relationship
to organizational strategy, and is (or will be) supported by the organization.
A benchmarking study of project selection found that most organiza-tions (89 percent of respondents) prioritized Six Sigma projects on the basis of financial savings (Evans and Lindsey 2015). The survey also found that the existence of formal project selection processes, process documentation, and rigorous requirements for project approval were all important to the success of Six Sigma projects.
In the definition phase, internal and external customers of the process are identified and their “critical to quality” characteristics (CTQs) are determined. CTQs are the key measurable characteristics of a product or process for which minimum performance standards desired by the customer can be determined. Often, CTQs must be translated from a qualitative customer statement to a quantitative specification. In this phase, the team also defines project boundar-ies and maps the process (mapping is discussed in chapter 6).
MeasureIn the measurement phase, team members must understand how well the pro-cess they are analyzing meets the requirements set by the customer. To gain this knowledge, the team determines the current capability and stability of the process. Using the function Y = f(x) as a mechanism to understand how process outputs (Y) are affected by certain activities or tasks (x), the team begins by collecting data on the key process output variables. Once the key variables are identified, reliable metrics are determined for them (exhibit 9.3). The inputs to the process are identified and prioritized. Root-cause analysis (RCA) or failure mode and effects analysis (FMEA) is sometimes used here to determine the key process input variables. Valid, reliable metrics are determined for the
Chapter 9: Qual i ty Management—Focus on Six Sigma 231
input variables as well. A data collection plan for the process is established and implemented related to the input and output variables. The purpose of this phase of the project is to establish the current state of the process to evaluate the impact of any changes to it.
AnalyzeIn the analysis phase, the team studies the data that have been collected to determine true root causes, or which of the many input variables can be best used to eliminate variation or failure in the process and improve the outcomes.
ImproveIn the improvement phase, the team identifies, evaluates, and implements the improvement solutions. Possible solutions are identified and evaluated in terms of their probability of successful implementation. A plan for deployment of solutions is developed, and the solutions are put in place. Here, actual results should be measured to quantify the impact of the project.
Critical to the improvement phase is ensuring that the tested and imple-mented solutions address the problems identified in the project. People on Six Sigma teams often arrive with preconceived notions on how to solve problems and may manipulate the data results to justify their solution (Bednarz 2012). For example, say a director wants to hire a new doctor. To justify his request, he looks at the results and points to a lack of capacity in the system. However, the constraint in the system may not be due to physician understaffing, and the director’s solution, if implemented, would increase rather than reduce spending and have no impact on system capacity.
In other words, the solutions that are put in place should address the issues uncovered in the data analysis phase of the project. When team members push solutions that do not resolve the issues uncovered by the data, inferior solutions may reduce performance and increase costs. Over time, as inferior solutions continue to be implemented, the organization abandons programs like Six Sigma because of lack of positive results from the program. The implementation and control phases of any project are the most difficult in which to achieve success.
CUSTOMERS
Key process
inputvariables
Key processoutput
variables
Critical to
quality
INPUT OUTPUTPROCESS
EXHIBIT 9.3Six Sigma Process Metrics
Healthcare Operat ions Management232
ControlIn the control phase, controls (discussed in chapter 6) are put in place to ensure that process improvement gains are maintained and the process does not revert to the “old way of doing things.” The improvements are institutionalized through modification of structures and systems (training, incentives, monitoring). This hardwiring of the change eventually becomes the new baseline for the system.
Seven Basic Quality ToolsThe seven fundamental tools used in quality management and Six Sigma were first popularized by Kauro Ishikawa (1985), who believed that up to 95 percent of quality-related problems could be solved with the following seven funda-mental tools (see exhibit 9.4):
• Fishbone diagram—tool for analyzing and illustrating the root causes of an effect (chapter 6)
• Check sheet—simple form used to collect data in which hatch marks are used to record frequency of occurrence for various categories; frequently used to produce histograms and Pareto charts (chapter 6)
• Histogram—graph used to show frequency distributions (chapter 7)• Pareto chart—sorted histogram, used to separate the vital few from the
trivial many, founded on the idea that 80 percent of quality problems are due to 20 percent of causes (chapter 7)
• Flowchart—process map (chapter 6)• Scatter plot—graphical technique to analyze the relationship between
two variables (chapter 7)• Run chart—plot of a process characteristic, in chronological sequence,
used to examine trends; control charts, discussed in the next section, are a type of run chart
Run chart
Scatter plot
Fishbonediagram
Check sheet
FlowchartPareto chart
Histogram
EXHIBIT 9.4Seven Quality
Tools
Chapter 9: Qual i ty Management—Focus on Six Sigma 233
Statistical Process ControlSPC is a statistics-based methodology for determining when a process is mov-ing out of control. All processes have variation in output, some of it caused by factors that can be identified and managed, known as assignable or special variation, and some of it inherent in the process, called common variation. SPC aims to discover variation due to assignable causes so that adjustments can be made and “bad” output is not produced.
In SPC, samples of process output are taken over time, measured, and plotted on a control chart. From statistics theory, we know that the sample means follow a normal distribution. From the central limit theorem, 99.7 per-cent of sample means have a sample mean within a positive or negative three standard errors (±3 SE) of the overall mean and 0.3 percent have a sample mean outside those limits. If the process works as intended, only 3 times out of 1,000 would a sample mean outside the ±3 SE limits be obtained. These ±3 SE limits (sx) are the control limits on a control chart.
If the sample means fall outside the control limits (or follow statistically unusual patterns), the process is likely experiencing variation due to assignable or special causes and is out of control. The special causes should be found and corrected. After the process is fixed, the sample means should fall within the control limits and the process should again be in control.
Some statistically unusual patterns that indicate a process is out of con-trol are shown in exhibit 9.5. A more complete list can be found in Pyzdek and Keller (2014).
Often, the sample mean (X , called X-bar) or X-bar chart is used in conjunction with a range (r) chart. r-Charts follow many of the same rules as X-bar charts and can be used as an additional check on the status of a process. In addition, c-charts are used when the measured process output is the count of discrete events (e.g., number of occurrences in a day), and p-charts are used when the output is a proportion. Lim (2003) describes more sophisticated types of control charts that can be used in healthcare organizations.
A control chart may also be set up using individual values rather than sample means. However, this step often is not taken for two reasons. First, the individual values must be normally distributed. Second, data collection can be expensive; typically, collecting samples of the data costs less than collecting all of the data.
Riverview Clinic Statistical Process ControlThe Riverview Clinic of Vincent Valley Hospital and Health System (VVH) is undertaking a Six Sigma project to reduce its waiting times. In the measurement phase of the project, data have been collected on waiting time and a control
Control limitsCommon variation limits that are ±3 standard deviations from the mean.
X-bar chartMeasures process performance of sample means for continuous data.
Range (r) chartMeasures process performance of sample ranges for continuous data.
Healthcare Operat ions Management234
Observation
Observation
Observation
Observation
8 or more samples above (or below) mean
14 or more samples oscillating
6 or more samples increasing (or decreasing)
UCL = 3
LCL = –3
X = 0
UCL = 3
LCL = –3
X = 0
UCL = 3
LCL = –3
X = 0
UCL = 3
LCL = –3
X = 0
One sample more than 3 standard errors from mean
Note: LCL = lower control limit; UCL = upper control limit.
EXHIBIT 9.5Out-of-Control
Observation Patterns
Chapter 9: Qual i ty Management—Focus on Six Sigma 235
chart format selected to help management understand the current situation. Six observations of waiting time are made over 20 days. At randomly chosen times throughout each of the 20 days, the next patient to enter the clinic is chosen. The time from when this patient enters the clinic until he exits is recorded (exhibit 9.6).
Observations of Wait Times (minutes)
Observation
Day 1 2 3 4 5 6 Sample Mean Sample Range
1 29 29 22 31 29 31 28.50 9
2 24 29 40 26 36 30 30.83 16
3 28 33 25 26 28 33 28.83 8
4 26 31 38 30 23 28 29.33 15
5 36 29 24 29 26 32 29.33 12
6 26 27 32 25 30 29 28.17 7
7 22 33 30 31 37 34 31.17 15
8 40 29 26 29 32 30 31.00 14
9 32 32 21 34 28 29 29.33 13
10 34 26 35 27 31 26 29.83 9
11 35 30 29 30 31 27 30.33 8
12 31 39 32 32 30 31 32.50 9
13 36 24 30 29 31 26 29.33 12
14 25 23 29 31 25 23 26.00 8
15 38 43 37 35 38 32 37.17 11
16 35 29 30 25 28 30 29.50 10
17 26 29 20 33 30 28 27.67 13
18 22 29 26 30 36 28 28.50 14
19 33 33 34 37 28 30 32.50 9
20 26 26 34 34 25 36 30.17 11
Standard Deviation = 4.42 Overall Mean = 30.00
EXHIBIT 9.6Riverview Clinic Wait Times, in Minutes
Healthcare Operat ions Management236
Riverview uses the standard deviation of all of the observations to esti-mate the standard deviation of the population. The three-sigma control limits for the X-bar chart are calculated as follows:
x −zα/2×σx ≤µ≤ x + zα/2×σx
x −zα/2×sn≤µ≤ x + zα/2×
sn
30−3×4.4
6≤µ≤30+3×
4.46
30−5.4≤µ≤30+5.4
24.6≤µ≤35.4.
Looking at the control chart (exhibit 9.7), it appears that day 15 was out of control. An investigation found that on day 15 the clinic was short-staffed because of a school holiday. The control chart cannot be used as is because of the out-of-control point. Knowing that they may either continue to collect data until all points are in control or recalculate the control chart limits excluding day 15, Riverview leaders choose to recalculate, and the new three-sigma limits are
x −zα/2×σx ≤µ≤ x + zα/2×σx
x −zα/2×sn≤µ≤ x + zα/2×
sn
30−3×4.1
6≤µ≤30+3×
4.16
30−5.0≤µ≤30+5.0
25.0≤µ≤35.0.
Unless the system is changed, 50 percent of Riverview patients will experience a wait time longer than 30 minutes (50 percent will experience a wait time of less than 30 minutes), and 10 percent of Riverview patients will experience a wait time of greater than 35.3 minutes (90 percent will experience a wait time of less than 35.3 minutes).
µ ≤ X + zα × σx
µ ≤ x + α × s; z0.9 = 1.3
µ ≤ 30 + (1.3 × 4.1)
µ ≤ 30 + 5.3
µ ≤ 35.3.
Chapter 9: Qual i ty Management—Focus on Six Sigma 237
If Riverview’s goal for its Six Sigma project is to ensure that 90 percent of patients experience a wait time of no more than 30 minutes, the clinic needs to improve the system. The Six Sigma team’s aim would be to reduce mean wait time to 24.7 minutes if the process variation remains the same (exhibit 9.8).
µ ≤ x + zα × σx
µ ≤ x + zα × s; z0.9 = 1.3
µ ≤ 30
µ ≤ 24.7 + 5.3; x = 24.7
15 20 25 30
Wait Time (minutes)
35 40 45
CURRENT WAIT TIME
50% of patients
wait more than 30 minutes
15 20 25 30
Wait Time (minutes)
35 40 45
WAIT TIME GOAL
10% of patients
wait more than 30 minutes
EXHIBIT 9.8 Riverview Clinic Wait Time
20
25
30
35
40
0 5 10 15 20 25 30
Day
Mea
n W
ait T
ime
(min
utes
)
±1 ±2 ±3
Out-of-control sample
EXHIBIT 9.7Riverview Clinic Wait Times: X-Bar Control Chart
Healthcare Operat ions Management238
Process Capability and Six Sigma QualityProcess capability measures how well a process can produce output that meets desired standards or specifications. This critical measurement in Six Sigma sys-tems determines how well the internal processes conform to customer require-ments. Process capability is measured by comparing the natural (or common) variability of an in-control process, the process width, to the specification width. Specifications are determined by outside forces (such as customers or management), but process variability is not determined—it is simply a natural part of any process. A capable process is one that produces few defects, where a defect is defined as an output outside specification limits.
The two common measures of process capability are Cp and Cpk. Cp is used when the process is centered on the specification limits; the mean of the process is the same as the mean of the specification limits. Cpk is used when the process is not centered. A capable process shows a Cp or Cpk
greater than 1. At a Cp
of 1, the process produces about 3 defects per 1,000 attempts or opportunities.
C Cs
Cx x
Cx
sx
s
USL LSL6
and is estimated by ˆ USL LSL6
minLSL
3or
USL3
and is estimated by ˆ minLSL
3or
USL3
,
p p
pk
pk
σ
σ σ
=−
=−
=− −
=− −
where USL = upper specification limit and LSL = lower specification limit.Recall that Six Sigma quality is defined as fewer than 3.4 DPMOs. This
definition can be somewhat confusing, as it corresponds to the 4.5σ one-tail probability limit for the normal distribution. Six Sigma allows for a 1.5σ shift in the mean of the process and Cpk
= 1.5 (exhibit 9.9).
Riverview Clinic Process CapabilityRiverview Clinic management has decided that no patient should wait more than 40 minutes, or a waiting time USL of 40 minutes. The Six Sigma team wants to determine if the process is capable of achieving that wait time thresh-old. Note that because no lower specification limit is specified (waiting time less than some lower limit would not be considered a defect), Cpk
is the correct measure of process capability.
Ĉ pk =USL−x
3s⎛⎝⎜⎜⎜
⎞⎠⎟⎟⎟=
40−303×4.1
=10
12.3= 0.81
Process capabilityA measure of how well a process can produce output that meets desired standards or specifications.
Chapter 9: Qual i ty Management—Focus on Six Sigma 239
The Cpk is less than 1. Therefore, the process is not capable, and 7,000 DPMOs are expected [x ~ N(30, 16.8), P(x > 40) = 0.007].
The team determines that to ensure Six Sigma quality, the specification limit needs to be 48.8 minutes:
= =−
=−
=−
∴ =Cx
sˆ 1.5
USL3
USL 3012.3
48.8 3012.3
USL 48.8.pk
If the Riverview Six Sigma team determines that Six Sigma quality with a specification limit of 40 minutes is a reasonable goal, it may reduce average wait time, reduce the variation in the process, or seek some combination of both.
Cx
sx
x
ss
ˆ USL3
USL12.3
40 2112.3
21
1.540 30
310
3 2.22.2
pk
=−
=−=
−∴ =
=−
=×
∴ =
Average wait time must be reduced to 21 minutes, or the standard deviation of the process reduced to 2.2 minutes, to reach the goal.
Rolled Throughput YieldRolled throughput yield (RTY) measures overall process performance. It is the probability that a unit (of product or service) will pass through all process
Rolled throughput yield (RTY)The probability that a unit (of product or service) will pass through all process steps free of defects.
–7 –6 –5 –4 –3 –2 –1 0 1 2 3 4 5 6 7
1.5 Shift
LowerSpecification
Limit
UpperSpecification
Limit
3.4 DPMO
EXHIBIT 9.9 Six Sigma Process Capability Limits
Healthcare Operat ions Management240
steps free of defects. For example, consider a process composed of four steps or subprocesses. If each step has a 5 percent probability of producing an error or defect (95 percent probability of an error-free outcome), the RTY of the overall process is 81 percent—considerably lower than that in the individual steps (exhibit 9.10).
Additional Quality Tools
In addition to the quality tools and techniques commonly associated with Six Sigma, many other tools can be used in process improvement. Quality function deployment (QFD) and Taguchi methods are often applied in the development of new products or processes to ensure quality outcomes. However, they can also be used to improve existing products and processes. Benchmarking helps to determine best practices and to adapt them to the organization to achieve superior performance. Mistake proofing, or poka-yoke, is used to minimize the possibility of an error occurring. All of these tools are essential components of an organization’s quality toolbox.
Quality Function DeploymentQFD is a structured process for identifying customer needs and wants and translating them to a product or process that meets those needs. This tool is most often used in the development phase of a new product or process, but it can also be applied to redesign an existing product or process. Typically, QFD is found in a design for Six Sigma (DFSS) project, where the goal is to design the process to meet Six Sigma goals. The QFD process uses a matrix called the house of quality (exhibit 9.11) to organize data in a usable fashion.
The first step in QFD is to determine customer requirements. Customer requirements represent the VOC and are often stated in customer terms, not technical terms. A particular product or service can have many customers, and the voice of all must be heard.
Market research tools are used to capture the VOC. The many customer needs discovered are organized into a few key customer requirements, which
Quality function deployment (QFD)A technique that translates customer requirements to specific product or process requirements.
Step 2 Step 3 Step 486 in,
81 error-free
productsout
Step 1100 in,
95 error-free
productsout
95 in, 90 error-
free products
out
90 in, 85 error-
free products
out
EXHIBIT 9.10 Rolled
Throughput Yield
Chapter 9: Qual i ty Management—Focus on Six Sigma 241
are weighted on the basis of their relative importance to the customer. Typi-cally, a scale of 1 to 5 is used, with 5 representing the most important. The customer requirements and their related importance are listed on the left side of the QFD diagram.
A competitive analysis of the identified customer needs is also per-formed. The question here is how well competitors meet customer needs. Typically, a scale of 1 to 5 is used here as well, with 5 indicating that the competitor completely meets the particular need. The competitive analysis is used to focus the development of the service or product on areas that pre sent opportunities to gain competitive advantage and where the organization is at a competitive disadvantage. This assessment can help the development team focus on important strategic characteristics of the product or service. The competitors’ scores on each customer requirement are listed on the right side of the QFD diagram.
Technical requirements of the product or process that relate to customer requirements are determined next. For example, if customers want speedy service time, a related technical requirement might be that 90 percent of all service times are less than 20 minutes. The technical requirements are listed horizontally across the top of the QFD diagram, and the relationship between the customer requirements and technical requirements is evaluated. Usually, the relationships are evaluated as strong, medium, or weak, and symbols represent these relationships in the relationship matrix. Numeric values are assigned to
Correlation matrix
Technical requirements
Customer requirements
Competitive analysis
Relationship matrix
Specifications or
target values
Impo
rtan
ce
Importance weight
EXHIBIT 9.11 House of Quality Correlation Matrix
Healthcare Operat ions Management242
the relative weights (5 = strong, 3 = medium, 1 = weak), and these values are placed in the matrix.
Positive and negative interactions among the technical requirements are evaluated as strongly positive, positive, strongly negative, and negative. Another set of symbols represents these relationships in the “roof,” or correla-tion matrix, of the house of quality. This framework makes clear the trade-offs involved in product and process design.
Customer importance weights are multiplied by relationship weights and summed for each technical requirement to determine the overall impor-tance weights. Target values are then developed from the house of quality that emerges from the process.
Historically, QFD was a phased process. The above-described process is the planning phase; for product development, planning is followed by the construction of additional houses of quality related to parts, process, and pro-duction. For service development and improvement, using only the first house of quality, or the first house of quality followed by the process house, is often sufficient. For examples of QFD applications in healthcare environments, see Sarker and colleagues (2010), and for a complete review of QFD applications in healthcare and other industries, see Sharma and Rawani (2010).
Riverview Clinic Quality Function DeploymentMany diabetes patients at Riverview Clinic do not return for routine preventive exams. The team formed to address this problem has decided to use QFD to improve the process and begins by soliciting the VOC via focus groups. The team finds the following patient needs and wants:
• To know (or be reminded) that they need to schedule a preventive exam
• To know why an office visit is needed• A convenient means to schedule their appointments• That their appointments be on time• To know that their appointments will last a certain length of time
Next, patient rankings of the importance of these needs and wants are determined via patient surveys.
A competitive analysis of Riverview Clinic’s two main competitors follows to assess how they are meeting the determined needs and wants of Riverview’s patients with diabetes. The team develops related technical requirements and evaluates the interactions between them. The resulting house of quality is shown in exhibit 9.12. On-time appointments emerge as the highest importance ranking because they affect both appointment time and appointment length.
Chapter 9: Qual i ty Management—Focus on Six Sigma 243
The team then evaluates various process changes and improvements related to the determined technical requirements. To meet these technical requirements, it decides to notify patients via postcard and to follow this method with e-mail and phone notification if needed. The postcard and e-mail contain information related to the need for an office visit and direct patients to the clinic’s website for more information. Appointment scheduling is made available via the Internet as well as by phone. Staffing levels and appointment times are adjusted to ensure that appointments take place on time as scheduled and are approximately the same length. Training is conducted to help physicians and nurses understand the need to maintain appointment lengths, and tools are provide to them to ensure consistent length.
Exhibit 9.13 outlines these process changes and related technical require-ments. After the changes are implemented, the team checks to ensure that the technical requirements determined by the house of quality are being met.
Taguchi MethodsTaguchi methods refer to two related ideas first introduced by Genichi Tagu-chi (Barsalou 2013). Rather than deeming the quality of a product or service as good or bad, whereby good falls within some specified tolerance limits and bad falls outside those limits, quality is related to the distance from some target value; further from the target is worse. Taguchi developed experimental design
Taguchi methodsApproaches to quality whereby product development focuses on “perfect” rather than on conformance to specifications.
Competitor B
Competitor A
Our service
On-tim
e appointm
ent 90%
8 minutes
8 minutes
Yes
3
3 channels
Appointm
ent length range
Time to
schedule
Information
on need
Subsequent notification
Initial notification
3
3
3
3
3
4
3
3
5
5
2
2
3
3
3
29 27 20 15 15 25
5 3
3
4
3
4
3
5 5
Appointment time
5 Appointment length
5 Convenient
5 Why knowledge
3 Time knowledge
+ +
EXHIBIT 9.12 Riverview Clinic House of Quality for Patients with Diabetes
Healthcare Operat ions Management244
techniques in which the target value and the associated variation are important factors. The optimal process design is not necessarily where the target value is maximized but where variation is minimal in relation to the target. In other words, the process is robust and performs well under less-than-ideal conditions.
Taguchi methods are often applied in DFSS where the product or ser-vice is designed to be error free while meeting or exceeding the needs of the customer. Rather than fixing an existing product or service, the design process of the product or service ensures quality from the start.
BenchmarkingAccording to the American Productivity and Quality Center (2005), bench-marking is “the process of identifying, understanding, and adapting outstand-ing practices and processes from organizations anywhere in the world to help your organization improve its performance.” Benchmarking focuses on how to improve any given process by finding, studying, and implementing best practices.
These best practices may be found in the organization, in competi-tor organizations, and even in organizations outside the particular market or industry. Best practices are everywhere—the challenge is to find them and adapt them to the organization.
The benchmarking process consists of deciding what to benchmark, determining how to measure it, gathering information and data, and then implementing the best practice in the organization. Benchmarking can be an important part of a quality improvement initiative, and many healthcare organizations are involved in benchmarking (Olson et al. 2008). The journal Healthcare Benchmarks and Quality Improvement provides information on many of these initiatives.
Technical Requirement Process Change
Initial notification Postcard mailed
Subsequent notifications E-mail and/or phone call
Information on need Website
Time to schedule Website and phone
Appointment length range Staff levels adjusted
On-time appointment Staff training
Note: QFD = quality function deployment
EXHIBIT 9.13Riverview
QFD Technical Requirement and Related
Process Change
Chapter 9: Qual i ty Management—Focus on Six Sigma 245
Poka-YokePoka-yoke (a Japanese phrase meaning to avoid inadvertent errors), or mis-take proofing, is a way to prevent errors from occurring. A poka-yoke is a mechanism that either prevents a mistake from being made or makes the mis-take immediately obvious so that no adverse outcomes are experienced. For example, all of the instruments required in a surgical procedure are placed on an instrument tray with unique indentations for each instrument. After the procedure is complete, the instruments are replaced in the tray. This process provides a quick means to visually check that all instruments are removed from the patient before closing the patient’s incision.
Another example of mistake proofing is locating the controls for a mam-mography machine in such a way that the technician cannot start the machine unless she is shielded from radiation by a wall that separates her from the machine. In FMEA, identified fail points are good candidates for poka-yoke.
Technology can often enable poka-yoke. When patient data are put into a system, the software is often programmed to provide an error message if the data are incorrect. For example, a Social Security number is nine digits long; no more than nine digits can be entered into the Social Security field, and an error message appears if fewer than nine digits are entered. In the past, surgi-cal sponges were counted before and after a procedure to ensure that none were left in a patient. Now, the sponges can be radio frequency identification tagged, eliminating the error-prone counting process, and a simple scan can determine if any sponges remain in the patient.
Riverview Clinic Six Sigma Generic Drug Project
Riverview Clinic’s management team has determined that meeting pay-for-performance goals related to prescribing generic drugs is a strategic objective for the organization, and a project team has been organized to meet this goal. Benchmarking is performed to help the team determine which pay-for-performance measure to focus on and to define reasonable goals for the project. The team has found that 10 percent of nongeneric prescription drugs could be replaced with generic drugs, an approach taken by other clinics that have successfully met this goal.
DefineIn the definition phase, the team articulates the project goals, scope, and business case. This activity includes developing the project charter, determining customer requirements, and diagramming a process map. (The charter for a similar project is found in chapter 5; it defines the project’s goals, scope, and business case.)
Poka-yokeA mechanism that prevents mistakes or makes them immediately obvious to prevent adverse outcomes.
Healthcare Operat ions Management246
The team identifies the health plans and patients as customers of the process. The outputs of the process are identified as prescriptions and the effi-cacy of those prescriptions. The process inputs are physician judgment and the information technology (IT) system for drug lists. Additionally, pharmaceutical firms provide input on drug efficacy. The process map developed by the team is shown in exhibit 9.14.
MeasureThe team decides to quantify the outcomes using the percentage of generic (versus nongeneric) drugs prescribed and the percentage of prescription changes following the prescribing of a generic drug. Additionally, the team tracks and records data on all nongeneric drugs prescribed by each individual clinician for one month.
AnalyzeAfter one month, the team analyzes the data and finds that, overall, clinicians prescribed 65 percent generic drugs (exhibit 9.15) and prescription changes were needed for 3 percent of all prescriptions. A sample of the data collected is shown in exhibit 9.16.
The team generates a Pareto analysis by clinician and drug to determine if particular drugs or clinicians were more problematic than others. The analysis shows that some drugs caused more problems leading to represcribing but that all clinicians showed roughly the same outcomes (exhibit 9.17).
The team reexamines its stated goal of increasing generic drug prescrip-tions by 4 percent in light of the data collected. If all prescriptions for the top
Information on drugs
Clinician prescribes
drug
Type of drug
Drug doesn’t work
Generic Drug efficacy
Drug efficacy
Drug works
Drug works
End
End
Patient needs drug
Nongeneric
Drug doesn’t work
EXHIBIT 9.14 Riverview Clinic
Prescription Process
Chapter 9: Qual i ty Management—Focus on Six Sigma 247
65%
15%
20% Generic
Nongeneric, generic availableNongeneric, generic not available
35%
EXHIBIT 9.15 Riverview Generic Drug Project: Drug Type and Availability
Date Clinician Drug Drug TypeGeneric
Available Represcribe
1-Jan Smith F Nongeneric Yes No
1-Jan Davis G Generic Yes No
1-Jan Jones L Generic Yes No
1-Jan Anderson F Nongeneric No No
1-Jan Swanson R Generic Yes Yes
1-Jan Smith S Nongeneric Yes No
1-Jan Swanson U Generic Yes No
1-Jan Jones P Generic Yes No
1-Jan Jones S Nongeneric No No
1-Jan Swanson A Generic Yes No
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
31-Jan Anderson F Nongeneric Yes No
31-Jan Anderson E Nongeneric No No
31-Jan Davis T Generic Yes No
31-Jan Smith Y Generic Yes No
31-Jan Jones D Generic Yes No
31-Jan Swanson J Generic Yes No
31-Jan Swanson I Nongeneric Yes No
31-Jan Smith T Generic Yes No
31-Jan Davis G Generic Yes No
31-Jan Anderson H Generic Yes No
EXHIBIT 9.16Riverview Clinic Generic Drug Project Sample Data
Healthcare Operat ions Management248
four nongeneric drugs for which a generic drug is available could be changed to generics, Riverview would increase generic prescriptions by 5 percent. There-fore, management decides that the original goal is still reasonable.
ImproveThe team conducts an RCA of the reasons for prescribing nongeneric drugs and determines that the major cause was the clinicians’ lack of awareness of a generic replacement for the prescribed drug. In addition to adapting the IT system to identify approved generic drugs, the team publishes a monthly top five list (on the basis of data from the previous month) of nongeneric drugs for which an approved generic exists. The team continues to collect and ana-lyze data after these changes are implemented and finds that prescriptions for generic drugs have risen by 4.5 percent after six months.
ControlTo measure progress and ensure continued compliance, the team sets up a weekly control chart for generic prescriptions and continues to monitor and publish the top five list. It conducts an end-of-project evaluation to document the steps taken and results achieved and to ensure that learning from the project is retained in the organization.
Clinician Prescriptions
0
5
10
15
20
Davis Jones Smith Swanson AndersonC L I N I C I A N
Non
gene
ric
pres
crip
tion
s w
here
ther
e is
a g
ener
icav
aila
ble/
mon
th
Nongeneric Prescriptions Where There is a Generic Available
0
5
10
15
20
O W J V B H M A C I G D K L T U N P Q R
D R U G
Pres
crip
tion
s/m
onth
EXHIBIT 9.17 Riverview Clinic
Generic Drug Project: Pareto
Diagrams
Chapter 9: Qual i ty Management—Focus on Six Sigma 249
Tool or Technique Define Measure Analyze Improve Control
7 quality control tools
Cause-and-effect diagram x
Run chart x x
Check sheet
Histogram x x
Pareto chart x x x
Scatter plot x x
Flowchart x x
Other tools and techniques
Mind mapping/brainstorming x x x
5 Whys/RCA x
FMEA x x
Pie chart x
Hypothesis testing x
Control chart x x x
Process capability x x x
QFD x x x
Benchmarking x x x x
Poka-yoke x
Gantt chart x
Project planning x x x
Charters x
Tree diagram x
Force field analysis x x
Balanced scorecard x x x x x
Note: FMEA = failure mode and effects analysis; QFD = quality function deployment; RCA = root-cause analysis.
EXHIBIT 9.18Quality Tools and Techniques Selector Chart
Healthcare Operat ions Management250
Conclusion
The Six Sigma DMAIC process is a framework for improvement. At any point in the process, revisiting an earlier step in the process may be necessary to ensure that improvement is achieved. For example, what the process improvement team thought was the root cause of a problem of interest may be found not to be the true root cause. Or when attempting to analyze the data, insufficient or incorrect data may have been collected. In both cases, the team may need to go back in the DMAIC process to ensure that a project is successful.
At each step in the DMAIC process, various tools can be used. The choice of tool is related to the problem and possible solutions. Exhibit 9.18 outlines suggestions for when to choose a particular tool or technique. This chart is only a guideline—you should use whatever tool is most appropriate for the situation.
Discussion Questions
1. Read the executive summary of the IOM (1999) report To Err Is Human (www.nap.edu/read/9728/chapter/2?term=executive+summary) and answer the following questions: a. Why did this report spur an interest in quality management in the
healthcare industry? b. What does IOM recommend to address these problems? c. Conduct a search and determine how much progress has been made
since 1999.2. What does quality in healthcare mean to your organization? To you
personally?3. Discuss a real example of each of the four costs of quality in a healthcare
organization.4. List at least three poka-yokes currently used in the healthcare industry.
Can you think of a new one for your organization?
Exercises
1. Clinicians at VVH have been complaining about the turnaround time for blood work. The laboratory manager decides to investigate the problem and collects turnaround time data on five randomly selected requests every day for one month (shown in the chart on the next page).
Chapter 9: Qual i ty Management—Focus on Six Sigma 251
a. Construct an X-bar chart using the standard deviation of the observations to estimate the population standard deviation. Construct an X-bar chart and r-chart using the range to calculate the control limits. (The Excel template on the book’s companion website performs this calculation for you.)
b. Is the process in control? Explain.c. If the clinicians feel that any time over 100 minutes is unacceptable,
what are the Cp and Cpk of this process?
d. What are the next steps for the laboratory manager?2. Riverview Clinic has started a customer satisfaction program. In
addition to other questions, each patient is asked if she is satisfied with her overall experience at the clinic. Patients can respond “yes” if they were satisfied or “no” if they were not satisfied. Typically, 200 patients are seen at the clinic each day. The data collected for two months are shown on the next page.
Observation Observation
Day 1 2 3 4 5 Day 1 2 3 4 5
1 44 41 80 51 25 16 14 44 35 52 76
2 28 32 58 42 18 17 52 84 55 63 15
3 54 83 59 50 46 18 28 20 67 76 69
4 57 53 63 15 52 19 25 23 35 21 23
5 30 50 62 68 42 20 46 74 24 10 47
6 42 40 50 49 73 21 33 54 62 14 72
7 26 17 50 47 91 22 64 55 62 14 72
8 54 39 39 82 28 23 53 49 72 49 61
9 46 62 53 64 57 24 15 16 18 35 78
10 49 71 34 42 43 25 64 9 51 47 70
11 53 64 12 35 43 26 36 21 51 40 57
12 75 43 43 50 64 27 24 58 19 88 16
13 74 19 52 55 59 28 75 66 34 27 71
14 91 40 66 15 73 29 60 42 20 59 60
15 59 32 59 49 71 30 52 28 85 39 67
On the web at ache.org/books/OpsManagement3
Healthcare Operat ions Management252
Day
Proportion of patients who were
unsatisfied Day
Proportion of patients who were
unsatisfied Day
Proportion of patients who were
unsatisfied
1 0.17 15 0.15 28 0.18
2 0.13 16 0.14 29 0.19
3 0.15 17 0.13 30 0.14
4 0.22 18 0.15 31 0.19
5 0.16 19 0.15 32 0.10
6 0.13 20 0.22 33 0.17
7 0.17 21 0.19 34 0.15
8 0.17 22 0.15 35 0.17
9 0.11 23 0.12 36 0.15
10 0.16 24 0.16 37 0.15
11 0.15 25 0.18 38 0.15
12 0.17 26 0.14 39 0.14
13 0.17 27 0.17 40 0.19
14 0.12
a. Construct a p-chart using the collected data.b. Is the process in control?c. On average, how many patients are satisfied with Riverview Clinic’s
service? If Riverview wants 90 percent (on average) of patients to be satisfied, what should the clinic do next?
3. Think of a problem in your organization that Six Sigma could help solve. Map the process and determine the key process input variables, the key process output variables, the CTQs, and exactly how you can measure them.
4. Use QFD to develop a house of quality for the VVH emergency department (you may need to guess the numbers you do not know). The Excel template labeled QFD.xls, available on the companion website, may be helpful in completing this problem.
On the web at ache.org/books/OpsManagement3
Chapter 9: Qual i ty Management—Focus on Six Sigma 253
References
American Productivity and Quality Center. 2005. “Glossary of Benchmarking Terms.” Accessed January 30, 2006. www.apqc.org/portal/apqc/ksn/Glossary%20of%20Benchmarking%20Terms.pdf?paf_gear_id=contentgearhome&paf_dm=full&pageselect=contentitem&docid=119519.
Barsalou, M. 2013. “The Legacies of Genichi Taguchi.” Published March 21. www.quality digest.com/inside/quality-insider-article/legacies-genichi-taguchi.html.
Bednarz, T. F. 2012. “Strategies and Solutions for Solving Team Problems: Teams That Run Smoothly Can Concentrate on Their Primary Goals.” Quality Digest. Published February 2. www.qualitydigest.com/inside/quality-insider-article/strategies-and-solutions-solving-team-problems.html.
Crosby, P. B. 1979. Quality Is Free: The Art of Making Quality Certain. Boston: McGraw-Hill.Evans. J., and W. Lindsey. 2015. An Introduction to Six Sigma and Process Improvement,
2nd edition. Boston: Cengage Learning.Fineberg, H. 2012. “A Successful and Sustainable Health System: How to Get There from
Here.” New England Journal of Medicine 366: 1020–27.Garvin, D. A. 1987. “Competing on the Eight Dimensions of Quality.” Harvard Business
Review 65 (6): 101–10.Goldstein, S. M., and A. R. Iossifova. 2011. “Ten Years After: Interference of Hospital
Slack in Process Performance Benefits of Quality Practices.” Journal of Operations Management 20 (1–2): 44–54.
Grossmann, C., W. A. Goolsby, L. Olsen, and J. M. McGinnis. 2011. Engineering a Learn-ing Healthcare System: A Look at the Future. A Workshop Summary. Washington, DC: National Academies Press.
Institute of Medicine (IOM). 2001. Crossing the Quality Chasm—A New Health System for the 21st Century. Washington, DC: National Academies Press.
. 1999. To Err Is Human: Building a Safer Health System. Washington, DC: National Academies Press.
Ishikawa, K. 1985. What Is Total Quality Control? Translated by D. J. Lu. Englewood Cliffs, NJ: Prentice-Hall.
Juran, J. M., and J. A. De Feo. 2010. Juran’s Quality Handbook: The Complete Guide to Performance Excellence, 6th edition. New York: McGraw-Hill Education.
Khanna, S., and D. S. Pardi. 2012. “Clostridium difficile Infection: New Insights into Management.” Mayo Clinic Proceedings 87 (11): 1106–17.
Lim, T. O. 2003. “Statistical Process Control Tools for Monitoring Clinical Performance.” International Journal for Quality in Health Care 15 (1): 3–4.
Olson, J. R., J. A. Belohlav, L. S. Cook, and J. M. Hays. 2008. “Examining Quality Improve-ment Programs: The Case of Minnesota Hospitals.” Health Services Research 43 (5): 1781–86.
Healthcare Operat ions Management254
Parasuraman, A., V. A. Zeithaml, and L. L. Berry. 1988. “SERVQUAL: A Multiple-Item Scale for Measuring Consumer Perceptions of Service Quality.” Journal of Retail-ing 64 (4): 12–40.
Pyzdek, T., and P. Keller. 2014. The Six Sigma Handbook, 4th edition. New York: McGraw-Hill.Quality Assurance Project. 2003. “Dimensions of Quality.” Accessed January 15, 2006.
www.qaproject.org/methods/resdimension.html.Sarker, S., A. Al Masud, M. A. Habib, and A. K. M. Masud. 2010. “Application of QFD
for Improving Customer Perceived Quality of Synthetic Fiber: A Case of Beximco Synthetics Ltd.” Journal of Business. Accessed May 18, 2012. www.journalofbusi-ness.org/index.php/GJMBR/article/view/133.
Sharma, J. R., and A. M. Rawani. 2010. “From Customers Requirements to Customers Satisfaction: Quality Function Deployment in Service Sector.” International Journal of Productivity and Quality Management 5 (4): 428–39.
Suver, J. D., B. R. Neumann, and K. E. Boles. 1992. “Accounting for the Costs of Quality.” Healthcare Financial Management 46 (9): 28–37.
Walsh, N. 2012. “C. Difficile Inpatient Stays Long, Costly.” Published December 8. MedPage Today. www.medpagetoday.com/MeetingCoverage/ASHP/36339.
CHAPTER
255
THE LEAN ENTERPRISE
Operations Management in Action
Park Nicollet (PN), a healthcare system in Minne-sota, has been using Lean tools to help improve the flow of its processes since 2003. In 2009, PN used Lean tools integrated with clinician guidance to design a new care model to manage patients taking prescribed anticoagulants.
Anticoagulants, such as warfarin, can be dangerous medications. Warfarin is often pre-scribed to cardiac patients to prevent blood clots and is also used to treat or prevent venous throm-bosis and pulmonary embolism. However, the drug can cause bleeding that may be life threatening. As a result, most hospitals have specialized units that deal with the dangers of warfarin.
PN used the Lean tools to help allevi-ate issues with administering anticoagulants to patients. The primary metric that PN focused on was international normalized ratio (INR) time in desired range. The INR is a measurement estab-lished by the World Health Organization (WHO) for reporting the results of blood coagulation tests. PN’s tests saw results in the desired range just 38 percent of the time.
To improve its anticoagulant delivery sys-tem, PN standardized several policies related to the administration of warfarin. First, it established centralized dosing models in which only certain individuals—in PN’s case, nurse clinicians—had the authority to administer the medications. The cen-tralized dosing models greatly improved PN’s ability to track the amount of warfarin given to patients.
10OVE RVI EW
Lean tools and techniques have been employed extensively
in manufacturing organizations since the 1990s to improve
the efficiency and effectiveness of those organizations’
activities. Since that time, many healthcare organizations
realized the transformative potential of Lean to improve
patient safety and financial performance (Dobrzykowski,
McFadden, and Vonderembse 2016).
The healthcare industry faces increasing pressure
to use resources in an effective manner to reduce costs
and increase patient satisfaction. This chapter provides
an introduction to the Lean philosophy as well as the vari-
ous Lean tools and techniques used by many healthcare
organizations today. The major topics covered include the
following:
• The Lean philosophy
• Defining waste
• Kaizen
• Value stream mapping
• Other Lean tools, techniques, and ideas, including
the five Ss, spaghetti diagrams, kaizen events, takt
time, kanbans, rapid changeover, heijunka, jidoka,
andon, standardized work, and pull
• The Lean–Six Sigma merge
After completing this chapter, readers should have
a basic understanding of Lean tools, techniques, and phi-
losophy. This background should help them recognize how
Lean may be used in their organizations and enable them
to employ its tools and techniques to facilitate continuous
improvement.
Healthcare Operat ions Management256
Next, PN decentralized management of each patient to his or her local clinic. This step ensures that each patient receives personalized care and attention because the drugs are ordered by the patient’s primary doctor.
Specific Lean tools used in the improvement process include visual man-agement and standardization for orders, poka-yoke to limit errors, standardized work protocols for the triage of phone calls, and kaizen (introduced in chapter 4) to improve practices in the system. In addition, a consistent formal education program was deployed to help reduce these types of issues in the future.
These improvements helped PN increase the INR percentage to an in-range standing of higher than 70 percent. The average cost to administer the medication per patient per year decreased from a baseline measure of $1,300 to an average $442 per patient per year. Finally, the hospital admission rate of patients using warfarin decreased from 15.9 percent to 11.2 percent.
Source: Trajano, Mattson, and Sanford (2011).
What Is Lean?
As described in chapter 2, Lean production was developed by Taiichi Ohno, Toyota’s chief of production after World War II. The Toyota Production System (TPS) was studied by researchers at Massachusetts Institute of Technology and documented in the book The Machine That Changed the World (Womack, Jones, and Roos 1990). The Lean system originated from just-in-time production and became widely adopted in many manufacturing operations. Lean spread quickly to healthcare organizations because the removal of waste in the system has been shown to improve the clinical measure of safety (Caldwell, Brexler, and Gillem 2005; Chalice 2005; Spear 2005).
Whereas Six Sigma, total quality management, and continuous quality improvement create customer value by eliminating defects, Lean creates seam-less flow to the customer by eliminating waste. Although Six Sigma and Lean are different programs, their methodologies, tools, and outcomes are similar. Both have Japanese roots, as evidenced by the terminology associated with them, and they use many of the same tools and techniques.
TPS, or the Lean Production House (exhibit 10.1), is built on a foun-dation of stability and standardization. The pillars of the house represent the systems that create value for the customer (the roof of the house). The left side of the structure represents producing what you need just in time for the customer. To execute this model correctly, the system must remove waste. The right side of the structure represents automation, or designing the system to stop when defects are produced and remove them. The middle section is the human factor that links the two systems. The ultimate goal is to produce as much value for the customer as possible.
Chapter 10: The Lean Enterpr ise 257
A Lean organization is focused on eliminating all types of waste. Like Six Sigma, Lean has been defined as a philosophy, methodology, and set of tools. The Lean philosophy is to produce only what is needed, when it is needed, and with no waste. The Lean methodology begins by examining the system or process to deter-mine where value is added and where it is not; steps in the process that do not add value are eliminated, and those that do add value are optimized. Lean tools include value stream mapping, the five Ss, spaghetti diagrams, kaizen events, kanbans, rapid changeover (originating with the single-minute exchange of die), heijunka, jidoka, and standardized work, all of which are explored in more detail later.
Types of Waste
In Lean, waste is called muda, which comes from the Japanese term for waste. Many types of waste are found in organizations. As an engineer at Toyota after
Lean• Flow• Heijunka• Takt time• Pull system• Kanban• Visual order
(5S)• Robust
process• Involvement
Jidoka• Poka-yoke• Visual order
(5S)• Problem
solving• Abnormality
control• Separate
human and machine work
• Involvement
Standardized work, 5S, jidoka
TPM, heijunka, kanban
Stability
Involvement• Standardized
work• 5S• TPM• Kaizen circles• Suggestions• Safety activities
Goal Customer Focus
• Takt, heijunka• Involvement, Lean design, A3 thinking
Standardized work, kanban, A3 thinking Visual order (5S)Standardization
EXHIBIT 10.1 Lean Production House
Source: Adapted from Pascal (2007).
Note: 5S = the five Ss of workplace practice; TPM = Toyota’s production method, Toyota Production System.
Healthcare Operat ions Management258
World War II, Ohno created TPS to eliminate waste and inefficiencies in the company’s production system (Economist 2009). Since that time, these wastes have been categorized and reinterpreted as follows for services and healthcare:
• Overproduction—producing more than is demanded or producing before the product is needed to meet demand. Printing reports and preparing meals when they are not needed are examples of overproduction in healthcare.
• Waiting—time during which value is not being added to the product or service. Waiting in healthcare can refer to either the patient sitting idle in a waiting room or the provider waiting for a patient to arrive. When waiting occurs, the resources in the system are not productive or adding value to the end customer in the system.
• Transportation—unnecessary travel of the primary product in the system. In healthcare, transport is so common that the word describes an entire department, whose staff are typically called to move patients in clinics and hospitals to different areas of the facility. Other forms of transportation include bringing equipment and supplies to various locations.
• Inventory—holding or purchasing raw materials, work in process (WIP), and finished goods that are not immediately needed. In healthcare, wasted inventory includes supplies and pharmaceuticals. Too much inventory costs money and limits the organization’s ability to be profitable. In addition, the probability of having outdated drugs on-site increases, creating a greater risk to patients.
• Motion—actions of providers or operators that do not add value to the product (including repetitive motion that causes injury). In healthcare, wasted motion includes unnecessary travel of the service provider to obtain supplies or information.
• Overprocessing—unnecessary processing, or steps and procedures that do not add value to the product or service. Numerous examples of overprocessing in healthcare relate to record keeping and documentation. Many computerized provider order entry systems also require overprocessing to work smoothly.
• Defects—production of a part or service that is scrapped or requires rework. In healthcare, defect waste ranges from mundane errors, such as misfiling documents, to serious errors resulting in the death of a patient. The Joint Commission (2016) classifies catastrophic defects that lead to death or serious injury due to mistakes as sentinel events.
Effective Lean systems focus on eliminating all waste through continuous improvement.
Chapter 10: The Lean Enterpr ise 259
Kaizen
Kaizen is the Japanese term for “change for the better,” or continuous improvement. Kaizen has become the vehicle by which Lean systems adjust and improve. The philosophy of kaizen involves all employees making sug-gestions for improvement and then implementing those suggestions quickly. Because Lean systems target removing waste, opportunity to improve should occur immediately and perpetually.
Kaizen is based on the assumptions that everything can be improved and that many small incremental changes result in an improved system. Absent kaizen, organizations generally operate under the maxim, “If it isn’t broken, leave it alone.” Those that have adopted a kaizen philosophy believe, “Even if it isn’t broken, it can be improved.” An organization that does not focus on continuous improvement is unable to compete with those that continuously improve.
Kaizen can be both a general philosophy of improvement centering on the entire system or value stream and a specific improvement technique for a particular process. The kaizen philosophy of continuous improvement consists of five basic steps:
1. Specify value. Identify activities that provide value from the customer’s perspective.
2. Map and improve the value stream. Determine the sequence of activities or current state of the process and the desired future state. Eliminate non-value-added steps and other waste.
3. Facilitate flow. Enable the process to progress as smoothly and quickly as possible.
4. Allow for pull. Enable the customer to derive products or services.5. Enable perfection. Repeat the process to ensure a focus on continuous
improvement.
The kaizen philosophy is supported by the various tools and techniques of Lean.
Value Stream Mapping
A value stream map is a big-picture view of how a system transforms supplies into finished goods for the customer. Effective value stream maps include all of the steps in the process—both the value-adding and the non-value-adding steps—and their related measurements in producing and delivering a product or service. Both information processing and transformational processing steps are included in a value stream map.
KaizenContinuous improvement based on the beliefs that everything can be improved and that incremental changes result in an enhanced system.
Value stream mapAn overview of how a system transforms supplies into finished goods for the customer.
Healthcare Operat ions Management260
The value stream map shows process flow from a systems perspective and can help in determining how to measure and improve the system or process of interest. Value stream mapping enables the organization to focus on the entire value stream rather than just a specific step or piece of the stream. Without a view of the entire stream, individual parts of the system tend to be optimized according to the needs of those parts, and the resulting system is suboptimal. This short-sightedness occurs frequently in healthcare organizations that are separated by departments. One department, such as lab or X-ray, may make a decision that helps its own processes but has an adverse impact on other areas of the organization, such as the operating rooms or emergency department.
Value stream mapping in healthcare is typically performed from the perspective of the patient, where the goal is to optimize her journey through the system. Information, material, and patient flows are captured in the value stream map. Each step in the process is classified as value-added or non-value-added. Value-added activities are those that change the item being worked on in some way that the customer desires. Using the value stream methodology, value is classified in terms of the following questions:
• Does the patient care about the activity?• Does the activity transform the end product in some way?• Is the activity performed correctly the first time?
If all three questions cannot be answered in the affirmative, the activity is con-sidered non-value-added and should be removed from the system.
Non-value-added activities can be further classified as necessary or unnec-essary. An example of a necessary non-value-added activity that organizations must perform is payroll. Payroll activities do not add value for customers, but employees must be paid. Activities that are classified as non-value-added and unnecessary should be eliminated. Activities that are necessary but non-value-added should be examined to determine if they can be made unnecessary and eliminated. Value-added and necessary non-value-added activities are candidates for improvement and waste reduction. The value stream map enables organiza-tions to see all of the activities in a value stream and focus their improvement efforts (Rother and Shook 1999).
A common measurement for the progress of Lean initiatives is percent value added. The total time for the process to be completed is also measured. These metrics can be captured by measuring the time a single item, customer, or patient spends to complete the entire process. At each step in the process, the value-added time is measured using the following ratio:
% Value added = × 100.Value-added time
Total time in system
Chapter 10: The Lean Enterpr ise 261
The goal of Lean is to increase percent value added by increasing this ratio. Many processes have a percent value added of 5 percent or less. Best-in-class value-added time is often 20 percent or less.
Value streams help organizations focus on flow and not on waiting. Value streams with low value-added percentages are often full of wait times. Traditional healthcare processes involving several departments having less than 1 percent total value-added time are not uncommon.
Once the value stream map is generated, kaizen activities can be identi-fied that allow the organization to increase the percent-value-added time and employ resources in the most effective manner possible.
Vincent Valley Hospital and Health System Value Stream MappingVincent Valley Hospital and Health System (VVH) has identified its birthing center as an area in need of improvement and is using Lean tools and techniques to accomplish its objectives. The goals for the Lean initiative are to decrease costs and increase patient satisfaction. Project management tools (chapter 5) are used to ensure success.
VVH has formed a team to improve the operations of the birthing center. The team consists of the manager of the birthing unit (the project manager), two physicians, three nurses (one from triage, one from labor and delivery, and one from postpartum), and the manager of admissions. All team members have been trained in Lean tools and techniques. They begin the project by developing a high-level value stream map over the course of several weeks (exhibit 10.2). In it, the team maps patient and information flows in the birthing center, and it collects data related to staffing type and level as well as length of time for the various process steps. The high-level value stream map helps the team decide where to focus its efforts; it then develops a plan for the coming year on the basis of the opportunities identified.
Additional Measures and ToolsTakt TimeTakt is a German word meaning rhythm or beat. It is often associated with the rhythm set by a conductor to ensure that the orchestra plays in unison. Takt time determines the speed with which customers must be served to satisfy demand for the service. The calculation is as follows:
Takt time =
.
Takt timeThe speed with which customers must be served to satisfy demand for the service.Available work time/Day
Customer demand/Day
Healthcare Operat ions Management262
EXH
IBIT
9.2
VV
H B
irth
ing
Cent
er V
alue
Str
eam
Map
NO
TE
: Cre
ated
with
eV
SM s
oftw
are
from
Gum
shoe
KI,
Inc
., a
Mic
roso
ft V
isio
add
-on.
Nur
ses’
tim
e
spen
t on
non-
pa
tien
t car
e
Long
wai
t af
ter c
lear
ed
to d
isch
arge
Room
s no
t av
aila
ble
Room
s un
avai
labl
e
Vagi
nal v
s.
C-se
ctio
n
Inco
rrec
t pa
tien
t for
ms
Slow
tu
rnar
ound
Food
qu
alit
y
Educ
atio
n la
te
LOS
Radi
olog
y La
b H
ouse
– ke
epin
g Ph
arm
acy
Supp
lies
Ane
sthe
s-
iolo
gy
Soci
al
serv
ices
D
iete
tics
Port
er
Kitc
hen
Sta
biliz
e
Pati
ents
A
dmit
ting
# #/
da
y
%
%
min
%
min
min
6
30 –
90
min
0 –
2 hr
1 –
3 hr
1 –
3 hr
1 –
60 h
r
rh 5 – 1 rh 8 – 1
rh 3 rh 08 – 02
30
82
85
20
90
20
na
FTEs
Firs
t tim
e co
rrec
t
Arr
ival
rate
Pati
ent
sati
sfac
tion
On-
tim
e de
liver
y
Chan
ge-
over
tim
e
Cycl
e ti
me
Wai
t tim
e
Tria
ge
# %
%
hrs
%
min
min
20
88
90
1 92
25
45
FTEs
Firs
t tim
e co
rrec
t Pa
tien
t sa
tisf
acti
on
On-
tim
e de
liver
y
Chan
ge-
over
tim
e
Cycl
e ti
me
Wai
t tim
e
Labo
r and
D
eliv
ery
# %
%
hrs
%
hrs
hrs
90
90
ns
na
87
5 4
FTEs
Firs
t tim
e co
rrec
t Pa
tien
t sa
tisf
acti
on
On-
tim
e de
liver
y
Chan
ge-
over
tim
e
Cycl
e ti
me
Wai
t tim
e
Post
part
um
# %
%
hrs
%
hrs
hrs
75
80
85
50
75
5 4
FTEs
Firs
t tim
e co
rrec
t Pa
tien
t sa
tisf
acti
on
On-
tim
e de
liver
y
Chan
ge-
over
tim
e
Cycl
e ti
me
Wai
t tim
e
Dis
char
ge
# %
%
hrs
%
hrs
hrs
10
83
85
2 75
2 na
FTEs
Firs
t tim
e co
rrec
t Pa
tien
t sa
tisf
acti
on
On-
tim
e de
liver
y
Chan
ge-
over
tim
e
Cycl
e ti
me
Wai
t tim
e
EXH
IBIT
10.
2 V
VH
Bir
thin
g Ce
nter
Val
ue S
trea
m M
ap
Not
e: C
reat
ed w
ith
eVSM
sof
twar
e fr
om G
umsh
oeK
I, In
c., a
Mic
roso
ft V
isio
add
-on.
FTE
= fu
ll-ti
me
equi
vale
nt; L
OS
= le
ngth
of s
tay;
na
= n
ot a
pplic
able
.
Chapter 10: The Lean Enterpr ise 263
Cycle time is the time needed for a system to accomplish a task in that system. Cycle time for a system is equal to the longest task-cycle time in that system. Cycle time is often referred to as the “drip rate” of the system, as with a leaky faucet: The cycle time is the rate at which water drips from the faucet. In a perfect Lean system, cycle time and takt time are equal. If cycle time is greater than takt time, demand is not satisfied and customers or patients are required to wait. If cycle time is less than takt time in a manufacturing environment, inventory is generated; in a service environment, resources are underutilized. In a Lean system, the rate at which a product or service can be produced is set by customer demand, not by the organization’s ability (or inability) to supply the product or service.
Throughput TimeThroughput time is the time needed for an item to complete the entire pro-cess. It includes waiting time and transport time as well as actual processing time. In a healthcare clinic, for example, throughput time is the total time the patient spends at the clinic, starting when he walks through the door and end-ing when he walks out. It includes not only the time the patient is interacting with a clinician but also time spent idle in the waiting and examining rooms. In a perfectly Lean system, no waiting time is experienced, and throughput time is thus minimized. In most instances, throughput time is dictated by the non-value-added activities and not by the provider–patient interaction.
Riverview Clinic Timing IssuesVVH’s Riverview Clinic has collected the data shown in exhibit 10.3 for a typical patient visit. Here, the physician exam and consultation involves the longest task time, 20 minutes; therefore, the cycle time for this process is 20 minutes. Assuming that the physician is available to work with the patients and not performing other tasks, every physician should be able to “output” one patient from this process every 20 minutes. However, the throughput time is equal to the total amount of time a patient spends in the system:
3 + 15 + 2 + 15 + 5 + 10 + 20 = 70 minutes.
The available work time per physician day is 5 hours (Riverview Clinic physicians work 10 hours per day, but only 50 percent of that time is spent with patients), the clinic has 8 physicians, and 100 patients are expected at the clinic every day:
Takt time = = 0.4 physician hours/patient
= 24 physician minutes/patient.
Cycle timeThe time required to accomplish a task in a system.
Throughput timeThe time required for an item to complete the entire process, including waiting time and transport time.
8 physicians × 5 hours/day
100 patients/day
Healthcare Operat ions Management264
Therefore, to meet demand, the clinic needs to serve one patient every 24 minutes. Because cycle time (20 minutes) is less than takt time (24 minutes), the clinic can meet demand.
Assuming that (1) patient check-in is necessary but non-value-added and (2) both the nurse preliminary exam (5 minutes) and the physician exam and consultation (20 minutes) are value-added tasks, the value-added time for this process is
5 + 20 = 25 minutes
and the percent-value-added time is
25 minutes ÷ 70 minutes = 36%.
This example assumes that all of the steps of the check-in process are value-added. The reality is that many of the steps we perform in any given activ-ity in a process are non-value-added. A Lean system works toward decreasing throughput time and increasing percent-value-added time. The tools discussed in the following sections can aid in achieving these goals as building blocks to the overall Lean system.
Five SsThe five Ss are workplace practices that constitute the foundation of other Lean activities; the Japanese words for these practices all begin with S. The five Ss essentially are ways to ensure a clean and organized workplace. Often,
Patientcheck-in
3 minutes
Move toexamining
room2 minutes
Wait15
minutes
Wait15
minutes
Nurse doespreliminary
exam5 minutes
Wait10
minutes
Physicianexam and
consultation20 minutes
Visitcomplete
Note: Created with Microsoft Visio.
EXHIBIT 10.3 Riverview
Clinic Cycle, Throughput,
and Takt Times
Chapter 10: The Lean Enterpr ise 265
they are seen as obvious and self-evident—a clean and organized workplace is more efficient than a cluttered area is. However, without a continuing focus on these five practices, workplaces often become disorganized and inefficient.
The five practices, with their Japanese names and the English terms typically used to describe them, are as follows:
• Seiri (sort)—Separate necessary from unnecessary items, including tools, parts, materials, and paperwork, and remove the unnecessary items.
• Seiton (set in order)—Arrange the necessary items neatly, providing visual cues to where items should be placed.
• Seiso (shine)—Clean the work area.• Seiketsu (standardize)—Standardize the first three Ss so that cleanliness
is maintained.• Shitsuke (sustain)—Ensure that the first four Ss continue to be
performed on a regular basis.
Many hospitals and healthcare organizations have adopted a sixth S in the system, safety, considered paramount in the design of the sustainable process (EPA 2011).
Adopting the five (or six) Ss is often the first step an organization takes in its Lean journey because so much waste can be eliminated by establishing and maintaining an organized and efficient workplace. An effective five S program requires that the organization build discipline to continue the efforts in the long term. If an organization cannot sustain a simple mechanism to keep an area clean and organized, it will struggle with more complex systems. Five S systems can be easy to build but are difficult to maintain. Exhibit 10.4 displays a form for scheduling regular audits to make sure the system is sustainable.
Spaghetti DiagramA spaghetti diagram is a visual representation of the movement or travel of materials, employees, or customers. In healthcare, a spaghetti diagram is often used to document or investigate the movements of caregivers or patients. Typically, the patient or caregiver spends a significant amount of time moving from place to place and often backtracks. A spaghetti diagram (exhibit 10.5) helps find and eliminate wasted movement in the system.
Kaizen Event or BlitzA kaizen event or blitz (sometimes referred to as a rapid process improvement workshop) is a focused, short-term project aimed at improving a particular process. A kaizen event is usually performed by a cross-functional team of eight to ten people, always including at least one person who works with or in the
Spaghetti diagramA visual representation of the movement or travel of materials, employees, or customers.
Kaizen eventA focused, short-term project aimed at improving a particular process.
Healthcare Operat ions Management266
5S A
UD
IT F
OR
MAr
ea:
D
ate:
Sort
Yes
No
Dat
e co
rrec
ted
Com
men
ts:
1S
Onl
y re
quir
ed it
ems
for t
he w
ork
area
are
pre
sent
Hig
h us
e it
ems
are
read
ily a
vaila
ble
App
ropr
iate
item
s ar
e gr
oupe
d to
geth
er
Set i
n or
der
Com
men
ts:
2S
Mat
eria
ls p
rope
rly
stor
ed s
o th
ey c
an b
e us
ed a
nd re
turn
ed e
asily
All
floor
loca
tion
s an
d w
ork
area
s cl
earl
y m
arke
d w
ith
tape
or p
aint
Cab
inet
s, s
helv
es a
nd c
onta
iner
s ar
e la
bele
d an
d or
gani
zed
Shin
eCo
mm
ents
:
3S
Flo
ors
and
enti
re a
rea
are
clea
n an
d fr
ee o
f clu
tter
All
equi
pmen
t is
clea
n, fr
ee fr
om le
aks,
chi
ps a
nd s
oile
d sp
ots
All
wor
k su
rfac
es, c
abin
ets
and
shel
ves
are
clea
n an
d fr
ee fr
om c
lutt
er
Stan
dard
ize
Com
men
ts:
4S
App
ropr
iate
inve
ntor
y on
han
d A
ppro
pria
te n
umbe
r of e
ach
type
of e
quip
men
t pre
sent
Pic
ture
s an
d/or
labe
ls a
re s
tand
ardi
zed
and
visi
ble
Sust
ain
Com
men
ts:
5S
Life
Saf
ety
Code
requ
irem
ents
com
plia
nt L
abel
s, s
igna
ls a
nd a
rea
dem
arca
tion
s ar
e m
aint
aine
d 5
S co
rrec
tive
act
ions
com
plet
ed fr
om la
st a
udit
Imm
edia
te C
orre
ctiv
e Ac
tion
(s):
Not
e: 5
S =
the
five
Ss o
f wor
kpla
ce p
ract
ices
.
EXH
IBIT
10.
4Sa
mpl
e 5S
Aud
it F
orm
, Vet
eran
s H
ealt
h Ad
min
istr
atio
n, M
inne
apol
is
Chapter 10: The Lean Enterpr ise 267
process. The rest of the team should include personnel from other functional areas and even nonemployees with an interest in improving the process. In healthcare organizations, staff, nurses, doctors, and other professionals, as well as management personnel from across departments, should be represented.
Typically, a kaizen event consists of the following steps, based on the plan-do-check-act improvement cycle of Deming and Juran (see chapter 2):
1. Determine and define the objective(s).2. Determine the current state of the process by mapping and measuring
the process. Measurements are related to the desired objectives and may include such factors as cycle time, waiting time, WIP, throughput time, and travel distance.
3. Determine the requirements of the process (takt time), develop target goals, and design the future state or ideal state of the process.
4. Create a plan for implementation, including who, what, when, and so on.5. Implement the improvements.6. Check the effectiveness of the improvements.7. Document and standardize the improved process.8. Report the results of the event on an A3 reporting form (discussed below).9. Continue the cycle.
Cabinetsunder
counterCabinets
undercounter
4M 134 4M 132
9” � 24” 9” � 24”
9” � 24”9” � 24”
9” � 36”
24” � 9”
9” � 24”
Fridge Cabinet Cabinet
EXHIBIT 10.5 Spaghetti Map for Setting Up Education Room
Healthcare Operat ions Management268
The kaizen event is based on the notion that most processes can be quickly (and relatively inexpensively) improved, in which case it makes sense to “just do it” rather than be paralyzed by resistance to change. A kaizen event is typically one week long and begins with training in the tools of Lean, followed by analysis and measurement of the current process and generation of possible ideas for improvement. By midweek, a proposal for changes to improve the process should be completed. The proposal includes the improved process flow and metrics for determining the impacts of the changes. The proposed changes are implemented and tested during the remainder of the week. At the end of the week, a team reports the results on an A3 reporting form.
The A3 is a summary of the project results presented on a one-page, standard letter size A3 sheet of paper. By the following week, the new process should be in place.
A kaizen event can be a powerful way to quickly and inexpensively improve processes. The results are usually a significantly enhanced process and increased employee pride and satisfaction.
Vincent Valley Hospital and Health System Kaizen EventThe value stream map developed for the VVH birthing center highlights the fact that nursing staff spend a significant amount of time on activities not related to actual patient care. This situation has resulted not only in dissatis-fied patients, physicians, and nurses but also in increased staffing costs to the hospital. A kaizen blitz is planned to address this problem in the postpartum area of the birthing center.
The nursing administrator is charged with leading the kaizen event. She forms a team consisting of a physician, a housekeeper, two nurses’ assistants, and two nurses. On Monday morning, the team begins the kaizen event with four hours of Lean training. That afternoon, team members develop a spaghetti diagram for a typical nurse and begin collecting data related to the amount of time nursing staff spend on various activities. They also collect historical data on patient load and staffing levels.
On Tuesday morning, the team continues to collect data. In the after-noon, its members analyze the data and note that nursing staff spend only 50 percent of their time in actual patient care. A significant amount of time—one hour per eight-hour shift—is spent locating equipment, supplies, and informa-tion. The team decides that a 50 percent reduction in this time measure is a reasonable goal for the kaizen event.
On Wednesday morning, the team performs a root-cause analysis to determine the reasons nursing staff spend so much time locating and moving equipment and supplies. They find that one of the major causes is general disorder in the supply and equipment room and in patient rooms.
On Wednesday afternoon, the team organizes the supply and equipment room. Team members begin by determining what supplies and equipment are
Chapter 10: The Lean Enterpr ise 269
necessary to performing their work and removing those that are unnecessary. Next, they organize the supply and equipment room by identifying which items are needed most frequently and locating those items together. All storage areas are labeled, and specific locations for equipment are designated visually. White boards are installed to enable the tracking and location of equipment. The team also develops and posts a map of the room so that the location of equipment and supplies can be easily viewed.
On Thursday, the team works on reorganizing all of the patient rooms, standardizing the layout and location of items in each one. First, team members observe the activity taking place in one of the patient rooms and determine the equipment and supply needs of physicians and nurses. All nonessential items are removed, creating more space. Additionally, rooms are stocked with supplies used on a routine basis to reduce trips to the central supply room. A procedure is also established to restock supplies daily.
On Friday morning, the kaizen team again collects data on the amount of time nursing staff spend on various activities. It finds that after implement-ing the changes, the time nursing staff spent locating and moving supplies and equipment has been reduced to approximately 20 minutes in an eight-hour shift, a 66 percent reduction. Friday afternoon is spent documenting the kai-zen event and putting systems in place to ensure that the new procedures and organizational approach are maintained.
Standardized WorkStandardized work is an essential part of Lean that provides the baseline for continuous improvement. Standardized work refers to the methods by which a process is executed. All effective standardized work procedures include written documentation of the precise way every step in a process should be performed. It should not be seen as a rigid system of compliance, but rather as a means of communicating and codifying current best practices in the organization. Standardized work is critical to developing an effective Lean system as it rep-resents the baseline against which all future improvements will be measured.
All relevant stakeholders of the process should be involved in establishing standardized work. Standardizing work in this way assumes that the people most intimately involved with the process have the most knowledge of how to best perform the work. Such involvement can promote employee buy-in, owner-ship of the process, and responsibility for improvement. Clear documentation and specific work instructions ensure that variation and waste are minimized.
Standardized work should be seen as a step on the road to improvement. It allows doctors and nurses to perform activities at their licensure level more often than in nonstandard work because basic business processes run effectively using standardized work (Lowe et al. 2012). This allowance to work at top of license then leads to standardized measures that lead to cost-effectiveness and improvement of patient outcomes.
Standardized workDocumentation of the precise way in which every step in a process should be completed.
Healthcare Operat ions Management270
In the healthcare industry, examples of standardized work include treat-ment protocols and the establishment of care paths. (Care paths are also examples of evidence-based medicine, which is explored in chapter 3.) A care path is “an optimal sequencing and timing of interventions by physicians, nurses, and other staff for a particular diagnosis or procedure, designed to minimize delays and resource utilization and at the same time maximize the quality of care” (Wheelwright and Weber 2004). Care paths define and docu-ment specifically what should happen to a patient the day before surgery, the day after surgery, and on following postsurgical days.
As part of an overall program to improve practices and reduce costs, Massachusetts General Hospital developed and implemented a care path for coronary artery bypass graft (CABG) surgery. The care path was not intended to dictate medical treatment but to standardize procedures as much as possible to reduce variability and improve the quality of outcomes (Wheelwright and Weber 2004).
The team that developed the care path was composed of 25 participants representing the various areas involved in treatment. It spent more than a year developing the initial care path. Because of its breadth of inclusion and applicability, resistance to implementation was minimal. The care path resulted in an average length of stay reduction of 1.5 days, and significant cost savings were associated with that reduction. After the successful implementation of the CABG surgical care path, Massachusetts General established more than 50 additional care paths related to surgical procedures and medical treatments (Wheelwright and Weber 2004).
Standardized work processes can be used in clinical, support, and admin-istrative operations of healthcare organizations. The development and docu-mentation of standardized processes and procedures can be a powerful way to engage and involve everyone in the organization in continuous improvement.
Jidoka and AndonIn Lean systems, jidoka refers to the ability to stop the process in the event of a problem. The term stems from the weaving loom invented by Sakichi Toyoda, founder of the Toyota Group. The loom stopped itself if a thread broke, eliminating the possibility that defective cloth would be produced.
Jidoka prevents defects from being passed from one step in the system to the next and enables the swift detection and correction of errors. If the system or process is stopped when a problem is found, everyone in the process works quickly to identify and eliminate the source of the error.
In ancient Japan, an andon was a paper lantern used as a signal; in a Lean system, an andon is a visual or audible signaling device used to indicate a problem in the process. Andons are typically used in conjunction with jidoka.
In his book The Checklist Manifesto, Atul Gawande (2009) highlights the benefits that hospitals gain by using simple checklists prior to inducing a
Care pathA sequence of best practices for healthcare staff to follow for a diagnosis or procedure, designed to minimize waste and maximize quality of care.
JidokaThe ability to prevent defects by stopping a process when an error occurs.
AndonA visual or audible signaling device used to indicate a problem in the process, typically used in conjunction with jidoka.
Chapter 10: The Lean Enterpr ise 271
patient into an anesthetized state for surgery. These checklists are a mechanism to make sure everyone in the surgical suite is in agreement on the details of the patient and procedure about to take place, and they give the surgical team a chance to “stop the line” if protocol has not been properly followed.
Virginia Mason Medical Center implemented a jidoka-andon system called the Patient Safety Alert System (Womack et al. 2005). If a caregiver believes something is not right in the care process, not only can she stop the process but she is obligated to do so. The person who has noticed the problem alerts the patient safety department. The appropriate process stakeholders or relevant managers move immediately to determine and correct the root cause of the problem. After two years, the number of alerts per month rose from 3 to 17, enabling Virginia Mason to correct most problems in the process before they became more serious. The alerts are primarily related to systems issues, medication errors, and problems with equipment or facilities.
KanbanKanban is a Japanese term for signal. A kanban uses containers of a certain size to signal the need for more production or the movement of product. The customer indicates that he wants a product, a kanban is released to the last operation in the system to signal the customer demand, and that station begins to produce the product in response. As incoming material is consumed at the last workstation, another kanban is emptied and sent to the previous worksta-tion to signal that production should begin at that station. The empty kanbans go backward through the production system to signal the need to produce in response to customer demand (see exhibit 10.6). This system ensures that production is only undertaken in response to customer demand, not simply because production capacity exists.
In a healthcare environment, kanbans can be used for supplies or phar-maceuticals to signal the need to order more. For example, a pharmacy would
KanbanA visual signal that triggers the movement of inventory or product in a system.
EmptyKanban
EmptyKanban
FullKanban
FullKanban
Customer OrderTask 2
Workstation 2Task 1
Workstation 1
Note: Created with Microsoft Visio.
EXHIBIT 10.6 Kanban System
Healthcare Operat ions Management272
have two kanbans; when the first kanban is emptied, this signals the need to order more of the drug and an order is placed. The second kanban is emptied while waiting for the order to arrive. Ideally, the first kanban is received from the supplier at the point that the second kanban is empty and the cycle con-tinues. The size of the kanbans is related to demand for the pharmaceutical during lead time for the order. The number and size of the kanbans determine the amount of inventory in the system.
In a healthcare environment, kanbans can be used to control the flow of patients, ensuring continuous movement. For example, for patients needing both an echocardiography (echo) procedure and a computed tomography (CT) scan, where the echo procedure is to be performed before the CT scan, the CT scan could pull patients through the process. When a CT is performed, a patient is taken from the pool of patients between CT and echo. A kanban (signal) is sent to the echo station to indicate that another patient should receive an echo (see exhibit 10.7). This method keeps a constant pool of patients between the two processes. The patient pool should be large enough to ensure that the CT is busy even when disturbances in the echo process occur. However, its size must be balanced with the need to keep patients from waiting for long periods. Eventually, in a Lean system, the pool size is reduced to one.
Rapid ChangeoverThe rapid changeover, or single-minute exchange of die (SMED) system, was developed by Shigeo Shingo (1985) of Toyota. Originally, it was used by manufacturing organizations to reduce changeover or setup time—the time between producing the last good part of one product and the first good part
PatientsPatients CTEcho
langiSlangiS
EXHIBIT 10.7 Kanban for
Echo/CT Scan
Note: Created with Microsoft Visio. CT = computed tomography; echo = echocardiogram.
Chapter 10: The Lean Enterpr ise 273
of a different product. Currently, the technique is used to reduce setup time for both manufacturing and services. In healthcare, the SMED system trans-lates better as rapid changeover. In healthcare environments, setup is the time needed, or taken, between the completion of one procedure and the start of the next or between the checkout of one patient and the arrival of a new patient.
The rapid changeover technique consists of three steps:
1. Separating internal activities from external activities2. Converting internal setup activities to external activities3. Streamlining all setup activities
Internal activities are those that must be performed in the system; they cannot be done offline. For example, cleaning an operating room (OR) prior to the next surgery is an internal setup activity; it cannot be completed outside the OR. However, organizing the surgical instruments for the next surgery is an external setup, as it can be completed outside the OR to allow for speedier changeover of the OR.
Setup includes finding and organizing instruments, gathering supplies, cleaning rooms, and obtaining paperwork. In the healthcare environment, rapid changeover can help alleviate surgery suite backlogs and cancelations because the room can be turned over quickly and the surgery teams can maximize the amount of time they are in surgery (AHRQ 2007).
To streamline activities, Lean teams must look for opportunities to per-form tasks in parallel and find ways to automate the process. For example, many manufacturers have facilitated the turnover of surgery rooms by manufacturing disposable sleeves that cover all of the lights and fixtures in the room. Instead of having to scrub all of those fixtures, a team simply replaces the sleeves.
Heijunka and Advanced AccessHeijunka is a Japanese term meaning to make flat and level. It refers to eliminating variations in volume and variety of production to reduce waste. In healthcare environments, making flat and level often means determining how to level out patient demand. Producing goods or services at a steady rate allows organizations to be increasingly responsive to customers and make optimal use of their own resources. In healthcare, advanced access provides a good example of the benefits of heijunka.
Advanced-access scheduling reduces the time between scheduling an appointment for care and the actual appointment. It is based on the principles of Lean and aims for swift, even patient flow through the system. Heijunka helps reduce the wait time for appointments, decrease patient no-show rates, and improve both patient and staff satisfaction. As a result, clinics increase their revenue and reduce administrative costs because fewer patients are rescheduled.
HeijunkaThe process of eliminating variations in volume and variety of production to reduce waste.
Healthcare Operat ions Management274
Although the benefits of advanced access are valuable, implementation can be difficult because the concept challenges established practices and beliefs. However, if the delay between making an appointment and the actual appoint-ment is relatively constant, implementing advanced access should be feasible.
Centra Health, a multisite primary care organization, was able to reduce access time to three days or less. As a result, patient satisfaction increased from 72 percent to 85 percent, and continuity of care was significantly increased, such that 75 percent of visits occurred with a patient’s primary physician, compared to 40 percent prior to advanced access. The most significant issue encountered was the greater demand for popular clinicians than for others and the need to address this inequity on an ongoing basis (Murray et al. 2003).
Successful implementation of advanced access requires that supply and demand be balanced. Accurate estimates of both supply and demand are needed, backlog must be reduced or eliminated, and the variety of appointment types needs to be minimized. Once supply and demand are known, demand profiles may need to be adjusted and the availability of bottleneck resources increased (Murray and Berwick 2003). The Institute for Healthcare Improvement (2006) offers extensive online resources to aid healthcare organizations in implement-ing advanced access, and chapter 12 discusses the concept in more detail.
The Merging of Lean and Six Sigma Programs
Many organizations now combine the philosophies and tools of Lean and Six Sigma into Lean Six Sigma (George 2002). Although proponents of Lean or Six Sigma might tout their differences and champion one over the other, the two methods are complementary, and combining them can be an effective approach to improvement.
Exhibit 10.8 provides a classic illustration of how the two continuous improvement programs may be used together. Here, the water represents waste in the system. The high water (waste) buffers the rocks so the boat can move down-stream without encountering any issues. In healthcare systems, this waste often shows up in one of two forms: excess supplies and inventory or too much demand on the system. This buffering might seem helpful, because once the water is removed, the rocks become exposed, making travel dangerous. But the rocks represent major issues in our systems, such as sentinel events and excessive overtime paid to nurses and other staff. To sail the boat without crashing (encountering issues), the rocks (problems) must be eliminated (by removing variance in the system). Perhaps too much overtime is being paid to the staff in the surgical suite of a hospital. Analysis finds that staff are spending excess time looking for equipment, which delays surgeries and forces the overtime. To get the boat to sail smoothly, the problems of looking for equipment must be reduced and removed.
Chapter 10: The Lean Enterpr ise 275
The Lean system focuses on eliminating waste and streamlining flow. In the previous example, the waste in the system was identified as excessive idle time as a result of waiting for the equipment, which may lead to hiring extra people to make sure the equipment reaches the OR suite on time. The Six Sigma program focuses on creating value to the customer, eliminating defects, and reducing variation. It identifies the reasons that equipment arrives late to the OR and systematically reduces and removes those sources of variance. Both Lean and Six Sigma are ultimately focused on continuous improvement of any system.
Excess hides problems
Reducing excess makes problem visible
Reduce problems/remove variation
EXHIBIT 10.8Lean Six Sigma Approach
Healthcare Operat ions Management276
The Six Sigma process, featuring the define-measure-analyze-improve-control structure, always begins with defining the issues or problems as they relate to the customer. The focus on reducing variance in the eyes of the cus-tomer allows Six Sigma programs to create customer value.
The kaizen philosophy of Lean begins with determining what customers value, followed by mapping and improving the process to achieve flow and pull. Lean thinking enables identification of the areas causing inefficiencies. How-ever, to truly achieve Lean, variation in the processes must be eliminated—Six Sigma helps achieve its elimination. Focusing on the customer and eliminating waste not only results in increased customer satisfaction but also reduces costs and increases the profitability of the organization.
Together, Lean and Six Sigma can provide the philosophies and tools needed to ensure that the organization is continuously improving. Research supports the idea that the implementation of continuous improvement is a gradual addition of skill sets and not the selection of a specific system like Lean or Six Sigma (Belohlav et al. 2010).
Conclusion
Lean systems have been used in many industries to remove inefficiencies and waste related to production of goods and services. Healthcare systems have also adopted Lean to enhance safety and improve the quality of care. The removal of outdated medicines, expired supplies, and clutter makes the environment safer for patients. These simple concepts related to waste reduction work well for most healthcare systems. Lean will continue to be a focal point in healthcare as the pressure mounts to reduce cost. The waste reduction approaches will allow the US healthcare system to be increasingly cost-effective and safe for patients.
Discussion Questions
1. What are the drivers of the healthcare industry’s focus on patient satisfaction and on employing resources in an effective manner?
2. What are the differences between Lean and Six Sigma? The similarities? Would you like to see both applied in your organization? Why or why not?
3. From your own experiences, discuss a specific example of each of the seven types of waste.
4. From your own experiences, describe a specific instance in which standardized work, kanban, jidoka and andon, and rapid changeover would enable an organization to improve its effectiveness or efficiency.
Chapter 10: The Lean Enterpr ise 277
5. Does your primary care clinic have advanced-access scheduling? Should it? To determine supply and demand and track progress, what measures would you recommend to your clinic?
6. Are any drawbacks inherent to Lean Six Sigma? Explain.
Exercises
1. A simple value stream map for patients requiring a colonoscopy at an endoscopy clinic is shown in the graphic below. Assume that patients recover in the same room where the colonoscopy is performed and the clinic has two colonoscopy rooms. What is the cycle time for the process? What is the throughput time? What is the percent value added in this process? If the clinic operates 10 hours a day and demand is 12 patients per day, what is the takt time? If demand is 20 patients per day, what is the takt time? What would you do in the second situation?
20 Min
5 Min 15 Min
15 Min 0 Min
40 Min
10 Min
5 Min30 Min
DischargePatient
recoveryColonoscopy
Patientprep
Patientcheck-inColonoscopy
patients
No./day
ValueDemand
Note: Created with eVSM software from GumshoeKI, Inc., a Microsoft Visio add-on.
2. Draw a high-level value stream map for your organization (or a part of your organization). Pick a part of this map and draw a more detailed value stream map for it. On each map, be sure to identify the information you would need to complete the map and exactly how you might obtain that information. What are the takt and throughput times of your process? Identify at least three kaizen opportunities on your map.
3. For one of the kaizen opportunities listed in exercise 2, describe the kaizen event you would plan if you were the kaizen leader.
References
Agency for Healthcare Research and Quality (AHRQ). 2007. Managing and Evaluating Rapid-Cycle Process Improvements as Vehicles for Hospital System Redesign. AHRQ Publication No. 07-0074-EF, September. Rockville, MD: AHRQ.
Healthcare Operat ions Management278
Belohlav, J. A., L. S. Cook, J. R. Olson, and D. E. Drehmer. 2010. “Core Values in Hospitals: A Comparative Study.” Quality Management Journal 17 (4): 36–50.
Caldwell, C., J. Brexler, and T. Gillem. 2005. Lean-Six Sigma for Healthcare: A Senior Leader Guide to Improving Cost and Throughput. Milwaukee, WI: ASQ Quality Press.
Chalice, R. 2005. Stop Rising Healthcare Costs Using Toyota Lean Production Methods: 38 Steps for Improvement. Milwaukee, WI: ASQ Quality Press.
Dobrzykowski, D. D., K. L. McFadden, and M. A. Vonderembse. 2016. “Examining Pathways to Safety and Financial Performance in Hospitals: A Study of Lean in Professional Service Organizations.” Journal of Operations Management 42–43 (Special Issue): 39-51.
Economist, The. 2009. “Taiichi Ohno.” Published July 3. www.economist.com/node/ 13941150.
Environmental Protection Agency (EPA). 2011. “Lean and Environment Toolkit.” Accessed May 21, 2012. www.epa.gov/ lean/environment/toolkits/environment/ch5.htm.
Gawande, A. 2009. The Checklist Manifesto. New York: Metropolitan Books.George, M. 2002. Lean Six Sigma: Combining Six Sigma Quality with Lean Production
Speed. New York: McGraw-Hill.Institute for Healthcare Improvement. 2006. “Managing Patient Flow: Smoothing OR
Schedule Can Ease Capacity Crunch, Researchers Say.” OR Manager 19 (1): 9–10.Joint Commission, The. 2016. “Sentinel Event Policy and Procedures.” Published January
6. www.jointcommission.org/sentinel_event_policy_and_procedures/.Lowe, G., V. Plummer, A. P. O’Brien, and L. Boyd. 2012. “Time to Clarify—the Value of
Advanced Practice Nursing Roles in Health Care.” Journal of Advanced Nursing 68 (3): 677–85.
Murray, M., and D. M. Berwick. 2003. “Advanced Access: Reducing Waiting and Delays in Primary Care.” Journal of the American Medical Association 290 (3): 332–34.
Murray, M., T. Bodenheimer, D. Rittenhouse, and K. Grumbach. 2003. “Improving Timely Access to Primary Care: Case Studies in the Advanced Access Model.” Journal of the American Medical Association 289 (8): 1042–46.
Pascal, D. 2007. Lean Production Simplified, 2nd edition. New York: Productivity Press.Rother, M., and J. Shook. 1999. Learning to See: Value Stream Mapping to Add Value and
Eliminate Muda. Brookline, MA: Lean Enterprise Institute.Shingo, S. 1985. A Revolution in Manufacturing: The SMED System. Translated by A. Dil-
lon. New York: Productivity Press.Spear, S. J. 2005. “Fixing Health Care from the Inside, Today.” Harvard Business Review
83 (9): 78–91.Wheelwright, S., and J. Weber. 2004. Massachusetts General Hospital: CABG Surgery (A).
Harvard Business Review Case 9-696-015. Boston: Harvard Business School Publishing.
Womack, J. P., A. P. Byrne, O. J. Fiume, G. S. Kaplan, and J. Toussaint. 2005. Going Lean in Health Care. Cambridge, MA: Institute for Healthcare Improvement.
Womack, J. P., D. T. Jones, and D. Roos. 1990. The Machine That Changed the World: Based on the Massachusetts Institute of Technology 5-Million Dollar 5-Year Study on the Future of the Automobile. New York: Rawson Associates.
PART
IVAPPLICATIONS TO CONTEMPORARY HEALTHCARE OPERATIONS ISSUES
CHAPTER
281
PROCESS IMPROVEMENT AND PATIENT FLOW
Operations Management in Action
Cambridge Health Alliance Whidden Hospital in Ever-ett, Massachusetts, is a safety net hospital whose emergency department (ED) was experiencing long waits, inefficient processes, and poor patient satisfac-tion. Its leaders undertook two projects to improve patient flow: an ED facility expansion, and, two years later, a reorganization of patient flow and the estab-lishment of a rapid assessment unit (RAU).
In the period following the ED expansion, sig-nificant negative trends were observed: decreasing Press Ganey patient satisfaction percentiles (–4.1 percentile per quarter), increasing door-to-provider time (+4.9 minutes per quarter), increasing duration of stay (+13.2 minutes per quarter), and increasing percentage of patients leaving without being seen (+0.11 per quarter).
After the RAU was established, significant immediate impacts were observed for door-to-provider time (–25.8 minutes) and total duration of stay (–66.8 minutes). The trends for these indicators further suggested the improvements continued to be significant over time. Furthermore, the negative trends for the Press Ganey outcomes observed after ED expansion were significantly reversed and contin-ued to move in the positive direction after the RAU. The major conclusion from the project team was that the impact of process improvement and RAU imple-mentation is far greater than the impact of renovation and facility expansion.
Source: Sayah et al. (2016).
11OVE RVI EW
At the core of all organizations are their operating sys-
tems. Excellent organizations continuously measure,
study, and make improvements to these systems. This
chapter provides a methodology for measuring and
improving systems using a select set of the tools pre-
sented in the preceding chapters.
The terminology associated with process
improvement can be confusing. Typically, tasks com-
bine to form subprocesses, subprocesses combine
to form processes, and processes combine to form
a system. The boundaries of a particular system are
defined by the activity of interest. For example, the
boundaries of a supply chain system are more encom-
passing than those of a hospital system that is part of
that supply chain.
The term process improvement refers to
improvement at any of these levels, from the task level
to the systems level. This chapter focuses on process
and systems improvement.
Process improvement follows the classic plan-
do-check-act (PDCA) cycle (chapter 9), with the follow-
ing, more specific, key steps:
• Plan: Define the entire process to be improved
using process mapping. Collect and analyze
appropriate data for each element of the process.
• Do: Use a process improvement tool(s) to
improve the process.
• Check: Measure the results of the process
improvement.
(continued)
Healthcare Operat ions Management282
Problem Types
Continuous process improve-ment is essential for organiza-tions to meet the challenges of today’s healthcare environ-ment. The theory of swift and even flow (TSEF) (Schmenner 2001, 2004; Schmenner and Swink 1998) asserts that a pro-cess is more productive as the stream of materials (custom-ers or information) flows more swiftly and evenly. Productiv-ity rises as the speed of flow through the process increases and the variability associated with that process decreases.
Note that these phenomena are not independent. Often, decreasing system variability increases flow, and increasing flow decreases variability. For example, advanced-access (same day) scheduling increases flow by decreas-ing the elapsed time between when a patient schedules an appointment and when she has completed her visit with the provider. Applying this concept of interdependence to patient no-shows, advanced-access scheduling can decrease variability by decreasing the number of no-shows.
Solutions to many of the problems facing healthcare organizations can be found in increasing flow or decreasing variability. For example, a key oper-ating challenge in most healthcare environments is the efficient movement of patients in a hospital or clinic, commonly called patient flow. Various approaches to process improvement can be illustrated using the patient flow problem. Optimizing patient flow through EDs has become a top priority of many hospitals; therefore, the Vincent Valley Hospital and Health System (VVH) example at the end of this chapter focuses on improving patient flow through that organization’s ED.
Another key issue facing healthcare organizations is the need to increase the level of quality and eliminate errors in systems and processes. In other words, variation must be decreased. Finally, increasing cost pressures result in the need for healthcare organizations to improve processes and do so while reducing costs.
The tools and techniques presented in this book are aimed at enabling cost-effective process improvement. Although this chapter focuses on patient flow and elimination of errors related to patient outcomes, the discussion is equally applicable to other types of flow problems (e.g., information, paperwork)
OVE RVI EW (Continued)
• Act to hold the gains: If the process improvement
results are satisfactory, hold the gains (chapter 15).
If the results are not satisfactory, repeat the PDCA cycle.
This chapter discusses the types of problems
or issues faced by healthcare organizations, reviews
many of the operations tools discussed in earlier chap-
ters, and illustrates how these tools can be applied to
process improvement. The relevant tools include the
following:
• Basic process improvement tools
• Six Sigma and Lean tools
• Simulation software
Chapter 11: Process Improvement and Pat ient F low 283
and other types of errors (e.g., billing). Some tools are more applicable to increasing flow and others to decreasing variation, eliminating errors, or improv-ing quality, but all of the tools can be used for process improvement.
Patient Flow
Efficient patient movement in healthcare facilities can significantly improve the quality of care patients receive and substantially improve financial performance. A patient receiving timely diagnosis and treatment has a higher likelihood of obtaining a desired clinical outcome than a patient whose diagnosis and treat-ment are delayed. Because most current payment systems are based on fixed payments per episode of treatment, a patient moving more quickly through a system tends to generate lower costs and, therefore, higher margins.
Patient flow optimization opportunities occur in many healthcare set-tings. Examples include operating suites, imaging departments, urgent care centers, and immunization clinics. Advanced-access scheduling is a special case of patient flow and is examined in depth in chapter 12.
Poor patient flow has several causes; one culprit discovered by many investigators is variability of scheduled demand. For example, if an operating room is scheduled for a surgery but the procedure does not take place at the scheduled time, or it takes longer than scheduled to complete, the rest of the surgery schedule becomes delayed. These delays ripple through the entire hospital, including the ED.
As explained by Eugene Litvak, PhD (2003):
You have two patient flows competing for hospital beds—ICU or patient floor beds.
The first flow is scheduled admissions. Most of them are surgical. The second flow
is medical, usually patients through the emergency department. So when you have a
peak in elective surgical demand, all of a sudden your resources are being consumed
by those patients. You don’t have enough beds to accommodate medical demand.
If scheduled surgical demand varies unpredictably, the likelihood of inpatient overcrowding, ED backlogs, and ambulance diversions increases dramatically.
A number of management solutions have been introduced to improve patient flow. Separating low-acuity patients into a unique treatment stream can reduce the time these patients spend in the ED and improve overall patient satisfaction (Rodi, Grau, and Orsini 2006). Other tools and methods that have been employed to improve flow once a patient is admitted to the hospital relate to the discharge process. These approaches include creating a uniform discharge time (e.g., 11:00 a.m.), writing discharge orders the night before
Healthcare Operat ions Management284
release, communicating discharge plans early in the patient’s care, centralizing oversight of census and patient movement, changing physician rounding times, alerting ancillary departments when their testing procedures are critical to a patient’s discharge, and improving discharge coordination with social services (Clark 2005).
Investments in health information technology (IT) can improve patient flow as well. Devaraj, Ow, and Kohli (2013) studied 576 US hospitals to investigate the relationship between IT and investments in smooth and even flow. Using risk-adjusted length of stay (LOS) as their measure of smooth and even flow, they found that IT investments were positively related to smooth and even flow (shorter LOS) at the .05 level of significance.
They provide an example of how this result occurs (Devaraj, Ow, and Kohli 2013, page 190):
When the patient record is complete, the discharge IT system prompts the attending
physician to access the patient record from the cloud. After reviewing the record, the
attending physician can digitally sign the record and issue orders to discharge the
patient. Because the entire patient record resides in the cloud, the attending physician
can complete the entire process through a mobile device and discharge the patient
from anywhere. If a hospital automated the current process that requires attending
physicians to physically come to the hospital, often the next day, in order to review
and sign discharge orders, the LOS may not be significantly reduced. Therefore, it is
important for hospital managers to understand such complementarities (e.g., TSEF)
to ensure that IT is appropriately placed in the patient care “system.”
For patient flow to be carefully managed and improved, the formal methods of process improvement outlined in the next section need to be widely employed.
Process Improvement Approaches
Process improvement projects can use a variety of approaches and tools. Typi-cally, they begin with process mapping and measurement. Some simple tools can be initially applied to identify opportunities for improvements. Identifying and eliminating or alleviating bottlenecks in a system (theory of constraints) can quickly improve overall system performance. In addition, the Six Sigma tools described in chapter 9 can be used to reduce variability in process output, and the Lean tools discussed in chapter 10 can identify and eliminate waste. Finally, simulation (discussed later in this chapter) is a powerful tool that enables understanding and optimization of flow in a system.
Chapter 11: Process Improvement and Pat ient F low 285
All major process improvement projects should use the formal project management methodology outlined in chapter 5. An important first step is to identify a system’s owner: For a system to be managed effectively over time, it must have a designated individual who monitors the system as it operates, collects performance data, and leads teams to improve the system.
Many systems in healthcare do not have an owner and, therefore, operate inefficiently. For example, a patient may enter an ED, be assessed by the triage nurse, move to the admitting department, take a chair in the waiting area, be moved to an exam room, be seen by a floor nurse, have his blood drawn, and finally be examined by a physician. From the patient’s point of view, this is one system, but these various hospital departments may be operating autonomously. System ownership problems can be remedied by multidepartment teams with one individual designated as the overall system or process owner.
Problem Definition and Process MappingOnce the process owner is identified, the first step in improving a system is generally considered to be problem description and mapping of that pro-cess. However, the team should first ensure that the correct problem is being addressed. Mind mapping or root-cause analysis should be employed to ensure that the problem is identified and framed correctly; much time and money can be wasted finding an optimal solution to a process that is not problematic.
For example, suppose a project team is given the task of improving cus-tomer satisfaction with the ED. The team assumes that customer satisfaction is low because of high throughput time. It proceeds to optimize patient flow in the ED. Patient satisfaction does not improve.
Now, imagine that a second project team is assigned to improve customer satisfaction. It conducts an analysis of customer satisfaction, which reveals that customers are dissatisfied because of a lack of parking. The team solves the problem by following a different path than the first team because it has clearly understood and defined the issue, allowing team members to determine what process to map.
Processes can be described in a number of ways. The most common is the written procedure or protocol, typically constructed in the “directions” style. This type of process is sufficient for simple procedures—for example, “Turn right at Elm Street, go two blocks, and turn left at Vine Avenue.” Clearly written procedures are an important part of defining standardized work, as described in chapter 10.
However, when processes are linked to form systems, they become com-plex. These linked processes benefit from process mapping because process maps
• provide a visual representation that allows process improvement through inspection,
Healthcare Operat ions Management286
• enable branching in a process,• provide the ability to assign and measure the resources in each task in a
process, and• are the basis for modeling the process via computer simulation software.
Chapter 6 provides an introduction to process mapping. To review, the steps in process mapping are as follows:
1. Assemble and train the team.2. Determine the boundaries of the process (where it starts and ends) and
the level of detail desired.3. Brainstorm the major process tasks, and list them in order. (Sticky notes
are often helpful here.)4. Generate an initial process map (also called a flowchart). 5. Draw the formal flowchart using standard symbols for process
mapping.6. Check the formal flowchart for accuracy by all relevant personnel.7. Depending on the purpose of the flowchart, collect data needed or
include additional information.
Process Mapping ExampleA basic process map illustrating patient flow in VVH’s emergency department is displayed in exhibit 11.1.
Here, the patient arrives at the ED and is examined by the triage nurse. If the patient is very ill (high complexity level), she is immediately sent to the intensive care section of the ED. If not, she is sent to admitting and then to the routine care section of the ED.
The simple process map shown in exhibit 11.1 ends with the routine care step. In actuality, other processes now begin, such as admission into an inpatient bed or discharge from the ED to home with a scheduled clinical follow-up. The VVH emergency department process improvement project is detailed at the end of this chapter.
Process MeasurementsOnce a process map is developed, relevant data are collected and analyzed. The situation at hand dictates which specific data and measures should be employed. Important measures and data for possible collection and analysis include the following:
• Capacity of a process is the maximum possible amount of output (goods or services) that a process or resource can produce or transform.
Chapter 11: Process Improvement and Pat ient F low 287
Capacity measures can be based on outputs or on the availability of inputs. The capacity of a series of tasks is determined by the lowest-capacity task in the series.
• Capacity utilization is the proportion of capacity actually being used. It is measured as actual output divided by maximum possible output.
• Throughput time is the average time a unit spends in the process. Throughput time includes both processing time and waiting time and is determined by the critical (longest) path through the process.
• Throughput rate, sometimes referred to as drip rate, is the average number of units that can be processed per unit of time.
Triage–financial
EndDischargeWaitingWaiting
Waiting
Patientarrives
at the ED
IntensiveED care
AdmittingMedicaid
Triage–clinical
Complexity
Admittingprivate
insurance
Exam/treatment
Nursehistory/
complaint
Privateinsurance
Low
High
No
Yes
Waiting
EXHIBIT 11.1 VVH Emergency Department (ED) Patient Flow Process Map
Note: Created with Microsoft Visio.
Healthcare Operat ions Management288
• Service time or cycle time is the time to process one unit. The cycle time of a process is equal to the longest task cycle time in that process. The probability distribution of service times may also be of interest.
• Idle time or wait time is the time a unit spends waiting to be processed.• Arrival rate is the rate at which units arrive to the process. The
probability distribution of arrival rates may also be of interest.• Work-in-process, things-in-process, patients-in-process, or inventory
describes the total number of units in the process.• Setup time is the amount of time spent getting ready to process the next
unit.• Value-added time is the time a unit spends in the process where value is
being added to the unit.• Non-value-added time is the time a unit spends in the process where no
value is being added. Wait time is non-value-added time.• Number of defects or errors.
The art in process mapping is to provide enough detail to be able to measure overall system performance, determine areas for improvement, and measure the impact of these changes.
Tools for Process ImprovementOnce a system has been mapped, several techniques can be considered for improving the process. These improvements should result in a reduction in the duration, cost, or waste in a system.
Eliminate Non-Value-Added ActivitiesThe first step after a system has been mapped is to evaluate every element to ascertain whether each is necessary and provides value (to the customer or patient). If a system has been in place for a long period and has not been evalu-ated through a formal process improvement project, elements of the system can likely be easily eliminated. This step is sometimes referred to as “harvesting the low-hanging fruit.”
Eliminate Duplicate ActivitiesMany processes in systems have been added on top of existing systems without formally evaluating the total system, frequently resulting in duplicate activities. The most infamous redundant process step in healthcare is asking patients repeatedly for their contact information. Duplicate activities increase both time and cost in a system and should be eliminated whenever possible.
Chapter 11: Process Improvement and Pat ient F low 289
Combine Related ActivitiesProcess improvement teams should examine both the process map and the activity and swim lane map. If a patient moves back and forth between depart-ments, the movement should be reduced by combining these activities so he only needs to be in each department once.
Process in ParallelAlthough a patient can only be in one place at one time, other aspects of her care can be completed simultaneously. For example, medication preparation, physician review of tests, and chart documentation can all be performed at the same time. As more tasks are executed simultaneously, the total time a patient spends in the process is reduced. Similar to a chef who has a number of dishes on the stove synchronized to be completed at the same time, much of the patient care process can be completed simultaneously.
Another element of parallel processing is the relationship of subpro-cesses to the main flow. For example, a lab result may need to be obtained before a patient enters the operating suite. Many of these subprocesses can be synchronized through the analysis and use of takt time (chapter 10). This synchronization enables efficient process flow, thereby optimizing the process.
Balance WorkloadsIf similar workers perform the same task, a well-tuned system can be designed to balance the work among them. For example, a mass-immunization clinic should develop its system so that all immunization stations are active at all times. This aim can be accomplished by using a single queue that feeds into multiple immunization stations.
Load balancing (or load leveling, heijunka) is difficult when employ-ees can only perform a limited set of specific tasks (a consequence of the superspecialization of the healthcare professions). Load balancing is easier in environments that feature cross-training of employees than in those that limit employee tasks to singular functions.
Develop Alternative Process Flow Paths and Contingency PlansThe number and placement of decision points in the process should be evalu-ated and optimized. A system with few decision points has few alternative paths and, therefore, does not respond well to unexpected events. Alternative paths or contingency plans should be developed for these types of events. For example, a standard clinic patient rooming system should designate alternative paths for when an emergency occurs, a patient is late, a provider is delayed, or medical records are absent.
Healthcare Operat ions Management290
Establish the Critical PathFor complex pathways in a system, identifying the critical pathway with tools described in chapter 5 can be helpful. If a critical path can be identified, execu-tion of processes on the pathway can be improved (e.g., reduce average service time). In some cases, the process can be moved off the critical path and be performed in parallel to it. Either technique decreases the total time on the critical pathway. In the case of patient flow, moving this process off the critical pathway decreases the patient’s total time spent in the system.
Embed Information Feedback and Real-Time ControlSome systems have a high level of variability in their operations because they experience variability in the arrival of jobs or customers (patients) into the process and variability of the cycle time of each process in the system. High variability in the system can lead to poor performance. One tool to reduce variability is the control loop. Information can be obtained from one process and used to drive change in another. For example, the number of patients in the ED waiting area can be continuously monitored, and if it reaches a certain level, contingency plans—such as floating in additional staff from other por-tions of the hospital—can be initiated.
Ensure Quality at the SourceMany systems contain multiple reviews, approvals, and inspections. A system in which the task is performed correctly the first time should not require these redundancies. Deming (1998) first identified this problem in the process design of manufacturing lines that had inspectors throughout the assembly process. This expensive and ineffective system was one of the factors that gave rise to the quality movement in Japan and, later, the United States.
Systems should be designed to embed quality at their source or beginning to eliminate inspections. For example, a billing system that requires a clerk to inspect a bill before it is released does not have quality built into the process.
Match Capacity to DemandA common problem in 24-hour healthcare operations is having too few or too many staff for patient care demand. This problem is exacerbated if an organiza-tion only allows set shifts (e.g., eight hours).
To solve this problem, first graph and analyze demand on an hourly and daily basis. Then develop staffing patterns that match this demand. For example, a five-hour or seven-hour shift might be needed to correctly meet the demand.
Using the tools in chapter 7, you should be able to identify patterns of demand (e.g., high ED demand on Friday and Saturday evenings). Chapter 12 also provides details on capacity planning.
Chapter 11: Process Improvement and Pat ient F low 291
Let the Patient Do the WorkThe Internet and other advanced information technologies have allowed for increased self-service in service industries. Individuals are now comfortable booking their own airline reservations, buying goods online, and checking themselves out at retailers. This trend can be exploited in healthcare with tools that enable patients to be part of the process. For example, online tools are now available that allow patients to make their own clinic appointments. Letting the patient do the work reduces the work of staff and provides an opportunity for quality at the source—the data are more likely to be correct if the patients input them than if a staff member does so.
Use TechnologyThe electronic health record and other IT tools provide a platform to automate many tasks that were once performed manually. A good rubric through which to identify these tasks is to examine every daily task and ask where it ranks in complexity on the basis of your professional training. For those tasks that are low on this list, consider ways to automate them.
Today, work is an activity—not a place. The widespread use of smart-phones and tablets enables work to be performed outside the traditional work-place. Consider moving some tasks to these devices to improve your personal productivity.
Apply the Theory of ConstraintsChapter 6 discusses the underlying principles and applications of the theory of constraints, which can be used as a powerful process improvement tool. First, the bottleneck in a system is identified, often through the observation of queues forming in front of it. Once a bottleneck is identified, it should be exploited and everything else in the system subordinated to it. Specifically, other nonbottleneck resources (or steps in the process) should be synchronized to match the output of the constraint. Idleness at a nonbottleneck resource costs nothing, and nonbottlenecks should never produce more than can be consumed by the bottleneck resource. Often, this synchronization causes the bottleneck to shift and a new bottleneck is identified. However, if the original bottleneck remains, the possibility of elevating the bottleneck needs to be considered. Elevating bottlenecks requires additional resources (e.g., staff, equipment), so a comprehensive financial and outcomes analysis needs to be undertaken to determine the trade-offs among process improvement, quality, and costs.
Identify Best Practices and ReplicateAlthough this tip does not describe a formal operations management tool, it must be mentioned as a highly recommended management approach. As
Healthcare Operat ions Management292
health systems expand, they are likely to have many similar activities replicated in separate geographic sites. Good management practice is to identify high-performing sites (e.g., the best primary care clinic in a system) and replicate their core processes throughout the organization.
A similar approach can be taken with individual employees. For example, study the best billing clerk in a hospital to understand her processes and then replicate them with all the billers in a department.
The Science of Lines: Queuing Theory
Although most people are familiar with waiting in line, few are familiar with, or even aware of, queuing theory, or the theory of waiting lines. Most people’s experience with waiting lines is when they are actually part of those lines, for example, when waiting to check out in a retail environment. In a manufactur-ing environment, items wait in line to be worked on. In a service environment, customers wait for a service to be performed.
Queues, or lines, form because the resources needed to serve them (servers) are limited—deploying unlimited resources is economically unfeasible. Queuing theory is used to study systems to determine the best balance between service to customers (short or no waiting lines, implying many resources or servers) and economic considerations (few servers, implying long lines). A simple queuing system is illustrated in exhibit 11.2.
Customers (often referred to as entities) arrive and either are served (if there is no line) or enter the queue (if others are waiting to be served). Once they are served, customers exit the system.
The customer population, or input source, can be either finite or infi-nite. If the source is effectively infinite, the analysis of the system is easier than if it is finite because simplifying assumptions can be made.
The arrival process is characterized by the arrival pattern—the rate at which customers arrive (number of customers divided by unit of time)—or by the interarrival time (time between arrivals) and the distribution in time of those arrivals. The distribution of arrivals can be constant or variable. A
Queuing theory The mathematical study of wait lines.
Customerpopulation,input source
Buffer or queue
Server(s) ExitArrival
EXHIBIT 11.2 Simple Queuing
System
Chapter 11: Process Improvement and Pat ient F low 293
constant arrival distribution has a fixed interarrival time. A variable, or random, arrival pattern is described by a probability distribution. The queue discipline is the method by which customers are selected from the queue to be served. Often, customers are served in the order in which they arrived—first come, first served. However, many other queue disciplines are possible, and choice of a particular discipline can greatly affect system performance. For example, choosing the customer whose service can be completed most quickly (shortest processing time) usually minimizes the average time customers spend waiting in line. This result is one reason urgent care centers are often located near an ED—urgent issues can usually be handled more quickly than true emergen-cies can.
The service process is characterized by the number of servers and service time. Like arrivals, the distribution of service times can be constant or vari-able. Often, the exponential distribution (M) is used to model variable service times, μ is the mean service rate, λ is the mean arrival rate, and ρ is capacity utilization. (An exponential distribution creates data points that simulate a purely random process.)
Queuing NotationThe type of queuing system is identified with a specific notation in the form of A/B/c/D/E. The A represents the interarrival time distribution, and B represents the service time distribution. A and B together are represented as either a deterministic or a constant rate. The c represents the number of serv-ers, D is the maximum queue size, and E is the size of the input population. When both queue and input population are assumed to be infinite, D and E are typically omitted. An M/M/1 queuing system, therefore, has an exponential service time distribution, a single server, an infinite possible queue length, and an infinite input population; it assumes only one queue. An M/M/1 queue for VVH is used as an example throughout the remainder of the chapter.
Queuing SolutionsAnalytic solutions for some simple queuing systems at equilibrium or steady state (after the system has been running for some time and is unchanging, often referred to as a stable system) have been determined; however, the derivation of these results is outside the scope of this text. Refer to Cooper (1981) for a complete derivation and results for many other types of queuing systems.
Here, we focus primarily on the M/M/1 queuing system by presenting the results for an M/M/1 queue where λ < μ—the arrival rate is less than the service rate. Note that if λ ≥ μ (customers arrive faster than they are served), the queue becomes infinitely long, the number of customers in the system becomes infinite, waiting time becomes infinite, and the server experiences 100 percent capacity utilization (percentage of time the server is busy). The
Queue disciplineIn queuing theory, the method by which customers are selected from the queue to be served.
Healthcare Operat ions Management294
following formulas can be used to determine some characteristics of the queu-ing system at steady state.
Capacity utilization:
Wq =−λ
μ μ λ( )
ρλμ
= = =Mean arrival rateMean service rate
1 Meaan time between arrivals1/Mean service timeeMean service time
Mean time between arriv=
aals
Average waiting time in queue:
Wq =−λ
μ μ λ( )
ρλμ
= = =Mean arrival rateMean service rate
1 Meaan time between arrivals1/Mean service timeeMean service time
Mean time between arriv=
aals
Average time in the system (average waiting time in queue plus average service time):
Lq =−
=⎛
⎝
⎞
⎠ −
⎛
⎝
⎞
⎠λ
μ μ λλμ
λμ λ
2
( )
W Ws q= + =−
1 1μ μ λ
= Arrival rate × Time in the systemL Wss = −=
λμ λ
λ
Average length of queue (average number in queue):
Lq =−
=⎛
⎝
⎞
⎠ −
⎛
⎝
⎞
⎠λ
μ μ λλμ
λμ λ
2
( )
W Ws q= + =−
1 1μ μ λ
= Arrival rate × Time in the systemL Wss = −=
λμ λ
λAverage total number of customers in the system:Lq =
−=⎛
⎝
⎞
⎠ −
⎛
⎝
⎞
⎠λ
μ μ λλμ
λμ λ
2
( )
W Ws q= + =−
1 1μ μ λ
= Arrival rate × Time in the systemL Wss = −=
λμ λ
λ = Arrival rate × Time in the system
This last result is called Little’s law and applies to all types of queuing systems and subsystems. To summarize this result in plain language, in a stable system or process, the number of things in the system is equal to the rate at which things arrive to the system multiplied by the time they spend in the system. In a stable system, the average rate at which things arrive to the system is equal to the average rate at which things leave the system. If this were not true, the system would not be stable.
Little’s law can also be restated using other terminology:
Inventory (things in the system) = Arrival rate (or departure rate) × Throughput time (flow time)
Little’s lawThe relationship between the arrival rate to a system, the time an item (e.g., a patient) spends in the system, and the number of items in a system.
Chapter 11: Process Improvement and Pat ient F low 295
or
Throughput time = Inventory ÷ Arrival rate
Knowledge of two of the variables in Little’s law allows calculation of the third variable. Consider a clinic that serves 200 patients in an eight-hour day, or an average of 25 patients an hour. The average number of patients in the clinic (waiting room, exams rooms, etc.) is 15. Therefore, the average throughput time is
T = I/λ
=
= 0.6 hour,
where T is throughput time, λ is patients per hour, and I is number of patients. Hence, each patient spends an average of 36 minutes in the clinic.
Little’s law has important implications for process improvement and can be seen as the basis of many improvement techniques. Throughput time can be decreased by decreasing inventory or increasing departure rate. Lean initiatives often focus on decreasing throughput time (or increasing throughput rate) by decreasing inventory. The theory of constraints (chapter 6) focuses on identifying and eliminating system bottlenecks. The departure rate in any system is equal to 1 ÷ task cycle time of the slowest task in the system or process (the bottleneck). Decreasing the amount of time an object spends at the bottleneck task therefore increases the departure rate of the system and decreases throughput time.
Vincent Valley Hospital and Health System M/M/1 QueueVVH began receiving complaints from patients related to crowded conditions in the waiting area for magnetic resonance imaging (MRI) procedures. The organization has determined a goal to average just one patient waiting in line for the MRI. It has collected data on arrival and service rates and sees that, for MRIs, the mean service rate (μ) is four patients per hour, exponentially distributed. VVH also finds that the mean arrival rate (λ) is three patients per hour. To find the capacity utilization of MRI (percentage of time the MRI is busy), VVH uses the following formula:
⁄⁄
34
75% or11
15 minutes20 minutes
75%.ρλμ
ρμλ
= = = = = =
If one customer arrives every 20 minutes and assuming each MRI takes 15 minutes to complete, the MRI is busy 75 percent of the time.
15 patients
25 patients/hour
Healthcare Operat ions Management296
Next, VVH calculates patients’ average time waiting in line,
Ls = λWs = Arrival rate × Time in the system =3 Patients/Hour × 1 Hour = 3 Patients
Ls =−
=−
=λ
μ λ3
4 33 patients
Ws =−
=−
=1 1
4 3μ λ1 hour.
Wq =−
=−
= =λ
μ μ λ( ) ( )3
4 4 334
0.75 hour,
ρλμ
ρμλ
= = = = = =34
7511
1520
% orMinutesMinutes
775%
and average time spent in the system,
Ls = λWs = Arrival rate × Time in the system =3 Patients/Hour × 1 Hour = 3 Patients
Ls =−
=−
=λ
μ λ3
4 33 patients
Ws =−
=−
=1 1
4 3μ λ1 hour.
Wq =−
=−
= =λ
μ μ λ( ) ( )3
4 4 334
0.75 hour,
ρλμ
ρμλ
= = = = = =34
7511
1520
% orMinutesMinutes
775%
Finally, it determines average total number of patients in the system,
Ls = λWs = Arrival rate × Time in the system =3 Patients/Hour × 1 Hour = 3 Patients
Ls =−
=−
=λ
μ λ3
4 33 patients
Ws =−
=−
=1 1
4 3μ λ1 hour.
Wq =−
=−
= =λ
μ μ λ( ) ( )3
4 4 334
0.75 hour,
ρλμ
ρμλ
= = = = = =34
7511
1520
% orMinutesMinutes
775%
or
Ls = λWs = Arrival rate × Time in the system = 3 patients/hour × 1 hour = 3 patients,
and average number of patients in the waiting line,
Lq =−
=−
=
− = − = =
−
33
33
1
3 3 3 9
3
2 2
2 2
2
μ μ μ μ
μ μ μ μ
μ
( ) ( )
( )
μμ
μ
− =
=
9 04 85..
Lq =−
=−
=
= × − = −
+
λμ μ λ
λλ
λ λ λ
λ
2 2
2
2
4 41
4 4 16 4
( ) ( )
( )
44 16 02 47.λ
λ
− =
= .
Lq =−
=⎛
⎝⎜
⎞
⎠⎟
−
⎛
⎝⎜
⎞
⎠⎟ =
⎛⎝⎜
⎞⎠⎟
−λ
μ μ λλμ
λμ λ
2 34
34 3( )
⎛⎛⎝⎜
⎞⎠⎟
=−
= =3
4 4 394
2
( )2.25 patients.
To decrease the average number of patients waiting, VVH needs to decrease the utilization, ρ = λ ÷ μ, of the MRI process. In other words, the service rate must be increased or the arrival rate decreased. VVH may increase the service rate by making the MRI process more efficient so that the average time to perform the procedure is decreased and MRIs can be performed on a greater number of patients in an hour. Alternatively, the organization may decrease the arrival rate by scheduling fewer patients per hour.
To achieve its goal (assuming that the service rate is not increased), VVH needs to decrease the arrival rate to
Lq =−
=−
=
− = − = =
−
33
33
1
3 3 3 9
3
2 2
2 2
2
μ μ μ μ
μ μ μ μ
μ
( ) ( )
( )
μμ
μ
− =
=
9 04 85..
Lq =−
=−
=
= × − = −
+
λμ μ λ
λλ
λ λ λ
λ
2 2
2
2
4 41
4 4 16 4
( ) ( )
( )
44 16 02 47.λ
λ
− =
= .
Lq =−
=⎛
⎝⎜
⎞
⎠⎟
−
⎛
⎝⎜
⎞
⎠⎟ =
⎛⎝⎜
⎞⎠⎟
−λ
μ μ λλμ
λμ λ
2 34
34 3( )
⎛⎛⎝⎜
⎞⎠⎟
=−
= =3
4 4 394
2
( )2.25 patients.
Alternatively (assuming that the arrival rate is not decreased), VVH may increase the service rate to
Chapter 11: Process Improvement and Pat ient F low 297
Lq =−
=−
=
− = − = =
−
33
33
1
3 3 3 9
3
2 2
2 2
2
μ μ μ μ
μ μ μ μ
μ
( ) ( )
( )
μμ
μ
− =
=
9 04 85..
Lq =−
=−
=
= × − = −
+
λμ μ λ
λλ
λ λ λ
λ
2 2
2
2
4 41
4 4 16 4
( ) ( )
( )
44 16 02 47.λ
λ
− =
= .
Lq =−
=⎛
⎝⎜
⎞
⎠⎟
−
⎛
⎝⎜
⎞
⎠⎟ =
⎛⎝⎜
⎞⎠⎟
−λ
μ μ λλμ
λμ λ
2 34
34 3( )
⎛⎛⎝⎜
⎞⎠⎟
=−
= =3
4 4 394
2
( )2.25 patients.
VVH may also implement some combination of decreasing arrival rate and increasing service rate. In all cases, utilization of the MRI will be reduced to ρ = λ ÷ μ = 2.47 ÷ 4.00, or 3.00 ÷ 4.85 = 0.62.
Real systems are seldom as simple as an M/M/1 queuing system and rarely reach equilibrium. Often, simulation is needed to study these more complicated systems.
Discrete Event SimulationDiscrete event simulation (DES) is typically performed using commercially available software packages. As with Monte Carlo simulation, performing DES by hand is an option, albeit a tedious one. Two popular simulation software packages are Arena (Rockwell Automation 2016) and Simul8 (Simul8 Cor-poration 2016).
The terminology and general logic of DES are built on queuing theory. A basic simulation model consists of entities, queues, and resources, all of which can have various attributes. Entities are the objects that flow through the system; in healthcare, entities typically are patients, but they can be any object on which some service or task will be performed. For example, blood samples in the hematology lab are entities. Queues are the waiting lines that hold the entities while they await service. Resources (previously referred to as servers) can be people, equipment, or space for which entities compete.
The specific operation of a simulation model is based on states (variables that describe the system at a point in time) and events (variables that change the state of the system). Events are controlled by the simulation executive, and data are collected on the state of the system as events occur. The simulation jumps through time from event to event.
A simple example from the Vincent Valley Hospital and Health System M/M/1 MRI queuing discussion helps show the logic behind DES software. Exhibit 11.3 contains a list of the events as they happen in the simulation. The arrival rate is three patients per hour, and the service rate is four patients per hour. Random interarrival times are generated using an exponential distribu-tion with a mean of 0.33 hours. Random service times are generated using an exponential distribution with a mean of 0.25 hours (shown at the bottom of exhibit 11.10 later in this chapter).
Healthcare Operat ions Management298
Just
Fin
ishe
dVa
riab
les
Att
ribu
tes
Sta
tist
ics
Upc
omin
g Ev
ents
Enti
ty
No.
Tim
eEv
ent
Type
Leng
th o
f Q
ueue
Ser
ver
Bus
y
Arr
ival
Ti
me
in
Que
ue
Arr
ival
Ti
me
in
Ser
vice
Num
ber
of
Com
plet
e W
aits
in
Que
ue
Tota
l Wai
t Ti
me
in
Que
ue
Ave
rage
Q
ueue
Le
ngth
Uti
liza
tion
Enti
ty
No.
Tim
eEv
ent
10.
00Ar
r0
10.
000.
000
00
02
0.17
Arr
10.
21D
ep
20.
17Ar
r1
10.
170
00
1.00
10.
21D
ep
30.
54Ar
r
10.
21D
ep0
10.
000.
001
0.04
0.19
1.00
30.
54Ar
r
20.
77D
ep
30.
54Ar
r1
10.
541
0.04
0.07
1.00
20.
77D
ep
40.
90Ar
r
20.
77D
ep0
10.
170.
212
0.27
0.35
1.00
30.
79D
ep
40.
90Ar
r
30.
79D
ep0
00.
773
0.27
0.34
1.00
40.
90Ar
r
41.
27D
ep
40.
90Ar
r0
10.
903
0.27
0.30
0.88
41.
27D
ep
51.
49Ar
r
12
34
56
78
Inte
rarr
ival
tim
eEx
pon
(0.3
3)0.
170.
370.
360.
590.
140.
170.
240.
06
Tim
e of
arr
ival
0.00
0.17
0.54
0.90
1.49
1.63
1.80
2.04
2.10
Serv
ice
tim
eEx
pon
(0.2
5)0.
210.
560.
020.
370.
340.
111.
020.
010.
20
Not
e: A
rr =
arr
ival
; Dep
= d
epar
ture
; Exp
on =
exp
onen
t.
EXH
IBIT
11.
3Si
mul
atio
n Ev
ent L
ist
Chapter 11: Process Improvement and Pat ient F low 299
The simulation starts at time 0.00. The first event is the arrival of the first patient (entity); there is no line (queue), so this patient enters service. Upcoming events are the arrival of the next patient at 0.17 hours (the interar-rival between patients 1 and 2 is 0.17 hours) and the completion of the first patient’s service at 0.21 hours.
The next event is the arrival of patient 2 at 0.17 hours. Because the MRI on patient 1 is not complete, patient 2 enters the queue. The MRI has been busy since the start of the simulation, so the utilization of the MRI is 100 percent. Upcoming events are the completion of the first patient’s service at 0.21 hours and the arrival of patient 3 at 0.54 hours (the interarrival between patients 2 and 3 is 0.37 hours).
When the first patient’s MRI is completed at 0.21 hours, no one is waiting in the queue because once patient 1 has completed service, patient 2 can enter service. The total waiting time in the queue for all patients is 0.04 hours (the difference between when patient 2 entered the queue and entered service). The average queue length is 0.19 patients. No people were in line for 0.17 hours, and one person was in line for 0.04 hours:
= 0.19 people.
Upcoming events are the arrival of patient 3 at 0.54 hours and the departure of patient 2 at 0.77 hours (patient 2 entered service at 0.21 hours, and service takes 0.56 hours).
Patient 3 arrives at 0.54 hours and joins the queue because the MRI is still busy with patient 2. The average queue length has decreased from the previous event because more time has passed with no one in the queue—only one person has been in the queue for 0.04 hours, but total time in the simula-tion is 0.54 hours. Upcoming events are the departure of patient 2 at 0.77 hours and the arrival of patient 4 at 0.90 hours.
Patient 2 departs at 0.77 hours. No one is waiting in the queue at this point because patient 3 has entered service. Two people have departed the system. The total wait time in the queue for all patients is 0.04 hours for patient 2 plus 0.17 hours for patient 3 (0.77 hours − 0.54 hours) for a total of 0.21 hours. The average queue length is
= 0.35 people.
The MRI utilization is still at 100 percent because the MRI has been busy constantly since the start of the simulation. Upcoming events are the departure
0 people × 0.17 hours + 1 person × 0.04 hours
0.21 hours
0 people × 0.50 hours + 1 person × 0.21 hours
0.77 hours
Healthcare Operat ions Management300
of patient 3 at 0.79 hours (patient 3 arrived at 0.54 hours, and service takes 0.25 hours) and the arrival of patient 4 at 0.90 hours.
Patient 3 departs at 0.79 hours. Because no patients are waiting for the MRI, it becomes idle. Upcoming events are the arrival of patient 4 at 0.90 hours and the departure of patient 4 at 1.27 hours.
With patient 4 arriving at 0.90 hours and entering service, the utilization of the MRI has decreased to 88 percent because it was idle for 0.11 hours of the 0.90 hours the simulation has run. Upcoming events are the departure of patient 4 at 1.27 hours and the arrival of patient 5 at 1.49 hours. The simula-tion continues in this manner until the desired stop time is reached.
Even for this simple model, performing these calculations by hand takes a long time. Additionally, an advantage of simulation is that it uses process map-ping; many simulation software packages are able to import and use Microsoft Visio process and value stream maps. DES software allows process improvement teams to build, run, and analyze simple models in limited time; Arena software was used to build and simulate the present model (exhibit 11.4).
As before, the arrival rate is three patients per hour, the service rate is four patients per hour, and both rates are exponentially distributed. Averages over time for queue length, wait time, and utilization for a single replication are
SCANNER
AVERAGE NUMBER IN QUEUE AVERAGE WAIT IN QUEUE AND SYSTEM
3.0
0.0
2.0
0.0
1.0
0.0
0.0020.00.0020.0
0.0020.0
MRI UTILIZATION
Patientdemand
MRI exam Exit
9 8 51 9 5 2
03 : 57 : 26
Note: Created with Arena simulation software. M = exponential distribution; MRI = magnetic resonance imaging.
EXHIBIT 11.4 Arena
Simulation of VVH MRI
M/M/1 Queuing Example
Chapter 11: Process Improvement and Pat ient F low 301
shown in the plots in exhibit 11.12 later in the chapter. Each of 30 replications of the simulation is run for 200 hours. Replications are needed to determine confidence intervals for the reported values. Some of the output from this simulation is shown in exhibit 11.5. The sample mean plus or minus the half-width gives the 95 percent confidence interval for the mean. Increasing the number of replications reduces the half-width. The results of this simulation agree fairly closely with the calculated steady-state results because the process was assumed to run continuously for a significant period, 200 hours. A more realistic assumption might be that MRI procedures are only performed ten hours every day. The Arena simulation was rerun with this assumption, and the results are shown in exhibit 11.6. The average wait times, queue length, and utilization are lower than the steady-state values.
Category OverviewJuly 26, 20118:22:36 AM
Values across all replications
MRI Example
Replications: 30 Time unit: Hours
Key Performance Indicators
Average601
SystemNumber out
Entity
Time
Patient
Patient
TotalTime
AverageHalf-
WidthWaitTime
MinimumAverage
MinimumValue
MaximumAverage
MaximumValue
AverageHalf-
WidthMinimumAverage
MinimumValue
MaximumAverage
MaximumValue
AverageHalf-
WidthMinimumAverage
MinimumValue
MaximumAverage
MaximumValue
AverageHalf-
WidthMinimumAverage
MinimumValue
MaximumAverage
MaximumValue
0.7241
0.9734
0.08
0.08
0.5009
0.7427
1.3496 0.00 7.3900
1.6174 0.00001961 7.4140
2.1944 0.25 1.4326 4.2851 0.00 29.0000
0.7488 0.01 0.6767 0.8513 0.00 1.0000
Usage
InstantaneousUtilization
NumberWaiting
MRI exam queue
Resource
MRI
Arrival rate = 3 patients/hour; service rate = 4 patients/hour.
Queue
Other
Note: Created with Arena simulation software. M = exponential distribution; MRI = magnetic resonance imaging.
EXHIBIT 11.5Arena Output for VVH MRI M/M/1 Queuing Example: 200 Hours
Healthcare Operat ions Management302
Vincent Valley Hospital and Health System M/M/1 QueueVVH has determined that a steady-state analysis is not appropriate for its situa-tion because MRIs are only offered ten hours a day. The process improvement team assigned to this system decides to analyze the situation using simulation. Once the model is built and run, the model and simulation results are com-pared with actual data and evaluated by relevant staff to ensure that the model accurately reflects reality. All staff agree that the model is valid and can be used to determine how to achieve the stated goal. If the model had not been con-sidered valid, the team would have needed to build and validate a new model.
The results of the simulation (refer to exhibit 11.14 later in the chapter) indicate that VVH has an average of 1.5 patients in the queue. To reach the desired goal of only one patient waiting on average, VVH needs to decrease the arrival rate or increase the service rate. Using trial and error in the simulation,
Category OverviewJuly 26, 201112:19:03 PM
Values across all replications
MRI Example
Replications: 30 Time unit: Hours
Key Performance Indicators
Average28
SystemNumber out
Entity
Time
Patient
Patient
TotalTime
AverageHalf-
WidthWaitTime
MinimumAverage
MinimumValue
MaximumAverage
MaximumValue
AverageHalf-
WidthMinimumAverage
MinimumValue
MaximumAverage
MaximumValue
AverageHalf-
WidthMinimumAverage
MinimumValue
MaximumAverage
MaximumValue
AverageHalf-
WidthMinimumAverage
MinimumValue
MaximumAverage
MaximumValue
0.4778
0.7304
0.15
0.16
0.02803444
0.2407
1.4312 0.00 2.9818
1.7611 0.00082680 3.3129
1.5265 0.46 0.2219 4.5799 0.00 10.0000
0.7167 0.05 0.4088 0.9780 0.00 1.0000
Usage
InstantaneousUtilization
NumberWaiting
MRI exam queue
Resource
MRI
Arrival rate = 3 patients/hour; service rate = 4 patients/hour.
Queue
Other
Note: Created with Arena simulation software. M = exponential distribution; MRI = magnetic resonance imaging.
EXHIBIT 11.6Arena Output
for VVH MRI M/M/1 Queuing
Example: 10 Hours
Chapter 11: Process Improvement and Pat ient F low 303
the organization finds that decreasing the arrival rate to 2.7 or increasing the service rate to 4.4 will allow the goal to be achieved.
However, even using the improvement tools in this text, the team believes that the organization will only be able to increase the service rate of the MRI to 4.2 patients per hour. Therefore, to reach the goal, the arrival rate must also be decreased. Again using the simulation, VVH finds that it needs to decrease the arrival rate to 2.8 patients per hour. Exhibit 11.7 shows the results of this simulation.
The team recommends that (1) a kaizen event be held for the MRI process to increase service rate and (2) appointments for the MRI be reduced to decrease the arrival rate. However, the team also notes that implementing these changes will reduce the average number of patients served from 28 to 26 and reduce the utilization of the MRI from 0.72 to 0.69. More positively, average patient wait time will be reduced from 0.48 hours to 0.35 hours.
Category OverviewJuly 26, 20118:24:44 AM
Values across all replications
MRI Example
Replications: 30 Time unit: Hours
Key Performance Indicators
Average26
SystemNumber out
Entity
Patient
Patient
TotalTime
AverageHalf-
WidthWaitTime
MinimumAverage
MinimumValue
MaximumAverage
MaximumValue
AverageHalf-
WidthMinimumAverage
MinimumValue
MaximumAverage
MaximumValue
AverageHalf-
WidthMinimumAverage
MinimumValue
MaximumAverage
MaximumValue
AverageHalf-
WidthMinimumAverage
MinimumValue
MaximumAverage
MaximumValue
0.3507
0.6008
0.12
0.14
0.02449931
0.1899
1.4202 0.00 3.4973
1.7825 0.00097591 4.2210
1.0342 0.36 0.0928 4.2272 0.00 9.0000
0.6682 0.06 0.3314 0.9456 0.00 1.0000
Usage
InstantaneousUtilization
NumberWaiting
MRI exam queue
Resource
MRI
Arrival rate = 2.8 patients/hour; service rate = 4.2 patients/hour; 10 hours simulated.
Queue
Other
EXHIBIT 11.7Arena Output for VVH MRI M/M/1 Queuing Example: Decreased Arrival Rate, Increased Service Rate
Note: Created with Arena simulation software. M = exponential distribution; MRI = magnetic resonance imaging.
Healthcare Operat ions Management304
VVH is able to increase the service rate to 4.2 patients per hour and decrease the arrival rate to 2.8 patients per hour, and the results are as predicted by the simulation. The team now begins to investigate other solutions enabling VVH to increase MRI utilization while maintaining wait times and queue length.
Simulation and Queuing Theory FindingsSimulation is a powerful tool for modeling processes and systems to evaluate choices and opportunities. As is true of all of the tools and techniques presented in this text, simulation can be used in conjunction with other initiatives, such as Lean or Six Sigma, to enable continuous improvement of systems and processes.
In a series of studies, queuing theory has been used to analyze flow of EDs and operating rooms (Butterfield 2007; McManus et al. 2004). In many instances, surgical suites more than doubled the number of surgeries they are able to complete in a short time. Because surgeries are a prime source of revenue and margin for most hospitals, this improvement makes the hospital more profitable.
Process Improvement in Practice
In this section, we review methods and tools that, in addition to simulation, are key approaches to process improvement, and we apply them to an emergency department scenario at VVH.
Review of MethodologiesSix SigmaIf the primary goal of a process improvement project is to improve quality (reduce the variability in outcomes), the Six Sigma approach and tools described in chapter 9 yield the best results. As discussed previously, Six Sigma uses seven basic tools: fishbone diagrams, check sheets, histograms, Pareto charts, flowcharts, scatter plots, and run charts. It also includes statistical process control to provide an ongoing measurement of process output characteristics to ensure quality and enable the identification of a problem situation before an error occurs.
The Six Sigma approach also includes measuring process capability—whether a process is capable of producing the desired output—and benchmark-ing it against other similar processes in other organizations. Quality function deployment is used to match customer requirements (voice of the customer) with process capabilities given that trade-offs must be made. Poka-yoke is employed selectively to mistake-proof parts of a process.
A primary function of Six Sigma programs is to eliminate sources of artificial variance in processes and systems. Natural variance occurs in any system, such as heat, temperature, and patients getting sick or breaking a leg. Artificial variance is created by the people in the system and is completely
Chapter 11: Process Improvement and Pat ient F low 305
in their control. Six Sigma programs identify and eliminate those sources of artificial variance. For example, scheduling systems, overtime allocations, and business office processing systems can all be changed by people in the system. The secret to a successful Six Sigma program is removing all the artificial vari-ance and focusing on creating value for customers. Effective Six Sigma systems strategically employ Lean concepts to achieve this goal.
LeanProcess improvement projects focused on eliminating waste and improving flow in the system or process can use many of the tools that are part of the Lean approach (chapter 10). The kaizen philosophy, which is the basis for Lean, includes the following steps:
1. Specify value. Identify activities that provide value from the customer’s perspective.
2. Map and improve the value stream. Determine the sequence of activities or the current state of the process and the desired future state. Eliminate non-value-added steps and other waste.
3. Enable flow. Allow the process to flow as smoothly and quickly as possible.
4. Enable pull. Allow the customer to pull products or services.5. Perfect. Repeat the cycle to ensure a focus on continuous
improvement.
An important part of Lean is value stream mapping, which is used to define the process and determine where waste is occurring. Takt time measures the time needed for the process to occur. It is based on customer demand and can be used to synchronize flow in a process. Standardized work, an important part of the Lean approach, is written documentation of the precise way in which every step in a process should be performed and helps ensure that activities are completed the same way every time in an efficient manner.
Other Lean tools include the five Ss (a technique to organize the work-place) and spaghetti diagrams (a mapping technique to show the movement of customers, patients, workers, equipment, jobs, etc.). Leveling workload (hei-junka) so that the system or process flows without interruption can be used to improve the value stream. Kaizen blitzes or events are Lean tools used to improve the process quickly when project management is not needed (chapter 10).
Process Improvement Project: Vincent Valley Hospital and Health System Emergency DepartmentTo demonstrate the power of many of the process improvement tools described in this book, an extensive patient flow process improvement project at VVH is examined.
Healthcare Operat ions Management306
VVH has identified patient flow in the ED as an important area on which to focus process improvement efforts. The goal of the project is to reduce total patient time in the ED (both waiting and care delivery) while maintaining or improving financial performance.
The first step for VVH leadership is to charter a multidepartmental team using the project management methods described in chapter 5. The head nurse for emergency services has been appointed project leader. The team feels VVH should take a number of steps to improve patient flow in the ED and splits the systems improvement project into three major phases. First, team members will perform simple data collection and basic process improvement to identify low-hanging fruit and make obvious, straightforward changes.
Once the team feels comfortable with its understanding of the basics of patient flow in the department, it will work to understand the elements of the system more fully by collecting detailed data. Then, value stream mapping and the theory of constraints will be used to identify opportunities for improve-ment. Root-cause analysis will be employed on poorly performing processes and tasks; resulting changes will be adopted and their effects measured.
The third phase of the project will be the use of simulation. Because the team, by this stage in the improvement effort, will have complete knowledge of patient flow in the system, it will be able to develop and test a simulation model with confidence. Once the simulation is validated, the team will con-tinuously test process improvements in the simulation model and implement them in the ED.
The specific high-level tasks in this project are as follows.
Phase I
1. Observe patient flow and develop a detailed process map.2. Measure high-level patient flow metrics for one week:
• Patients arriving per hour• Patients departing per hour to inpatient• Patients departing per hour to home• Number of patients in the ED, including the waiting area and exam
rooms3. With the process map and data in hand, use simple process
improvement techniques to make changes in the process, then measure the results.
Phase II4. Set up a measurement system for each individual process, and take
measurements over one week.
Chapter 11: Process Improvement and Pat ient F low 307
5. Use value stream mapping and the theory of constraints to analyze patient flow and make improvements, then measure the effects of the changes.
Phase III6. Collect data needed to build a realistic simulation model.7. Develop the simulation model and validate it against real data.8. Use the simulation model to conduct virtual experiments on process
improvements. Implement promising improvements, and measure the results of the changes.
Phase IVVH process improvement project team members observe patient flow and record the needed data. With the information collected, the team creates a detailed process map. Team members measure the following high-level operat-ing statistics related to patient flow:
• Patients arriving per hour = 10• Patients departing per hour to inpatient = 2• Patients triaged to routine emergency care per hour = 8• Patients departing per hour to home = 8• Average number of patients in various parts of the system (sampled
every 10 minutes) = 20• Average number of patients in ED exam rooms = 4
Using Little’s law, the average time in the ED (throughput time) is calculated as
Throughput time = T = I/λ
=
= 3 hours.
Hence, each patient spends an average of 3 hours, or 180 minutes, in the ED. However, Little’s law only gives the average time in the department
at steady state. Therefore, the team measures total time in the system for a sample of routine patients and determines an average of 165 minutes. It also observes that the number of patients in the waiting room varies from 0 to 20 and the actual time to move through the process varies from one hour to more than five hours.
24 patients
8 patients/hour
Healthcare Operat ions Management308
Initially, the team focuses on the ED admitting subsystem as an opportu-nity for immediate improvement. Exhibit 11.8 shows the complete ED system, with the admitting subsystem highlighted.
The team develops the following description of the admitting process from its documentation of patient flow:
Patients who did not have an acute clinical problem were asked if they had health
insurance. If they did not have health insurance, they were sent to the admitting clerk
who specializes in Medicaid (to enroll them in a Medicaid program). If they had health
insurance, they were sent to the other clerk, who specializes in private insurance. If
a patient had been sent to the wrong clerk by triage, he was sent to the other clerk.
Triage–financial
RoutineED care
End
Patientarrives
at the ED
IntensiveED care
Triage–clinical
Complexity
Low
High
Waiting
Admitting Subsystem
Waiting
AdmittingMedicaid
Admittingprivate
insurance
Privateinsurance
No
Yes
EXHIBIT 11.8VVH Emergency
Department (ED) Admitting
Subsystem
Note: Created with Microsoft Visio.
Chapter 11: Process Improvement and Pat ient F low 309
The team determines that one process improvement change could be to cross-train the admitting clerks on both private insurance and Medicaid eligibility. This training would provide for load balancing, as patients would automatically go to the free clerk. In addition, this system improvement would eliminate triage staff errors in sending patients to the wrong clerk, hence pro-viding quality at the source.
Phase IIPhase I produced some gains in reducing patient time in the ED. However, the team feels more detailed data are needed to improve further. As a first step in collecting these data, the team measures various parameters of the depart-ment’s processes. Initially, it focuses on the period from 2:00 p.m. to 2:00 a.m., Monday through Thursday, as this is the busy period in the ED and demand seems relatively stable during these times.
The team draws a more detailed process map (exhibit 11.9) and performs value stream mapping of this process (exhibit 11.10). First, team members evaluate each step in the process to determine if it is value-added, non-value-added, or non-value-added but necessary. Then, they measure the time a patient spends at each step in the process. The team finds that after a patient has given his insurance information, he spends an average of 30 minutes of non-value-added time in the waiting room before a nurse is available to take his history and record the presenting complaint, a process that takes an average of 20 minutes to complete. The percentage of value-added time for these two steps is
(Value-added time ÷ Total time) × 100 = [20 minutes ÷ (30 minutes + 20 minutes)] × 100 = 40%.
The team believes the waiting room process can be improved through automation. Patients are handed a tablet personal computer in the waiting area and asked to enter their symptoms and history via a series of branched questions. The results are sent via a wireless network to VVH’s electronic health record (EHR). This step takes patients an average of 20 minutes to complete. Staff know which patients have completed the electronic interview by checking the EHR and can prioritize which patient is to be seen next. This new procedure also reduces the time the nurse spends with the patient to 10 minutes because it enables the nurse to verify, rather than record, presenting symptoms and patient history. The percentage of value-added time for the new procedure is
(Value-added time ÷ Total time) × 100 = [(Patient history time + Nurse history time) ÷ (Patient history time
+ Wait time + Nurse history time)] × 100
= [(20 minutes + 10 minutes) ÷ (20 minutes + 10 minutes + 10 minutes)] × 100 = 75%.
Healthcare Operat ions Management310
The average throughput time for a patient in the ED is reduced by 10 minutes. The average time for patients to flow through the department (throughput time) prior to this improvement was 155 minutes. Because this step is on the critical path of the complete routine care ED process, throughput time for noncomplex patients is reduced to 145 minutes, a 7 percent produc-tivity gain. An analyst from the VVH finance department (a member of the project team) is able to demonstrate that the capital and software costs for the
EndDischarge
Patientarrives
at the ED
IntensiveED care
Admitting
Triage–clinical
Complexity
Exam/treatment
Nursehistory/
symptoms
Low
High
Waiting
Waiting
Focu
s
Note: Created with Microsoft Visio.
EXHIBIT 11.9VVH Emergency
Department (ED) Process
Map: Focus on Waiting and
History
Chapter 11: Process Improvement and Pat ient F low 311
tablet computers will be recovered within 12 months by the improvement in patient flow.
This phase of the project used three of the basic process improvement tools discussed in this chapter:
• Have the customer (patient) do it.• Provide quality at the source.• Gain information feedback and real-time control.
Although the process improvements already undertaken have had a visible impact on flow in the ED, the team believes more improvements are possible. Bottlenecks plague the process, as evidenced by two waiting lines, or queues: (1) the waiting room queue, where patients wait before being moved to an exam room, and (2) the most visible queue for routine patients, the discharge area, where patients occasionally must stand because all of the area’s chairs are occupied. In the discharge area, patients wait a significant amount of time for final instructions and prescriptions.
The theory of constraints suggests that the bottleneck be identified and optimized. However, alleviating or eliminating the patient examination and treatment or discharge bottlenecks would require significant changes in a long-standing process. Because this process improvement step seems to have the probability of a high payoff but would be a significant departure from existing practice, the team moves to phase III of the project and uses simula-tion to model different options to improve patient flow in the examination/treatment and discharge processes.
Discharge
m
%
#
Cycletime
FTEs
First-time
correct
Exam/treatment
m
%
#
Cycletime
FTEs
First-time
correct
Patients
#/hr
12Arrival
rate
0 min5 min 9 min
30 min
Hr
Hr
Hr
Hr
20 min
Nurse(history)
m
%
#
20
nm
2
Cycletime
FTEs
First-time
correct
Admitting(insurance)
IntensiveED care
m
%
#
9
nm
2
Cycletime
FTEs
First-time
correct
Triage
m
%
#
5
nm
1
Cycletime
FTEs
First-time
correct
EXHIBIT 11.10 VVH Emergency Department (ED) Value Stream Map: Focus on Waiting and History
Note: Created with eVSM software, a Microsoft Visio add-on from GumshoeKI, Inc. FTE = full-time equivalent; nm = number of patients in this step of the process.
Healthcare Operat ions Management312
Phase IIIFirst, the team reviews the basic terminology of simulation.
• An entity is what flows through a system. Here, the entity is the patient. However, in other systems, the entity can be materials (e.g., blood sample, drug) or information (e.g., diagnosis, billing code). Entities usually have attributes that affect their flow through the system (e.g., male/female, acute/chronic condition).
• Each individual process in the system transforms (adds value to) the entity being processed. Each process takes time and consumes resources, such as staff, equipment, supplies, and information.
• Time and resource use can be defined as an exact value (e.g., ten minutes) or a probability distribution (e.g., normal—mean, standard deviation). Most healthcare tasks and processes do not require the same amount of time each time they are performed—they require a variable amount of time. These variable usage rates are best described as probability distributions. (Chapter 7 discusses probability distributions in detail.)
• The geographic location of a process is called a station. Entities flow from one process to the next via routes. The routes can branch out on the basis of decision points in the process map.
• Finally, because a process may not be able to handle all incoming entities in a timely fashion, queues occur at each process and can be measured and modeled.
The team next develops a process map and simulation model for routine patient flow (exhibit 11.11) in the ED using Arena simulation software (see the companion website for links to videos detailing this model and its operation). The team focuses on routine patients rather than those requiring intensive emergency care because of the high proportion of routine patients seen in the department. Routine patients are checked in and their self-recorded his-
tory and presenting complaint(s) verified by a nurse. Then, patients move to an exam/treatment room and, finally, to the discharge area. Of the ten patients who arrive at the ED per hour, eight follow this process.
Next, to build a simulation model that accurately reflects this process, the team needs to determine the probability distributions of treatment time, admitting time, nurse history time, discharge time, and arrival rate for routine patients. To determine these probability distributions, team members collect data on time of arrival in the department and time to perform each step in the routine patient care process.
On the web at ache.org/books/OpsManagement3
Chapter 11: Process Improvement and Pat ient F low 313
Probability distributions are determined using the input analyzer func-tion in Arena. Input Analyzer takes raw input data and finds the best-fitting probability distribution for them. Exhibit 11.12 shows the output of Input Analyzer for 500 observations of treatment time for ED patients requiring routine care. Input Analyzer suggests that the best-fitting probability distribu-tion for these data is triangular, with a minimum of 9 minutes, mode of 33 minutes, and maximum of 51 minutes.
Patientarrives
Triage
Admitting
Patienthistory
Nursehistory
Exam andtreatment
Leave ED
IntensiveED care
Waiting room
Discharge area
Discharge
False
TrueComplexity
EXHIBIT 11.11 VVH Emergency Department (ED) Initial State Simulation Model
Note: Created with Arena simulation software.
Treatment Time (minutes)
12
24
Num
ber o
f Occ
urre
nces
159 33
EXHIBIT 11.12 Examination and Treatment Time Probability Distribution: Routine Emergency Department Patients
Healthcare Operat ions Management314
The remaining data are analyzed in the same manner, and the following best-fitting probability distributions are determined:
• Emergency routine patient arrival rate—exponential (7.5 minutes between arrivals)
• Triage time—triangular (2, 5, 7 minutes)• Admitting time—triangular (3, 8, 15 minutes)• Patient history time—triangular (15, 20, 25 minutes)• Nurse history time—triangular (5, 11, 15 minutes)• Exam/treatment time—triangular (14, 36, 56 minutes)• Discharge time—triangular (9, 19, 32 minutes)
The Arena model simulation is based on 12-hour intervals (2:00 p.m. to 2:00 a.m.) and replicated 100 times. Note that increasing the number of replications decreases the half-width and, therefore, gives tighter confidence intervals. The number of replications needed depends on the desired confi-dence interval for the outcome variables. However, as the model becomes more complicated, more replications take more simulation time; this model is fairly simple, so 100 replications take little time and are sufficient for this purpose.
Most simulation software, including Arena, is capable of using different arrival rate probability distributions for different times of the day and days of the week, allowing for varying demand patterns. However, the team believes that this simple model using only one arrival rate probability distribution represents the busiest time for the ED, having observed that by 2:00 p.m. on weekdays no queues are created in either the waiting room or the discharge area.
The results of the simulation are reviewed by the team and compared with actual data and observations to ensure that the model is, in fact, simulating the reality of the ED. The team is satisfied that the model accurately reflects reality.
The focus of this simulation is the queuing that occurs in both the waiting room and the discharge area and the total time in the system. Exhibit 11.13 shows the results of this base (current status) model. On average, a patient spends 2.4 hours in the ED.
The team next examines the discharge process in depth because patient waiting time is greatest there. The ED has two rooms devoted to discharge and uses two nurses to handle all discharge tasks, such as making sure prescriptions are given and home care instructions are understood. However, because of the limited number of nurses and exam rooms, queuing is inevitable. In addition, the patient treatment information must be handed off from the treatment team to the discharge nurse. The process improvement team simulates having the discharge process carried out by the examination and treatment team. Because the examination and treatment team knows the patient information, the handoff task can be eliminated. The team estimates that this change will save about five
Chapter 11: Process Improvement and Pat ient F low 315
minutes. To ensure that this is the correct outcome, team members simulate the new system by eliminating discharge as a separate process.
Team members estimate the probability distribution of the combined exam/treatment/discharge task by first estimating the probability distribution for handoff as triangular (4, 5, 7 minutes). The team uses Input Analyzer to simulate 1,000 observations of exam/treatment time, discharge time, and handoff time using the previously determined probability distributions for each. For each observation, it adds exam/treatment time to discharge time and subtracts handoff time to find total time. Input Analyzer finds the best-fitting
Replications: 100
TotalTime
WaitingTime
Routine patient
AverageMinimumAverage
MinimumValue
MaximumValue
MaximumAverage
Half-Width
2.4207 1.7953 1.2004 5.24483.40820.08
Admitting queueDischarge queueExam and treatment queueNurse history queueTriage queue
AverageMinimumAverage
MinimumValue
MaximumValue
MaximumAverage
Half-Width
0.005269300.39720.3382
0.017645410.06437939
0.000485530.064166920.041671220.002727150.01703829
0.000.000.000.000.00
0.22352.05312.57770.36940.6506
0.016686100.88651.1956
0.053097330.1402
0.000.260.380.010.05
WaitingTime
Admitting queueDischarge queueExam and treatment queueNurse history queueTriage queue
AverageMinimumAverage
MinimumValue
MaximumValue
MaximumAverage
Half-Width
0.034580322.24812.19300.1136
0.5394
0.002670400.28880.2062
0.012984610.1145
0.000.000.000.000.00
2.000013.000022.0000
5.000010.0000
0.10015.1713
9.44080.40691.7216
0.000.260.380.010.05
InstantaneousUtilization
Discharge nurse 1Discharge nurse 2Exam room 1Exam room 2Exam room 3Exam room 4Financial clerk 1Financial clerk 2History nurse 1History nurse 2Triage nurse
AverageMinimumAverage
MinimumValue
MaximumValue
MaximumAverage
Half-Width
0.82850.83600.84410.83290.81820.80750.46150.45800.52940.52400.6267
0.67150.66730.62530.65480.53580.61350.33200.32860.38860.39370.4861
0.000.000.000.000.000.000.000.000.000.000.00
1.00001.00001.00001.00001.00001.00001.00001.00001.00001.00001.0000
0.89720.91050.94970.92970.92000.91560.56360.58230.67960.71070.8373
0.010.010.010.010.020.020.010.010.010.010.01
Time Unit: Hours
Queue
Resource
Time
Other
Usage
EXHIBIT 11.13VVH Emergency Department Initial State Simulation Model Output
Note: Created with Arena simulation software.
Healthcare Operat ions Management316
probability distribution for the total time for the new process as triangular (18, 50, 82 minutes).
The team simulates the new process and finds that, under the new system, patients will spend an average of 2.95 hours in the ED—increasing the time spent there. However, it will eliminate the need for discharge rooms. The team decides to investigate the impact of converting the former discharge rooms to exam rooms and runs a new simulation incorporating this change (exhibit 11.14). The result of this simulation is shown in exhibit 11.15. Both the number of patients in the waiting room (examination and treatment queue) and the amount of time they wait are reduced substantially. The staffing levels are not changed, as the discharge nurses are now treatment nurses. Physician staffing also is not increased, as some delay inside the treatment process itself has always existed due to the need to wait for lab results, resulting in a delayed final physician diagnosis. Having more patients available for treatment fills this lab delay time for physicians to perform patient care.
Patientarrives
Triage
Admitting
Patienthistory
Nursehistory
Exam andtreatment
Leave ED
IntensiveED care
Waiting room
False
TrueComplexity
EXHIBIT 11.14 VVH Emergency
Department (ED) Proposed
Change Simulation
Model
Note: Created with Arena simulation software.
Chapter 11: Process Improvement and Pat ient F low 317
4:11:35 PM
Replications: 100
TotalTime
WaitingTime
Routine patient
AverageMinimumAverage
MinimumValue
MaximumValue
MaximumAverage
HalfWidth
1.8376 1.5459 1.0063 4.59892.87290.05
Admitting queueExam and treatment
and discharge queueNurse history queueTriage queue
AverageMinimumAverage
MinimumValue
MaximumValue
MaximumAverage
HalfWidth
0.00519434
0.20390.017917520.06635691
0.00041085
0.001972930.002445000.01863876
0.00
0.000.000.00
0.2235
2.29430.34170.8065
0.01364095
1.11050.07537764
0.2547
0.00
0.040.000.01
WaitingTime
Admitting queueExam and treatment
and discharge queueNurse history queueTriage queue
AverageMinimumAverage
MinimumValue
MaximumValue
MaximumAverage
HalfWidth
0.03400433
1.35710.10980.5629
0.00218978
0.008384960.01120623
0.1227
0.00
0.000.000.00
3.0000
19.00004.0000
11.0000
0.0946
7.82880.57162.5046
0.00
0.310.010.08
InstantaneousUtilization
Exam room 1Exam room 2Exam room 3Exam room 4Exam room 5Exam room 6Financial clerk 1Financial clerk 2History nurse 1History nurse 2Triage nurse
AverageMinimumAverage
MinimumValue
MaximumValue
MaximumAverage
HalfWidth
0.78270.76440.76260.74780.78590.80300.46060.45290.52360.51540.6226
0.54050.54680.55770.49840.54200.49900.32500.29680.36420.34030.4742
0.000.000.000.000.000.000.000.000.000.000.00
1.00001.00001.00001.00001.00001.00001.00001.00001.00001.00001.0000
0.93030.91030.90520.89930.93130.94720.59850.6119
0.67660.69820.8185
0.020.020.020.020.020.020.010.010.010.010.02
Time Unit: Hours
Values Across All Replications
February 8, 2012Category Overview
VVH Emergency
Entity
Queue
Resource
Time
Time
Other
Usage
EXHIBIT 11.15VVH Emergency Department (ED) Proposed Change Simulation Model Output
Note: Created with Arena simulation software.
Healthcare Operat ions Management318
The most significant improvement resulting from the process improve-ment initiative is that total patient throughput time now averages 1.84 hours (110 minutes). This 33 percent reduction in throughput time exceeds the team’s goal and is celebrated by VVH’s senior leadership. The summary of process improvement steps is displayed in exhibit 11.16.
Conclusion
The theory of swift, even flow provides a framework for process improvement and increased productivity. The efficiency and effectiveness of a process increase as the speed of flow through the process increases and the variability associated with that process decreases.
The movement of patients in a healthcare facility is one of the most critical and visible processes in healthcare delivery. Reducing flow time and variation in processes results in a number of benefits, including the following:
• Patient satisfaction increases.• Quality of clinical care improves as patients have reduced waits for
diagnosis and treatment.• Financial performance improves.
This chapter demonstrates many approaches to the challenges of reducing flow time and process variation. Starting with the straightforward process map, many improvements can be found immediately by inspection. In other cases, the powerful tool of computer-based discrete event simulation can provide a road map to sophisticated process improvements.
Ensuring quality of care is another critical focus of healthcare organiza-tions. The process improvement tools and approaches in this chapter may be
Process Improvement Change Throughput Time, Routine Patients
Baseline, before any improvement 165 minutes
Combine admitting functions 155 minutes
Patients enter their own history into computer
145 minutes
Combine discharge tasks into examination and treatment process, and convert discharge rooms to treatment rooms
110 minutes
EXHIBIT 11.16Summary of
VVH Emergency Department Throughput
Improvement Project
Chapter 11: Process Improvement and Pat ient F low 319
used to reduce process variation and eliminate errors. Healthcare organizations must employ the disciplined approach described in this chapter to achieve the needed improvements in flow and quality.
Discussion Questions
1. How do you determine which process improvement tools should be used in a given situation? What is the cost and return of each approach?
2. Which process improvement tool can have the most powerful impact, and why?
3. How can barriers to process improvement, such as staff reluctance to change, lack of capital, technological barriers, or clinical practice guidelines, be overcome?
4. How can the electronic health record be used to make significant process improvements for both efficiency and quality increases?
5. Describe several places or times in your organization where people or objects (paperwork, tests, etc.) wait in line. How do the characteristics of each example differ?
Exercises
1. Access the National Guideline Clearinghouse (www.guideline.gov/) and translate one of the guidelines described into a process map. Add decision points and alternative paths to account for unusual issues that might occur in the process. (Hint: Use Microsoft Visio or another similar application to complete this exercise.)
2. Access the following process maps on the companion website:• Operating Suite• Cancer Treatment ClinicUse basic improvement tools, theory of constraints, Six Sigma, or Lean tools to determine possible process improvements.
3. The hematology lab manager has received complaints that the turnaround time for blood tests is too long. Data from the past month show that the arrival rate of blood samples to one technician in the lab is five per hour and the service rate is six per hour. Using queuing theory, and assuming that (a) both rates are exponentially distributed and (b) the lab is at steady state, determine the following measures:
On the web at ache.org/books/OpsManagement3
Healthcare Operat ions Management320
• Capacity utilization of the lab • Average number of blood samples in the lab • Average time that a sample waits in the queue • Average number of blood samples waiting for testing • Average time that a blood sample spends in the lab
References
Butterfield, S. 2007. “A New Rx for Crowded Hospitals: Math.” ACP Hospitalist. Published December. www.acphospitalist.org/archives/2007/12/math.htm#sb1.
Clark, J. J. 2005. “Unlocking Hospital Gridlock.” Healthcare Financial Management 59 (11): 94–104.
Cooper, R. B. 1981. Introduction to Queueing Theory, 2nd edition. New York: North-Holland.Deming, W. E. 1998. “The Deming Philosophy.” Deming-Network. Accessed June 9, 2006.
http://deming.ces.clemson.edu/pub/den/deming_philosophy.htm.Devaraj, S., T. T. Ow, and R. Kohli. 2013. “Examining the Impact of Information Technol-
ogy and Patient Flow on Healthcare Performance: A Theory of Swift and Even Flow (TSEF) Perspective.” Journal of Operations Management 31 (4): 181–92.
Litvak, E. 2003. “Managing Patient Flow: Smoothing OR Schedule Can Ease Capacity Crunches, Researchers Say.” OR Manager 19 (November): 1, 9–10.
McManus, M., M. Long, A. Cooper, and E. Litvak. 2004. “Queuing Theory Accurately Models the Need for Critical Care Resources.” Anesthesiology 100 (5): 1271–76.
Rockwell Automation. 2016. Arena home page. Accessed September 21. www.arenasimula-tion.com/.
Rodi, S. W., M. V. Grau, and C. M. Orsini. 2006. “Evaluation of a Fast Track Unit: Align-ment of Resources and Demand Results in Improved Satisfaction and Decreased Length of Stay for Emergency Department Patients.” Quality Management in Healthcare 15 (3): 163–70.
Sayah, A., M. Lai-Becker, L. Kingsley-Rocker, T. Scott-Long, K. O’Connor, and L. F. Lobon. 2016. “Emergency Department Expansion Versus Patient Flow Improvement: Impact on Patient Experience of Care.” Journal of Emergency Medicine 50 (2): 339–48.
Schmenner, R. W. 2004. “Service Businesses and Productivity.” Decision Sciences 35 (3): 333–47.
. 2001. “Looking Ahead by Looking Back: Swift, Even Flow in the History of Manu-facturing.” Production and Operations Management 10 (1): 87–96.
Schmenner, R. W., and M. L. Swink. 1998. “On Theory in Operations Management.” Journal of Operations Management 17 (1): 97–113.
Simul8 Corporation. 2016. “Process Simulation Software.” Accessed September 21. www.simul8.com/.
Chapter 11: Process Improvement and Pat ient F low 321
Further Reading
Goldratt, E. M., and J. Cox. 1986. The Goal: A Process of Ongoing Improvement. New York: North River Press.
Kelton, W., R. Sadowski, and N. Swets. 2009. Simulation with Arena. New York: McGraw-Hill.
CHAPTER
323
SCHEDULING AND CAPACITY MANAGEMENT
Operations Management in Action
Once upon a time, a patient at Second Street Family Practice in Auburn, Maine, had to wait from 60 to 90 days to be seen for a routine check-up. Then, when the day of the appointment finally arrived, the patient might wait nearly 20 minutes in the waiting room and another 20 for the exam to begin. But thanks to strong leadership, impressive teamwork, and effective tools, patients wanting care from Second Street, even routine check-ups, are now seen the same day they call. The average time patients spend flipping through magazines in the waiting room has dropped to around seven minutes; the exam room wait is down to eight. What’s more, staff say they like the new system much better, and patient surveys show that about 90 percent of patients notice and are pleased with the changes as well.
[Clinic leadership], who had been reading and learning about advanced access scheduling, recognized it as the antidote for their frustrations. Devel-oped by Mark Murray, MD, and Cath-erine Tantau, RN, consultants in Sacra-mento, California, and promoted by [the Institute for Healthcare Improvement (IHI)] in its office practice programs and on its website, advanced access uses queuing theory to reengineer the stan-dard appointment scheduling system,
12OVE RVI EW
Matching the supply of goods or services to the demand for those
goods or services is a basic operational problem. In a manufactur-
ing environment, inventory can be used to respond to fluctuations
in demand. In the healthcare environment, safety stock can be used
to respond to fluctuations in demand for supplies (see chapter 13),
but stocking healthcare services is not possible. Therefore, capac-
ity must be matched to demand. If capacity is greater than demand,
resources are underutilized and costs are high. Idle staff, equipment,
or facilities increase organizational costs without increasing revenues.
If capacity is lower than demand, patients endure long waits or find
another provider.
To match capacity to demand, organizations can use demand-
influencing strategies or capacity management strategies. Pricing
and promotions are often deployed to influence demand and demand
timing; however, this strategy typically is not viable for healthcare
organizations. In the past, many clinics, hospitals, and health systems
used the demand-leveling strategy of appointment scheduling; more
recently, many have moved to advanced-access scheduling. Capac-
ity management strategies allow the organization to adjust capacity
to meet fluctuating demand; they include using part-time or on-call
employees, cross-training staff, and assigning overtime. Effective and
efficient scheduling of patients, staff, equipment, facilities, or jobs
can help leaders match capacity to demand and ensure that scarce
healthcare resources are used to their fullest extent.
This chapter outlines issues and problems faced in scheduling
and discusses tools and techniques that can be employed in schedul-
ing patients, staff, equipment, facilities, or jobs. Topics covered here
related to scheduling tools and approaches include
• hospital census and resource loading,
• staff scheduling,
• job and operation scheduling and sequencing rules,(continued)
Healthcare Operat ions Management324
leaving the majority of slots on any given day open for patients who call that day.
The benefits of advanced access go beyond improved scheduling, says IHI director Marie Schall. “It improves quality and continuity,” she says. “People can get problems checked sooner rather than later, and they see the same provider vir-tually every time. We know that continuity contributes to better overall quality.” Schall says that through its Breakthrough Series Collaboratives on Reducing Delays and Waiting Times and its IMPACT network, as well as its work with the Veterans Health Administration on improving access to care, IHI has worked with about 3,000 practices to introduce advanced access.
Source: Excerpted from IHI (2012).
Hospital Census and Rough-Cut Capacity Planning
For many healthcare organizations, the admittance rate and number of occupied beds provide a good indication of the demands being placed on the system. For hospitals, these numbers often can be measured on the basis of the overall patient census. Most hospitals report their census daily and hourly to manage the available beds in the system. However, what many healthcare organizations fail to understand is that the census also provides a view into the resource needs to appropriately staff a system. Exhibit 12.1 shows a three-month view of a census for Vincent Valley Hospital and Health System (VVH). The pattern is
OVE RVI EW (Continued)
• patient appointment scheduling models, and
• advanced-access patient scheduling.
The scheduling of patients is a unique, but important,
subproblem of patient flow. Since the mid-twentieth century, much
patient care delivery has moved from the inpatient setting to the
ambulatory clinic. Because this trend is likely to continue, matching
clinic capacity to patient demand becomes an even more critical
operating skill. Beyond operational considerations, if capacity
management can be deployed to meet a patient’s desired sched-
ule, marketplace advantage can be gained. Therefore, this chapter
focuses on advanced access (same-day scheduling) for ambulatory
patients. Related topics covered in this chapter include
• advantages of advanced access,
• implementation steps, and
• metrics for tracking the operations of advanced-access
scheduling systems.
Many of the operations tools and strategies detailed in
earlier chapters are demonstrated here to show how to optimize
the operations of an advanced-access clinic.
Chapter 12: Schedul ing and Capacity Management 325
remarkably similar to most hospitals in that a large amount of variance exists in the patient population on a daily basis. This variance can become magnified when observing the census on an hourly basis.
Rough-cut capacity planning is the process of converting the overall production plan into capacity needs for key resources. For a hospital, it means planning key resources for the demand schedule. While the day-to-day demand in healthcare systems is highly variable, the aggregate demand on a month-to-month basis can be predicted more precisely. When planning resources, hospital leaders generally consider two types of labor resources: full-time staff and contractors. By examining the census, an administrator should be able to determine, on an aggregate basis, the number of contactors needed dur-ing high-volume months. This approach is an example of rough-cut capacity planning. But many healthcare systems leave this planning until the need for additional resources arises. Because they have not paid enough attention to the required staffing levels to meet demand on an aggregate basis, these systems are forced to spend unnecessary costs to meet demand.
A hospital administrator may also use the daily census to assist in prepar-ing workforce schedules on a weekly or daily basis. Exhibit 12.2 shows a spike in the system at VVH occurring from hour 13 to hour 19, which in most situa-tions is the middle of the day. Many hospitals still schedule staff using standard morning, evening, and night shifts. Under that staffing model, VVH doctors and nurses are ending their shifts at the time of maximum demand on the sys-tem, resulting in increased potential for errors in handing off patients to new doctors, long patient wait times, and untimely completion of medical records.
A major cost savings can be gained for hospitals and clinics by simply matching the resources to the demand patterns in the system. In this case, staffing many doctors and nurses to overlap the peak times in the middle of the day is ideal.
Rough-cut capacity planningThe process of converting the overall production plan into capacity needs for key resources.
Time
Num
ber o
f Pat
ient
s
EXHIBIT 12.1Daily Census at VVH
Healthcare Operat ions Management326
From an operations perspective, this problematic issue is easy to fix. However, in practice, several obstacles may emerge, such as contractual terms agreed to by unions and conflicting physician block scheduling.
Staff Scheduling
For minor schedule-optimization problems, where demand is reasonably known and staffing requirements can be estimated with certainty, mathematical pro-gramming (chapter 6) may be used to optimize staffing levels and schedules. As these problems increase in complexity, however, developing and applying a mathematical programming model becomes time and cost prohibitive. In those cases, simulation can be used to answer what-if scheduling questions, such as “What if we added a nurse?” or “What if we cross-trained employees?” See chapter 11 and the advanced-access section of this chapter for examples of these types of applications.
A simple example of this type of issue, and how to solve it using linear programming, is illustrated in the paragraphs that follow. (For solutions to more complex staffing issues using linear programming, see Matthews [2005] and Trabelsi, Larbi, and Alouane [2012].)
Solving Riverview Clinic Urgent Care StaffingNurses who staff Riverview Urgent Care Clinic (UCC), the after-hours urgent care facility of VVH’s Riverview Clinic, have been complaining about their schedules. They would like to work five consecutive days and have two con-secutive days off every seven days. Different nurses prefer different days off
40
50
60
30
20
10
0
20161284 2319151173 2218141062 211713951 24
Hour
Num
ber o
f Pat
ient
s
EXHIBIT 12.2Hourly Census at VVH in One
Patient Care Unit
Chapter 12: Schedul ing and Capacity Management 327
and believe that their preferences should be accommodated on the basis of seniority, whereby the most senior nurses are granted their desired days off first.
Riverview UCC collects patient demand data by day of the week and knows how many nurses should be on staff each day to meet demand. River-view UCC managers want to minimize nurse payroll while reducing the nurses’ complaints about their schedules. They decide to apply linear programming to help determine a solution for this two-pronged problem. Target staffing levels and salary expense are shown in exhibit 12.3.
First, Riverview UCC needs to determine how many nurses should be assigned to each of the seven possible schedules (Monday and Tuesday off, Tuesday and Wednesday off, etc.).
The goal is to minimize weekly salary expense, and the objective func-tion is set up as follows.
Minimize:
($320 × Su) + ($240 × M) + ($240 × Tu) + ($240 × W) + ($240 × Th) + ($240 × F) + ($320 × Sa),
where Su is the number of nurses required on staff for Sundays, M is nurses needed Mondays, Tu is nurses needed Tuesdays, W is nurses needed Wednes-days, Th is nurses needed Thursdays, F is nurses needed Fridays, and Sa is nurses needed Saturdays.
The constraints are the following:
• The number of nurses scheduled each day must be greater than or equal to the number of nurses needed each day.
Su ≥ 5M ≥ 4Tu ≥ 3W ≥ 3Th ≥ 3F ≥ 4Sa ≥ 6
Linear programmingA mathematical technique used to find the optimal solution to a linear problem given a set of constrained resources.
Sunday Monday Tuesday Wednesday Thursday Friday Saturday
Nurses needed per day
5 4 3 3 3 4 6
Salary and benefits per nurse-day
$320 $240 $240 $240 $240 $240 $320
EXHIBIT12.3Riverview UCC Target Staffing Level and Salary Expense
Healthcare Operat ions Management328
• The number of nurses assigned to each schedule, where the schedules are denoted by a letter of the alphabet from A to G, must be greater than zero and an integer.
Number of nurses for schedule A (B, C, D, E, F, or G) ≥ 0Number of nurses for schedule A (B, C, D, E, F, or G) = integer
Exhibit 12.4 shows the Excel Solver setup of this problem.As illustrated in exhibit 12.5, Solver finds that the Riverview UCC needs
to employ six full-time equivalent nurses and should assign one nurse to sched-ules A, B, C, and D; two nurses to schedule E; and no nurses to schedules F and G. The total salary expense with this optimal schedule is calculated as follows.
Minimize:
($320 × 5) + ($240 × 4) + ($240 × 4) + ($240 × 4) + ($240 × 3) + ($240 × 4) + ($320 × 6) = $8,080 per week.
Next, Riverview UCC needs to determine which nurses to assign to which schedule on the bases of their preferences and seniority. Each nurse is asked to rank schedules A through E in order of preference. The nurses’ prefer-ences on a scale of 1 to 5, with 5 being the most preferred schedule, are then weighted by a seniority factor. Riverview UCC uses as the weighting factor the number of years a particular nurse has worked at the facility compared with the number of years the most senior nurse has worked there.
EXHIBIT 12.4Initial Excel
Solver Setup of Riverview UCC
Optimization
Chapter 12: Schedul ing and Capacity Management 329
The goal is to maximize the nurses’ total weighted preference scores (WPSs), and the objective function is set up as follows.
Maximize:
Mary’s WPS + Anne’s WPS + Susan’s WPS + Tom’s WPS + Cathy’s WPS + Jane’s WPS
The constraints are the following:
• The assignment is binary, meaning that each nurse must be either assigned or not assigned to a particular schedule.
Mary assigned to schedule A (B, C, D, or E) = 0 or 1Anne assigned to schedule A (B, C, D, or E) = 0 or 1Susan assigned to schedule A (B, C, D, or E) = 0 or 1Tom assigned to schedule A (B, C, D, or E) = 0 or 1Cathy assigned to schedule A (B, C, D, or E) = 0 or 1Jane assigned to schedule A (B, C, D, or E) = 0 or 1
• The number of nurses assigned to each schedule must adhere to the requirements established earlier.
Number of nurses assigned to schedule A (B, C, or D) = 1Number of nurses assigned to schedule E = 2
• Each nurse can only be assigned to one schedule.Mary (Anne, Susan, Tom, Cathy, or Jane) A + B + C + D + E = 1
Exhibit 12.5 shows the Excel setup of this problem.
EXHIBIT 12.5Riverview UCC Initial Solver Solution and Schedule Preference Setup
Healthcare Operat ions Management330
As shown in exhibit 12.6, Solver finds that Mary should be assigned to schedule D (her second choice), Anne to schedule E (her first choice), Susan to schedule C (her first choice), Tom to schedule E (his first choice), Cathy to schedule B (her second choice), and Jane to schedule A (her first choice). All of the nurses now have two consecutive days off every seven days and are assigned to either their first or their second choice of schedule. Note that even this simple problem has 20 decision variables and 41 constraints.
Job and Operation Scheduling and Sequencing Rules
Master production scheduling (MPS) is a technique used in most production-oriented environments that has direct application to the healthcare operations space. The concept behind MPS is to forecast needs for the future and build a schedule to fit those needs.
When building a master production schedule, time fences are set up to help avoid disruptions in the schedule. Typically, time fences depicted as “frozen,” “slushy,” or “liquid” are established to give the scheduling department informa-tion as to when a schedule can be adjusted. For example, a surgery center may aim for a frozen schedule for surgeries scheduled during the following week; a slushy schedule, where up to 20 percent may be adjusted, for surgeries sched-uled two to three weeks in advance; and a liquid, or open, schedule for surgeries scheduled one month or more into the future. By freezing a schedule for a set period, the surgery center is able to avoid unnecessary interruptions. Interruptions in scheduling eventually lead to fewer surgeries for a variety of reasons, including the variance in time related to surgeries, extra setup time of surgery rooms, and general impact of changing surgeries at the last minute. To handle urgent sur-geries when using MPS, a hospital should keep some capacity available for these situations. The net effect of this approach is increased output from the surgery because the variability associated with urgent surgeries does not affect the MPS.
EXHIBIT 12.6 Riverview UCC
Final Solver Solution for
Individual Schedules
Chapter 12: Schedul ing and Capacity Management 331
Job and operation scheduling views the problem of how to sequence a pool of jobs (or patients) through a particular operational activity. For example, a clinic laboratory constantly receives patient blood samples that need to be tested, and it must determine in what order it should conduct those tests. Similarly, a hospital typically has many patients waiting for their surgery to be performed, and it needs to decide the order in which those surgeries should occur.
The simplest sequencing problems consist of a pool of jobs waiting for only one resource to become available. Sequencing of those jobs is usually based on a desire to meet due dates (time at which the job is expected to be complete) by minimizing the number of jobs that are late, minimizing the average amount of time by which jobs are late, or minimizing the maximum late time of any job. Also desirable is to minimize the time jobs spend in the system or average completion time.
Various sequencing rules, also known as the queuing priority, may be used to schedule jobs through the system. Commonly used rules include the following:
• First come, first served (FCFS)—Jobs are sequenced in the same order in which they arrive.
• Shortest processing time (SPT)—The job that takes the least amount of time to complete is first, followed by the job that takes the next least amount time, and so on.
• Earliest due date (EDD)—The job with the earliest due date is first, followed by the job with the next earliest due date, and so on.
• Slack time remaining—The job with the least amount of slack (time until due date or processing time) is first, followed by the job with the next least amount of slack time, and so on.
• Critical ratio—The job with the smallest critical ratio (time until due date or processing time) is first, followed by the job with the next smallest critical ratio, and so on.
When only one resource or operation is available through which the jobs may be processed, the SPT rule minimizes average completion time, and the EDD rule minimizes average lateness and maximum lateness. However, no single rule accomplishes both objectives. When jobs (or patients) must be processed via a series of resources or operations, with different possible sequencing at each, the situation becomes complex and applying a particular rule does not result in the same outcome for the entire system as for the single resource. Simulation may be used to evaluate these complex systems and helps determine optimum sequencing.
For a busy resource, the SPT rule is often applied. It allows completion of a greater number of jobs in a shorter amount of time than do the other rules, but it may result in some jobs with long completion times never being
Sequencing rulesHeuristic rules that indicate the order in which jobs are processed from a queue. Also known as queuing priority.
Healthcare Operat ions Management332
finished. To alleviate this problem, the SPT rule may be used in combination with other rules. For example, in some emergency departments (EDs), less severe cases (those with a shorter processing time) are separated from more severe cases and fast-tracked to free up examination rooms quickly.
For time-sensitive operational activities, in which lateness is not toler-ated, the EDD rule is appropriate. Because it is the easiest to apply, the FCFS rule is typically used when the resource has excess capacity and no jobs will be late. In a Lean environment, sequencing rules become irrelevant because the ideal size of the pool of jobs is reduced to one and a kanban system (a form of FCFS) can be used to pull jobs through the system (chapter 10).
Vincent Valley Hospital and Health System Laboratory Sequencing RulesA technician recently has left the laboratory at VVH, and the lab manager, Jessica Simmons, does not believe she can find a qualified replacement for at least one month. This situation has greatly increased the workload in the lab, and physicians have been complaining that their requested blood work is not being completed in a timely manner.
In the past, Jessica has divided the blood testing among the technicians and requested they complete the tests on an FCFS basis. She is now consider-ing a different sequencing rule to satisfy the physicians. In anticipation of this change, she has asked each physician to enter a desired completion time on each request for blood testing. To investigate the effects of changing the sequenc-ing rules, she analyzes, under various scheduling rules, the first five requests completed by one of the technicians. For five jobs, 120 sequences are possible for their completion. Exhibit 12.7 shows the time to complete each blood work sample and the time of completion requested by the physician.
Exhibit 12.8 indicates the order in which jobs will be processed and results under different sequencing rules, and exhibit 12.9 compares the various sequencing rules. The FCFS rule performs poorly on all measures. The SPT
SampleProcessing Time
(minutes)Due Time
(minutes from now) Slack CR
A 50 100 100 – 50 = 50 100 50 = 2.00
B 100 160 160 – 100 = 60 160 100 = 1.60
C 20 50 50 – 20 = 30 50 20 = 2.50
D 80 120 120 – 80 = 40 120 180 = 1.50
E 60 80 80 – 60 = 20 80 60 = 1.33
Note: CR = critical ratio.
EXHIBIT 12.7VVH Laboratory
Blood Test Information
Chapter 12: Schedul ing and Capacity Management 333
SequenceStart Time
Processing Time
Completion Time Due Time Tardiness
FCFSA 0 50 50 100B 50 100 150 160C 150 20 170 50 170 – 50 = 120D 170 80 250 120 250 – 120 = 130E 250 60 310 80 310 – 80 = 230
Average 186 (120 + 130 + 230) 5 = 96
SPTC 0 20 20 50A 20 50 70 100E 70 60 130 80 130 – 80 = 50D 130 80 210 120 210 – 120 = 90B 210 100 310 160 310 – 160 = 150
Average 148 (50 + 90 + 150) 5 = 58
EDDC 0 20 20 50E 20 60 80 80A 80 50 130 100 130 – 100 = 30D 130 80 210 120 210 – 120 = 90B 210 100 310 160 310 – 160 = 150
Average 150 (30 + 90 + 150) 5 = 54
STRE 0 60 60 80C 60 20 80 50 80 – 50 = 30D 80 80 160 120 160 – 120 = 40A 160 50 210 100 210 – 100 = 110B 210 100 310 160 310 – 160 = 150
Average 164 (30 + 40 + 110 + 150) 5 = 66
CRE 0 60 60 80D 60 80 140 120 140 – 120 = 20B 140 100 240 160 240 – 160 = 80A 240 50 290 100 290 – 100 = 190C 290 20 310 50 310 – 50 = 260
Average 208 (20 + 80 + 190 + 260) 5 = 110
Note: All times shown in exhibit are in minutes. CR = critical ratio; EDD = earliest due date; FCFS = first come, first served; SPT = shortest processing time; STR = slack time remaining.
EXHIBIT 12.8VVH Laboratory Blood Test Sequencing Rules
Healthcare Operat ions Management334
rule minimizes average completion time, and the EDD rule minimizes aver-age tardiness. Under these two rules, three jobs are tardy and the maximum tardiness is 150 minutes. After considering these results, Jessica implements the EDD rule for laboratory blood tests to minimize the number of tardy jobs and the average tardiness of jobs. She hopes adopting this rule reduces physician complaints until a new technician can be hired.
Patient Appointment Scheduling Models
Appointment scheduling models attempt to minimize patient waiting time while maximizing utilization of the resource (clinician, machine, etc.) the patients are waiting to access. Soriano (1966) classifies appointment schedul-ing systems into four basic types: block appointment, individual appointment, mixed block-individual appointment, and other.
A block appointment scheme schedules the arrival of all patients at the start of a clinic session. Patients are usually seen FCFS, but other sequencing rules can be used in block appointment scheduling. This type of scheduling system maximizes utilization of the clinician, but patients may experience long wait times.
An individual appointment scheme assigns different, equally spaced appointment times to each patient. In a common modification of this type of system, different appointment lengths are available and assigned on the basis of the type of patient. This system reduces patient waiting time but decreases utilization of the clinician; in other words, increasing the interval between arrivals results in a reduction of both waiting time and utilization.
A mixed block-individual appointment scheme schedules a group of patients to arrive at the start of the clinic session, followed by equally spaced
Sequencing Rule
Average Completion Time
Average Tardiness
Number of Tardy Jobs
Maximum Tardiness
FCFS 186 96 3* 230
SPT 148* 58 3* 150*
EDD 150 54* 3* 150*
STR 164 66 4 150*
CR 208 110 4 260
*Best values.
Note: All times shown in exhibit are in minutes. CR = critical ratio; EDD = earliest due date; FCFS = first come, first served; SPT = shortest processing time; STR = slack time remaining.
EXHIBIT 12.9Comparison of
VVH Blood Test Sequencing
Rules
Chapter 12: Schedul ing and Capacity Management 335
appointment times for the remainder of the session. This type of system can be used to balance the competing goals of increased utilization and decreased waiting time.
Finally, other appointment schemes are modifications of the first three types.
Simulation has been used to study the performance of various appoint-ment scheduling models and rules. Although no scheduling rule or scheme has been found to be universally superior, the Bailey-Welch rule (Bailey and Welch 1952) performs well under most conditions. This rule schedules two patients at the beginning of a clinic session, followed by equally spaced appointment times for the remainder of the session.
Chow and colleagues (2011) demonstrate how to reduce the number of surgery cancellations by using an advanced computer simulation model to improve the allocation of open surgical slots in the appointment system. Using Monte Carlo simulation techniques, they increased surgical volume by more than 5 percent and reduced the number of overcapacity bed days by more than 9 percent.
Kaandorp and Koole (2007a, 2007b) developed a mathematic model, called the Optimal Outpatient Scheduling tool, to determine an optimal sched-ule using a weighted average of expected waiting times of patients, idle time of the clinician, and tardiness (the probability that the clinician has to work later than scheduled multiplied by the average amount of added time). This tool uses simulation to compare the optimal schedule found using the model to a user-defined schedule.
Riverview Clinic Appointment SchedulePhysicians at VVH’s Riverview Clinic typically see patients for six consecutive hours each day. Each appointment takes an average of 20 minutes; therefore, each clinician is scheduled to see 18 patients per day. The patient no-show rate is 2 percent. Currently, Riverview uses an individual appointment scheme with appointments scheduled every 20 minutes. However, clinicians have been complaining that they often have to work late but are idle at various points during the day. Riverview decides to use the Optimal Outpatient Scheduling tool (Kaandorp and Koole 2007b) to determine if another scheduling model can alleviate these complaints without increasing patient waiting time to an unacceptable level.
Exhibit 12.10 shows the results of this analysis when waiting time weight is 1.5, idle time weight is 0.2, and tardiness weight is 1.0. The optimal schedule follows the Bailey-Welch rule. Under this rule, patient waiting is increased by five minutes, but both idleness and tardiness are decreased. Riverview Clinic leaders do not believe that the additional waiting time is unacceptable and decide to implement this new appointment scheduling scheme.
Healthcare Operat ions Management336
EXHIBIT 12.10Riverview Clinic
Appointment Scheduling
Source: Kaandorp and Koole (2007b). Copyright © 2007 Guido Kaandorp and Ger Koole.
Chapter 12: Schedul ing and Capacity Management 337
Advanced-Access Patient SchedulingAdvanced Access for an Operating and Market AdvantageIn the early 1990s, Mark Murray, MD, and Catherine Tantau, RN, were among the early adopters of advanced-access scheduling at Kaiser Permanente in Northern California. Their goal was to eliminate long patient waits for appoint-ments and bottlenecks in clinic operations (Singer 2001). The principles they developed and refined have now been implemented by many leading healthcare organizations globally.
Because most clinics today use traditional scheduling systems, long wait times are prevalent and appointments may only be available weeks, or even months, into the future. The further in advance that visits are scheduled, the greater the fail (no-show) rate becomes. To compensate, providers double-book or even triple-book appointment slots. Long delays and queues occur when all the patients scheduled actually appear for the same appointment slot. This problem is compounded by patients who have urgent needs requiring that they be seen immediately. These patients are either worked into the schedule or sent to an ED, decreasing both continuity of care for the patient and revenue to the clinic. At the ED, patients are frequently told to see their primary care physician (PCP) in one to three days, further complicating the scheduling problem at the physician office.
Advanced access is implemented by beginning each day with a large portion of each provider’s schedule open for urgent, routine, and follow-up appointments. Patients are seen when they want to be seen. This scheme dra-matically reduces the fail rate, as patients do not have to remember clinic visits they booked long ago. Because no double or triple booking occurs, patients are seen on time and schedules run smoothly. Clinics using advanced access can provide patients with the convenience of walk-in or urgent care, with the added advantage of maintaining continuity of care with their own doctors and clinics.
Parente, Pinto, and Barber (2005) studied the implementation of advanced-access scheduling in a large midwestern clinic with a patient panel of 10,000. Following implementation, the average number of days between calling for an appointment and being seen by a doctor decreased from 18.7 to 11.8. However, the most significant finding was that 91.4 percent of patients saw their own PCP following implementation of the system, as opposed to 69.8 percent pre-implementation.
Implementing Advanced AccessChanging from a long-standing—albeit flawed—scheduling system to advanced access is challenging. However, an organization can increase its probability of success by following a few well-prescribed steps. In a study of large urban public hospitals, Singer (2001) developed the following methodology to implement advanced access.
Advanced-access schedulingA method of scheduling outpatient appointments that provides open time slots every day for seeing patients on the same day they request an appointment. Also known as same-day scheduling.
Healthcare Operat ions Management338
Obtain Buy-InLeadership is key to making this major change. The advanced-access system must be supported by senior leaders as well as providers. Touring other clinics that have implemented advanced access may help these groups understand how this system can work successfully.
For large systems, starting small in one or two clinical settings is best. Once initial operating problems are resolved and clinic staff are expressing posi-tive feelings about the change, advanced access can be carefully implemented in additional clinics in the system.
Predict DemandThe first quantitative step in implementation is to measure and predict demand from patients. For each day during a study period, demand is calculated as the number of patients requesting appointments (today or in the future), walk-in patients, patients referred from urgent care clinics or EDs, and calls deflected to other providers. After initial demand calculations are performed, additional factors may be included, such as day of the week, seasonality, demand for same-day versus scheduled appointments, and even clinical characteristics of patients.
Predict CapacityThe capacity of the clinic needs to be determined once demand is calculated. In general, capacity is the sum of appointment slots available each day. Capacity can vary dramatically from day to day, as providers usually have obligations for their time in addition to seeing patients in the clinic. Determining whether a clinic’s capacity can meet expected demand is relatively easy using Little’s law (described in detail in chapter 11).
That said, true capacity may not be readily apparent. Singer (2001) reports that, prior to close examination, leaders at many public hospital clinics felt that demand exceeded capacity in their operations. However, several of these clinics were able to find hidden capacity in their systems by using provid-ers effectively (e.g., by minimizing their paperwork) and converting storage space to examination areas.
Another opportunity to improve the capacity of a clinic is to standardize and minimize the length of visit times. A clinic with high variability in appoint-ment times may find that it has many small blocks of unused time.
Assess OperationsThe implementation of advanced access provides the opportunity to review and improve the core patient flow and operations in a clinic. The tools and tech-niques of process mapping and process improvement, particularly value stream mapping and the theory of constraints, should be applied before advanced access is implemented.
Chapter 12: Schedul ing and Capacity Management 339
Work Down the BacklogWorking down the backlog is one of the most challenging tasks in implement-ing advanced access, as providers are required to see more patients per day than usual until they have caught up to same-day access. For example, each provider may need to work one extra hour per day and see three additional patients until the backlog is eliminated.
The number of days needed to work down a backlog can be determined using this equation:
Days to work down backlog = Current backlog ÷ Increase in capacity,
where current backlog equals the number of appointments on the books divided by the average number of patients seen per day, and increase in capacity is the new service rate (patients per day) divided by the old service rate minus 1.
Go LiveOnce a clinic has completed the above steps, it is almost ready to go live with its advanced-access scheduling system. However, it must first determine how many appointment slots to reserve for same-day access. Singer and Regenstein (2003) report that public hospital clinics leave 40 percent to 60 percent of their slots available for same-day access while other types of clinics leave up to 75 percent of slots available.
Educating patients in anticipation of the shift to advanced access is important, as many will be surprised by the ability to see a provider the day they request an appointment. Many elderly patients may actually decline this option, as they may need more time to prepare for the appointment or arrange transportation to it.
No clinic operates in a completely stable environment, so prospectively developing contingency plans is useful. Contingencies can range from the unexpected, such as a provider being ill or called away on an emergency, to the predictable, such as increases in demand, as for routine physicals in the weeks preceding the start of school. Good contingency planning ensures the smooth and efficient operation of an advanced-access system.
Metrics for Evaluating Advanced AccessGupta and colleagues (2006) developed the following set of key indicators that can be used to evaluate the performance of advanced-access scheduling systems:
• PCP match—percentage of same-day patients who see their own PCP• PCP coverage—percentage of same-day patients seen by any physician• Wait time for next appointment (or third next available appointment)—
for example, if you are calling on Monday and an appointment is
Healthcare Operat ions Management340
available on Tuesday, Thursday, and Friday, the wait time for the third next available appointment is five days (Friday)
• Good backlog—appointments scheduled in advance because of patient preference
• Bad backlog—appointments waiting because of lack of slots
Most well-functioning advanced-access systems have high PCP match and PCP coverage. Depending on patient mix and preferences, the good backlog may be relatively large and still not be problematic, but a large or growing bad backlog can signal that capacity or operating systems in the clinic need to be improved.
Fears About Advanced Access and Their ResolutionPointing out the realities of same-day scheduling can help reduce physicians’ fears about change and help them make an effective adjustment to the new system. Gregg Broffman, MD, medical director of the 110-physician Lifetime Health Medical Group in Rochester and Buffalo, New York, whose group adopted same-day scheduling in the late 1990s, reported the following three common fears that physicians experience but that actually are unjustified (Olsen 2012):
• Insatiable demand. Physicians worry that opening their schedule will leave them swamped with work, but this is a false expectation. By carefully measuring and predicting supply and demand, advanced access ensures adequate coverage and can help determine the need to hire new clinicians to handle the workload.
• Fewer encounters. Use of same-day scheduling has been shown to decrease the number of annual encounters with individual patients. At the same time, it boosts the likelihood that patients will see their personal physician, rather than be worked in with the first available clinician. As a result, patients are more satisfied with their visits than they would be without advanced access. Furthermore, clinical outcomes rise while costs decrease, because a person’s regular practitioner is less likely to order unnecessary tests or prescribe medication than is a clinician who is unfamiliar with the patient’s history.
• Lower revenue. Decreased volume might suggest a dip in practice revenue, but the opposite has proven true. Clinicians who initially saw a 10 percent to 15 percent drop in encounters experienced about an 8 percent increase in relative value units, which are used to measure the robustness (or “dollar value”) of an office visit. For example, when a diabetic patient makes an unplanned visit, physicians can look ahead to her next scheduled appointment and “max pack” the initial visit by
Chapter 12: Schedul ing and Capacity Management 341
performing the future checkup that day. The visit can be coded at the higher level allowed by a more complicated encounter, and the max packing leaves an appointment open in two weeks to see a new patient.
Conclusion
Advanced-access scheduling is an efficient and patient-friendly method of sched-uling the delivery of ambulatory care. However, implementing and maintaining this and other capacity management techniques are difficult unless leadership and staff are committed to their success.
Discussion Questions
1. What job sequencing rule do you see most often in healthcare? Why? Can you think of any additional job sequencing rules not described in this text?
2. How could advanced-access techniques be used for the following types of facilities? a. An ambulatory surgery center b. A freestanding imaging center
3. What are the consequences of using advanced access in a multispecialty clinic? How might these tools be applied to provide same-day scheduling?
4. Can advanced-access techniques be used with appointment scheduling schemes? Why or why not?
Exercises
1. Two of the nurses (Mary and Tom) at Riverview UCC have decided to work part time rather than their previous full-time schedule. Each prefers to work only two (consecutive) days per week. Once they become part-time employees, salary and benefits per nurse-day for these nurses will be reduced to $160 on weekdays and $220 on weekend days. Considering this savings, Riverview UCC can hire an additional full-time nurse if needed. Should Riverview UCC agree to the two nurses’ request? If the clinic agrees, will additional nurses need to be hired? Assuming that part-time nurses and any new hires accept any schedule offered by Riverview UCC and that preferences for the
Healthcare Operat ions Management342
remainder of the nurses are the same as stated in the chapter, what new schedule would you recommend for each nurse?
2. The VVH radiology department currently uses FCFS to determine how to sequence patient X-rays. On a typical day, the department collects patient X-ray data, and these data are available on the book’s
companion website. Use the data to compare various sequencing rules. Assuming these data are representative, what rule should the radiology department adopt for sequencing, and why?
3. Use the Optimal Outpatient Scheduling tool (Kaandorp and Koole 2007b), provided on the companion website, to compare two appointment scheduling schemes—individual appointments and optimal scheduling—under the following assumptions. For the individual scheduling scheme, assume an 8-hour day that can be divided into 10-minute time blocks (48 time intervals), a 15-minute service time for patients, 24 patients seen according to the individual appointment scheme (a patient is scheduled to be seen every 20 minutes), and 5 percent no-shows. For the small neighborhood optimal schedule, assume a waiting time weight of 1, an idle time weight of 1, and a tardiness weight of 1. What are the differences in the two schedules? Which would you choose? Why? Now, increase the waiting time weight to 3 and recompute the small neighborhood optimal schedule. How is this optimal schedule different from the previous one? Finally, change the service time to 20 minutes and compare the individual appointment
schedule scheme to the small neighborhood optimal schedule with waiting time weights of 1 and 3. Which schedule would you choose, and why?
4. A clinic wants to work down its backlog to implement advanced access. The clinic currently has 1,200 booked appointments and sees 100 patients a day. The physician staff have agreed to extend their schedules and can now see 110 patients per day. What is their current backlog, and how many days will it take to reduce it to zero?
References
Bailey, N. T. J., and J. D. Welch. 1952. “Appointment Systems in Hospital Outpatient Departments.” Lancet 259: 1105–8.
Chow, V. S., M. L. Puterman, N. Salehirad, W. Huang, and D. Atkins. 2011. “Reducing Surgical Ward Congestion Through Improved Surgical Scheduling and Uncapacitated Simulation.” Production and Operations Management 20 (3): 418–30.
On the web at ache.org/books/OpsManagement3
On the web at ache.org/books/OpsManagement3
Chapter 12: Schedul ing and Capacity Management 343
Gupta, D., S. Potthoff, D. Blowers, and J. Corlett. 2006. “Performance Metrics for Advanced Access.” Journal of Healthcare Management 51 (4): 246–59.
Institute for Healthcare Improvement (IHI). 2012. “Advanced Access: Reducing Waits, Delays, and Frustrations in Maine.” Accessed October 4, 2016. www.ihi.org/resources/Pages/ImprovementStories/AdvancedAccessReducingWaitsDelaysand FrustrationinMaine.aspx.
Kaandorp, G. C., and G. Koole. 2007a. “Optimal Outpatient Appointment Scheduling.” Health Care Management Science 10 (3): 217–29.
. 2007b. “Optimal Outpatient Appointment Scheduling Tool.” Accessed June 24. http://obp.math.vu.nl/healthcare/software/ges.
Matthews, C. H. 2005. “Using Linear Programming to Minimize the Cost of Nurse Per-sonnel.” Journal of Healthcare Finance 32 (1): 37–49.
Parente, D. H., M. B. Pinto, and J. C. Barber. 2005. “A Pre-Post Comparison of Service Operational Efficiency and Patient Satisfaction Under Open Access Scheduling.” Health Care Management Review 30 (3): 220–28.
Singer, I. A. 2001. Advanced Access: A New Paradigm in the Delivery of Ambulatory Care Services. Washington, DC: National Association of Public Hospitals and Health Systems.
Singer, I. A., and M. Regenstein. 2003. Advanced Access: Ambulatory Care Redesign and the Nation’s Safety Net. Washington, DC: National Association of Public Hospitals and Health Systems.
Soriano, A. 1966. “Comparison of Two Scheduling Systems.” Operations Research 14 (3): 388–97.
Trabelsi, S., R. Larbi, and A. Alouane. 2012. “Linear Integer Programming for Home Health Care.” Lecture Notes in Business Information Processing 100 (2): 143–51.
CHAPTER
345
SUPPLY CHAIN MANAGEMENT
Operations Management in Action
Trinity Health, a multisite healthcare sys-tem based in Livonia, Michigan, reports that recently adopted aggressive supply chain management techniques will save the organization $20 million in costs. This cost decrease is being achieved largely through a relentless reduction of redundant inventory across the system and consolidation of many of its practices into efficiently streamlined services.
The central focus of the efforts at Trinity was to reduce the inventory in the supply chain. Previously, fulfilling the sup-ply preferences of physicians and nurses led to increased SKUs [stock keeping units] (representing individual products) and total dollars of inventory in the system. “As we’re bringing more organizations together, we naturally want to take advantage of econo-mies of scale,” says Lou Fierens, senior vice president overseeing supply chain at Trinity. “We had to rigorously reduce the amount of SKUs that we use inside the hospital” and centralize procurement of the medical goods, he says. This centralization gives the system better insights into inventory usage patterns across the entire system and the ability to adjust purchasing practices to take advantage of economies of scale and improve availability of inventory throughout the system.
13OVE RVI EW
In the current world of healthcare and healthcare reform, the
supply chain is rarely discussed as a source of improvement and
cost savings. However, health spending related to the supply
chain represents a substantial opportunity to save capital. A
groundbreaking study indicates a potential savings of 2 percent
to 8 percent of overall operating costs with an effective sup-
ply chain for tangible goods in hospitals and health systems
(McKone-Sweet, Hamilton, and Willis 2005). Johnson and Teplitz
(2009) demonstrated that procurement costs can be reduced
by more than 10 percent and quantity of items purchased by
more than 20 percent. With many hospital budgets exceeding
$500 million, this savings represents an enormous impact on an
organization’s bottom line.
As a result, efficient and effective supply chain manage-
ment (SCM) is increasingly important in healthcare. This chapter
introduces the concept of SCM and the various tools, techniques,
and theories that can enable supply chain optimization. The major
topics covered include the following:
• SCM basics
• Tools for tracking and managing inventory
• Forecasting
• Inventory models
• Inventory systems
• Procurement and vendor relationship management
• Strategic SCM
After completing this chapter, readers should have a
basic understanding of SCM. This knowledge will help them
determine how to apply SCM in their organizations and enable
them to employ SCM-related tools, techniques, and theories to
optimize supply chains.
Healthcare Operat ions Management346
The healthcare system is able to increase cash flow and reduce costs by increasing the inventory turnover in the system. The increase in inventory turnover reduces the amount of time between paying for inventory and receiving revenue from that inventory. These supply chain practices make a tremendous impact on the profitability of the healthcare systems.
Source: Adapted and excerpted from Chao (2016).
Supply Chain Management
The supply chain includes all of the processes involved in moving supplies and equipment from the manufacturer to patient care areas. Supply chain manage-ment is the handling and oversight of all activities and processes related to both upstream vendors and downstream customers in the value chain. Because SCM requires the effective management of relationships outside as well as inside an organization, this discipline constitutes a broad field of thought.
SCM aims to reduce costs and increase efficiencies associated with the supply chain. This effort carries substantial implications, as Duffy (2009) indicates that the average hospital can assume its expenditure on supplies, and on labor to manage supplies, is approximately 25 percent of its total operat-ing budget.
Effective SCM is enabled by new technologies as well as “old” meth-odologies for reducing supply-associated costs and effort and improving the efficiency of supply processes. Many techniques used to improve supply chain performance in other industries are applicable to healthcare. They may include technology-enabled solutions, such as electronic procurement, radio-frequency identification (RFID), bar coding, point-of-use data entry and retrieval, and data warehousing and management. These technologies have been used in other industries, and healthcare organizations increasingly find that they, too, can reduce costs and increase safety by using them.
A systems view of the supply chain can lead to an enhanced understanding of processes and how best to improve and optimize them. SCM is focused on managing relationships with vendors and customers to render the entire chain (rather than just pieces of it) as efficient as possible, and it results in benefits for all members of the chain.
For SCM to be effective, reliable and accurate data are required to deter-mine where the greatest improvements and gains can be made by improving the supply chain.
Supply chain managementThe management of all supplier, vendor, and distribution activities related to the production of value to end consumers.
Chapter 13: Supply Chain Management 347
Tracking and Managing Inventory
Inventory is the stock of items held by the organization either for sale or to support the delivery of a service. In healthcare organizations, inventory typi-cally includes supplies and pharmaceuticals. This stock allows organizations to cope with variations in supply and demand while making cost-effective ordering decisions.
Inventory management helps determine how much inventory to hold, when to order, and how much to order. Effective and efficient inventory management requires a classification system; an inventory tracking system; a reliable forecast of demand; knowledge of lead times; and reasonable estimates of holding, ordering, and shortage costs.
Inventory Classification SystemsNot all inventory is equal: Some items may be critical for the organization’s operations, some may be costly or relatively inexpensive, and some may be used in large volumes while others are seldom needed. A classification system can enable organizations to manage inventory effectively by allowing them to focus on the most important inventory items and place less emphasis on those items of low importance.
The ABC classification system divides inventory items into three catego-ries on the basis of the Pareto principle. Vilfredo Pareto studied the distribution of wealth in nineteenth-century Milan and found that 80 percent of the wealth was controlled by 20 percent of the people (Femia and Marshall 2012). This same idea of the vital few and the trivial many is found in quality management (chapter 9) and sales (80 percent of sales come from 20 percent of customers).
In ABC classification, the A items have a high-dollar volume (70–80 percent) but account for only 5–20 percent of items, B items have moderate-dollar (30 percent) and -item (15 percent) volume, and C items are low-dollar (5–15 percent) and high-item (50–65 percent) volume. The classification of items is not related to their unit cost; an A item may have high-dollar volume because of high usage and low cost or because of high cost and low usage. Items vital to the organization should be assigned to the A category even if their dollar volume is low to moderate.
The A items are the most important and, therefore, the most closely managed. The B and C items are less important and less closely managed. In a hospital setting, pacemakers are an example of A items and facial tissue might be a C item. The A items are likely ordered more often than B and C items, and their inventory accuracy is checked more often. These items are good candidates for bar coding and point-of-use systems. The C items do not need
Healthcare Operat ions Management348
to be as closely managed, and often, a two-bin system (discussed later in this chapter) is used for their management and control.
Inventory Tracking SystemsAn effective inventory management system requires a means of determining how much of a particular item is available. In the past, inventory records were updated manually and typically were not very accurate. Bar coding and point-of-use systems have eliminated much of the data input inaccuracy, but inven-tory records are still imperfect. A physical count must usually be performed to ensure that the actual and recorded amounts are the same.
Although many organizations perform inventory counts on a periodic basis (e.g., once a month), cycle counting is a more helpful technique in ensur-ing accuracy and eliminating errors. Highly accurate inventory records not only enable efficient inventory management but also help eliminate the hoarding that occurs when providers are concerned an item will be unavailable when needed. In a typical cycle counting system, a physical inventory is performed on a rotating schedule on the basis of item classification. The A items might be counted every day, whereas C items are counted once a month.
Electronic medication orders and matching allow an organization to track demand and improve patient safety—the patient and order are matched at the time of administration. Rules can be set up in the system to alert provid-ers to adverse drug interactions and thus eliminate errors. Systems are being developed that bring complete, current patient records to the bedside; the ready availability of patient and drug history can improve the quality and safety of the care delivered.
Radio-Frequency IdentificationRFID is a tool for identifying objects, collecting data about them, and storing those data in a computer system with little human involvement. RFID tags are similar to bar codes, but they emit a data signal that can be read without actually scanning the tag. RFID tags can also be used to determine the location of the object to which they are attached. However, using RFID tags is more expensive than bar coding.
BJC HealthCare, with hospitals in Illinois and Missouri, uses RFID to keep track of expensive equipment and supplies in the system. The RFID technology allows the organization to collect data and use those data to build increasingly effective inventory control systems. BJC has reported a 23 percent reduction in its inventory as a result of using the RFID technology (Chao 2015).
PinnacleHealth Harrisburg (Pennsylvania) Hospital has also successfully implemented RFID technology to track and locate expensive medical equip-ment (Wright 2007). The system can be queried to locate a particular piece of
Chapter 13: Supply Chain Management 349
equipment rather than staff having to search the hospital for it. The hospital’s real-time asset-tracking program saved PinnacleHealth $900,000 in its first 12 months of deployment (Radianse 2016).
Warehouse ManagementWarehouse management systems enable healthcare organizations to optimize operations, thereby decreasing their storage and facility costs. Functions such as bar coding and point-of-use systems help reduce the labor needed by automating data entry in the receiving area; automations such as these also reduce errors, allowing for more accurate determination of the inventory held. Information about demand trends at the organization, if available at the warehouse or stor-age facility, can be used to organize inventory so that the heavily demanded items are easily accessible. This process can significantly reduce labor costs associated with the storage facility.
Demand Forecasting
Knowledge of demand and demand variation in the system enables improved demand forecasting, which, in turn, can allow inventory reductions and enhance the probability that an item is available when needed. Bar coding and point-of-use systems allow organizations to track when and how many supplies are consumed, to use that information to forecast demand organization-wide, and to plan how to meet that demand effectively in the future.
Forecasting, or time series analysis, is used to predict what will happen in the future on the basis of data obtained at set intervals in the past. For example, forecasting can be used to predict the number of patients who will be seen in the emergency department in the next year (or month or day) based on the number of patients seen there in the past. Time series analysis accounts for the fact that data points collected over time may be related to one another and, therefore, violate the assumptions of linear regression. Forecasting methods range from simple to complex. Here, we describe the simpler methods; only a brief discussion of the more complicated methods is provided.
Averaging MethodsAll averaging methods assume that the variable of interest is stable or station-ary—not growing or declining over time and not subject to seasonal or cyclical variation.
Simple Moving AverageA simple moving average (SMA) takes the last p-values and averages them to forecast the value in the next period:
Healthcare Operat ions Management350
=+ + +− − −F
D D D
p,t
t t t p1 2
where Ft = forecast for period t (or the coming period), Dt–1 = value in the previous time period, and p = number of time periods.
Weighted Moving AverageIn contrast to SMA, where all values from the past are given equal weight, a weighted moving average (WMA) weights each previous time period. Typi-cally, the more recent periods are assumed to be more relevant and are assigned greater weight:
Ft = w1Dt–1 + w2Dt-2 + . . . + wpDt–p,
where Ft = forecast for period t (or the coming period), Dt–1 = value in the previous time period, wp = weight for time period p, and w1
+ w2 + . . . + wp
= 1.
Exponential SmoothingThe problem with the previous two methods, SMA and WMA, is that a large amount of historical data is required to compute the solutions. With single exponential smoothing (SES), the oldest data are eliminated once new data have been added. The forecast is calculated by using the previous forecast as well as the previous actual value with a weighting or smoothing factor, alpha (α). Alpha can never be greater than 1, and higher values of alpha put more weight on the most recent periods:
Ft = αDt-1 + (1 − α)Ft-1,
where Ft = forecast for period t (or the coming period), Dt–1 = value in the previous time period, and α = smoothing constant ≤ 1.
Trend, Seasonal, and Cyclical ModelsHolt’s Trend-Adjusted Exponential Smoothing TechniqueSES assumes that the data fluctuate around a reasonably stable mean; that is, no trend or consistent pattern of growth or decline is present. If the data contain a trend, Holt’s trend-adjusted exponential smoothing model can be used.
Trend-adjusted exponential smoothing works much like simple smooth-ing, except that two components—level and trend—must be updated each period. The level is a smoothed estimate of the value of the data at the end of each period, and the trend is a smoothed estimate of average growth at the end of each period. Again, the weighting or smoothing factors, α and delta
Single exponential smoothing (SES)A simple forecasting model that smooths data in a time series to predict the future.
Trend-adjusted exponential smoothingAn extension of a single exponential smoothing model that accounts for a trend when smoothing the data.
Chapter 13: Supply Chain Management 351
(δ), can never exceed 1, and higher values put more weight on more recent time periods:
FITt = F + Tt
and
Ft = αDt–1 + (1 − α)FITt–1Tt = Tt–1 + δ(Ft–1 – FITt–1),
where FITt = forecast for period t including the trend, Ft = smoothed forecast for period t, Tt = smoothed trend for period t, Dt–1 = value in the previous time period, 0 ≤ α = smoothing constant ≤ 1, and 0 ≤ δ = smoothing constant ≤ 1.
Linear RegressionAlternatively, when a trend exists in the data, regression analysis (chapter 7) is often used for forecasting. Demand is the dependent, or Y, variable, and the time period is the predictor, or X, variable. The regression equation
Ŷ = b(X) + a
can be restated using forecasting notation as
Ft = b(t) + a,
where Ft = forecast for period t, b = slope of the regression line, and a = Y inter-cept. To find b and a, D = actual demand, D = average of all actual demands, t = time period, and t = average of time periods, such that b = Σ (t – t )(D – D) ÷ Σ (t – t )2 and a = D – bt.
In time series forecasting, the predictor variable is time. Regression analysis is also used in forecasting when a causal relationship exists between a predictor variable (not time) and the demand variable of interest. For example, if the number of surgeries to be performed at some future date is known, that information can be used to forecast the number of surgical supplies needed.
Winter’s Triple Exponential Smoothed ModelIn addition to adjusting for a trend, Winter’s triple exponential smoothed model adjusts for a cycle or seasonality.
Autoregressive Integrated Moving Average ModelsAutoregressive integrated moving average (ARIMA) models, developed by Box and Jenkins (1976), model a wide variety of time series behavior. However,
Healthcare Operat ions Management352
ARIMA is a complex technique; although it often produces appropriate models, it requires a great deal of expertise to use.
Model Development and EvaluationForecasting models are developed on the basis of historical time series data using the previously described techniques. Typically, the “best” model is the simplest one available by which to minimize the forecast error associated with that model. Mean absolute deviation (MAD), mean squared error (MSE), or both may be used to determine error levels:
D F
nMAD
| |n
t t t1Σ
=−
=
D F
nMSE ,t
n
t t1
2Σ( )=
−=
where t = period number, F = forecast demand for the period, D = actual demand for the period, and n = total number of periods.
Many of these forecasting models are available as downloads on the companion website to this book.
Vincent Valley Hospital and Health System Diaper Demand ForecastingJessie Jones, purchasing agent for Vincent Valley Hospital and Health Sys-tem (VVH), wants to forecast demand for diapers. She gathers information related to past demand for diapers (exhibit 13.1) and plots it on a graph (exhibit 13.2). The plot of weekly demand shows no cycles or trends, so Jes-sie believes that an averaging method is most appropriate for achieving the accuracy desired. She obtains a five-period SMA forecast; a WMA forecast with weights of 0.5, 0.3, and 0.2; and an exponentially smoothed forecast with an alpha of 0.25.
SMA forecast:
FA A … A
p
FA A A A
tt t t p=
+ + +
=+ + + +
− − −1 2
1413 12 11 10 AA9
560 43 53 54 45
551= + + + + =
On the web at ache.org/books/OpsManagement3
Chapter 13: Supply Chain Management 353
Period Week of Cases of Diapers
1 1-Jan 70
2 8-Jan 42
3 15-Jan 63
4 22-Jan 52
5 29-Jan 56
6 5-Feb 53
7 12-Feb 66
8 19-Feb 61
9 26-Feb 45
10 5-Mar 54
11 12-Mar 53
12 19-Mar 43
13 26-Mar 60
EXHIBIT 13.1VVH Weekly Diaper Demand
Wee
kly
Dem
and
0
10
20
30
40
50
60
70
80
Period
1 2 3 4 5 6 7 8 9 10 11 12 13
EXHIBIT 13.2Plot of VVH Weekly Diaper Demand
Healthcare Operat ions Management354
WMA forecast:
× × ×
− − −F w A w A w A
F w A w A w A
= + + . . . +
= ( ) + ( ) + ( )
= (0.5 60) + (0.3 43) + (0.2 53) = 53.5
t t t p t p1 1 2 2
14 1 13 2 12 3 11
Exponentially smoothed forecast:
F A F
F A F
= + 1
= 0.25 + (1 0.25)
= 0.25 60 + 0.75 52 = 54
t t t1 1
14 13 13
α α )( −
× − ×
× ×
− −
Because each method results in a different forecast, Jessie compares the methods to determine which is best. She uses the Excel forecasting template (found on the companion website) to perform the calculations (exhibit 13.3).
She finds that both MAD and MSE are lowest with the WMA method and decides to use that method for forecasting. Therefore, she forecasts that 53.5 cases of diapers will be demanded the week of April 2, period 14.
Order Amount and Timing
Inventory management is concerned with the following questions:
• How much inventory should the organization hold?• When should an order be placed?• How much should be ordered?
To answer these questions, organizations need reasonable estimates of holding, ordering, and shortage costs. Knowledge of lead times and demand forecasts is also essential to determining the best answers to inventory questions.
Economic Order Quantity ModelIn the early 1900s, F. W. Harris (1913) developed the economic order quantity (EOQ) model to answer inventory questions. Although the assumptions of this model limit its usefulness in real situations, it provides important insights into effective and efficient inventory management.
To aid in understanding the model, definitions for some key inventory terms are provided.
On the web at ache.org/books/OpsManagement3
Economic order quantity (EOQ)An inventory model that indicates an optimal purchase quantity that will minimize total annual inventory costs.
Chapter 13: Supply Chain Management 355
• Lead time is the interval between placing an order and receiving it.• Holding (carrying) costs are associated with keeping goods in storage for
a period of time, usually one year. The most obvious of these costs are the cost of the space and the cost of the labor and equipment needed to operate the space. Less obvious costs include the opportunity cost of capital and those costs associated with obsolescence, damage, and theft of the goods. These costs are often difficult to measure and are commonly estimated as one third to one half the value of the stored goods per year.
• Ordering (setup) costs are the costs of ordering and receiving goods. They may also be the costs associated with changing or setting up to produce another product.
• Shortage costs are the costs of not having an item in inventory when it is needed.
• Independent demand is generated by the customer and is not a result of demand for another good or service.
• Dependent demand results from another demand. For example, the demand for hernia surgical kits (dependent) is related to the demand for hernia surgeries (independent).
• Back orders are orders that cannot be filled when received but are placed because the customer is willing to wait for the order to be filled.
• Stockouts occur when the desired good is not available.
Simple Moving Average Weighted Moving Average (3 periods) Single Exponential Smoothing
Weight3
Weight2
Weight1
Periods
LeastRecent
MostRecent5
6DAM7DAM
52.05.03.02.0
MAD 8ESMESM 86 75 MSE 135
Period Actual Forecast Error Period Actual Forecast Error Period Actual Forecast Error1 70 1 70 1 702 42 2 42 2 42 70 283 63 3 63 3 63 63 04 52 4 52 58 6 4 52 63 115 56 5 56 53 3 5 56 60 46 53 57 4 6 53 56 3 6 53 59 67 66 53 13 7 66 54 12 7 66 58 88 61 58 3 8 61 60 1 8 61 60 19 45 58 13 9 45 61 16 9 45 60 15
10 54 56 2 10 54 54 0 10 54 56 211 53 56 3 11 53 53 0 11 53 56 312 43 56 13 12 43 52 9 12 43 55 1213 60 51 9 13 60 48 12 13 60 52 814 51 14 53.5 14 54
EXHIBIT 13.3 Excel Forecasting Template Output: VVH Diaper Demand
Healthcare Operat ions Management356
The basic EOQ model is based on the following assumptions:
• Demand for the item in question is independent.• Demand is known and constant.• Lead time is known and constant.• Ordering costs are known and constant.• Back orders, stockouts, and quantity discounts are not allowed.
The EOQ inventory order cycle (exhibit 13.4) consists of stock or inven-tory being received at a point in time. An order is placed when the amount of stock on hand is just enough to cover the demand that will be experienced during lead time. The new order arrives at the exact point when the stock is completely depleted. The point at which new stock should be ordered, the reorder point (R), is the quantity of stock demanded during lead time:
R = dL
where d = average demand per time period and L = lead time.The EOQ inventory order cycle shows that the average amount of
inventory held will be
=QOrder quantity
2 2
and the number of orders placed in one year will be
DQ
Yearly demandOrder quantity
.=
Total costs are the sum of holding and ordering costs. Yearly holding costs are calculated as follows:
Cost to hold one item one year × Average inventory = h × Q2
.
Yearly ordering costs are
Cost to place one order × Yearly number of orders = o × Q2
.
Chapter 13: Supply Chain Management 357
Total yearly costs are then
× + ×hQ
oDQ2
.
Exhibit 13.5 illustrates these relationships. An inspection of this graph shows that total cost is minimized when holding costs equal ordering costs. (This relationship can also be proven using calculus.) In equation form, the order quantity that will minimize total costs is found with
× + ×hQ*
oDQ*2
,
where Q* is the EOQ.Rearranging this equation, the optimal order quantity is
Qo Dh
Qo Dh
2
*2
.
2 =× ×
=× ×
A key insight into inventory management can be gained from an exami-nation of this simple model. First, trade-offs are inherent between holding costs and ordering costs: As holding costs increase, optimal order quantity decreases, and as ordering costs increase, optimal order quantity increases.
Demandrate
Inve
ntor
yLe
vel
Orderplaced
Order qty, Q
Reorder point, R
Orderreceived
Orderplaced
Orderreceived
Leadtime
Leadtime
0
EXHIBIT 13.4EOQ Inventory Order Cycle
Healthcare Operat ions Management358
Many organizations, including those in the healthcare industry, believe that the costs of holding inventory are much higher than was previously thought. As a consequence, these organizations are decreasing order quantities and working to decrease order costs by streamlining procurement processes.
Vincent Valley Hospital and Health System Diaper Order QuantityJessie Jones, VVH’s purchasing agent, now wants to determine the optimal order quantity for diapers. From her forecasting work, she knows that annual demand, D, for the item (in this case, diapers), is
× = × =d Time period53.5 cases
week52 weeks
year2,782 cases
year.
Each case of diapers costs $5, and Jessie estimates holding costs at 33 percent. The transaction cost is $100 to place an order. Lead time for diapers is one week. She calculates the EOQ, Q*, as follows:
× ×=
× ×
= =
o Dh
2 2 $100 2,782 cases$1.67/case
333,174 cases 577 cases.2
She calculates the reorder point, R, as
= × =dL53.5 cases
week1 week 53.5 cases.
Annu
al C
ost (
$)
Minimumtotal cost
Total cost
Carrying cost = h � Q/2
Ordering cost = o � D/Q
Optimal orderquantity Q*
Order quantity (Q)
EXHIBIT 13.5Economic Order Quantity Model
Cost Curves
Chapter 13: Supply Chain Management 359
Jessie will need to place an order for 577 cases of diapers when the stock drops to 53.5 cases.
Fixed Order Quantity with Safety Stock ModelThe basic EOQ model assumes that demand is constant and known. In other words, the amount of stock carried in inventory need only match demand. In reality, demand is seldom constant, and excess inventory must be held to meet variations in demand and avoid stockouts. This excess inventory, called safety stock (SS), is the amount of inventory carried over and above expected demand. Exhibit 13.6 illustrates this model.
The SS model assumes that demand varies and is normally distributed (chapter 7). It also assumes that a fixed quantity equal to EOQ will always be ordered. The EOQ remains the same as in the basic model, but the reorder point differs because of the need for SS:
= +R dL SS.
The amount of SS to carry is determined by variation in demand and desired service level. Service level is defined as the probability of having an item on hand when needed. For example, suppose orders are placed at the beginning of a time period and received at the end of that period. If demand is expected to be 100 units in the next time period with a standard deviation of 20 units and 100 units on hand at the start of the period, the probability
Service levelThe probability of having an item on hand when needed.
Orderquantity (Q)
Inve
ntor
yLe
vel
Reorderpoint (R)
Safetystock (SS)
Leadtime
0
Leadtime
Time
EXHIBIT 13.6Variable Demand Inventory Order Cycle with Safety Stock
Healthcare Operat ions Management360
of stocking out is 50 percent and the service level is 50 percent. If demand is normally distributed, there is a 50 percent probability of its being higher than the mean and a 50 percent probability of its being lower than the mean. Demand is then greater than the stock on hand in half of the time periods.
To increase the service level, SS is needed. For example, if the stock on hand at the start of the time period is 120 units (20 units of SS), the service level increases to 84 percent and the probability of a stockout is reduced to 16 percent. Because demand is assumed to follow a normal distribution, and 120 units is exactly one standard deviation higher than the mean of 100 units, the probability of being less than one standard deviation above the mean is 84 percent. There is a 16 percent probability of being more than one stan-dard deviation above the mean (exhibit 13.7). A service level of 95 percent is typically used in industry. However, if one stockout every 20 time periods is unacceptable, a higher service level target is needed.
SS is the z-value associated with the desired service level (number of standard deviations above the mean) multiplied by the standard deviation of demand during lead time:
SS = z × σL.
Note that with this model, the only occasion in which demand variability may be problematic is during lead time. Because an order is triggered when a certain level of stock is reached, any variation in demand prior to that time does not affect the reorder point.
Probability of meeting demand duringlead time = service level = 84%
Probability of a stockout = 16%
R = reorderpoint
100
0 1
Example units
Z
Average demand during lead time = dL
120
EXHIBIT 13.7Service Level
and Safety Stock
Chapter 13: Supply Chain Management 361
This model also provides critical information about inventory manage-ment. Trade-offs exist between the amount of SS held and service level. As the desired service level increases, the amount of SS needed—and therefore the amount of inventory held—increases. As the variation in demand during lead time increases, the amount of SS increases. If demand variation or lead time can be decreased, the amount of SS needed to reach a desired service level also decreases. Many healthcare organizations continuously work with their suppliers to reduce lead time and, therefore, SS levels.
Vincent Valley Hospital and Health System Diaper Order QuantityAfter learning more about inventory models, Jessie Jones has realized that the reorder point she chose earlier by using the basic EOQ model will cause the hospital to run out of diapers during 50 percent of the order cycles. Because diapers are ordered five times per year, the hospital will stock out of diapers at least twice a year. Jessie believes this is an unacceptable amount of stockouts and determines that SS is needed to avoid them. She sets a service level of 95 percent, or one stockout every four years, as an acceptable threshold.
Jessie gathers additional information related to demand for diapers over the past year and finds that the standard deviation of demand during lead time is 11.5 cases of diapers. She calculates the amount of SS needed as follows:
z × σL = 1.64 × 11.5 = 18.9 cases.
Her new reorder point is
dL SS+ = ×
+ 18.9 cases = 72.4 cases.53 5
1. casesweek
week
Jessie needs to place an order for 577 cases of diapers when inventory drops to 72.4 cases. The forecasting template found on the companion website can be used to perform these calculations, and the output related to Jessie’s problem is shown in exhibit 13.8.
Additional Inventory ModelsMany inventory models that address some of the limiting assumptions of the EOQ have been developed. One is known as the fixed time period with SS model, whereby the order quantity varies and the time at which the order is placed is fixed. This type of model is applicable when vendors deliver on a set schedule or if one supplier is used for many different products and orders are bundled and delivered together on a set schedule. Generally, this situation requires more SS because stockouts are possible during the entire time between
On the web at ache.org/books/OpsManagement3
Healthcare Operat ions Management362
orders, not just the lead time for the order. Another inventory model is the fixed order quantity with SS model, in which the order quantity is fixed and the time at which the order is placed varies.
Models that account for quantity discounts and price breaks have also been developed. Information on these high-level models can be found in most inventory management textbooks.
Inventory Systems
In practice, various types of systems are employed for management and control of inventory. They range from simple to complex, and organizations typically employ a mixture of these systems.
Two-Bin SystemThe two-bin system is a simple, easily managed approach often used for B- or C-type items. In this system, inventory is separated into two bins. These are not necessarily actual bins or containers; the idea is to have in place a means of identifying the items as being in the first or second bin.
In the two-bin system, inventory is taken from the first bin. When that bin is emptied, an order is placed, and inventory from the second bin is used during the lead time. The amount of inventory held in each bin can be determined from the fixed order quantity with SS model. Inventory held in the first bin is ideally the EOQ minus the reorder point, and inventory in the second bin equals the reorder point.
Just-in-TimeJust-in-time (JIT) inventory systems are based on Lean concepts and employ a type of two-bin system called kanban. (See chapter 10 for a description of this type of system.) Because inventory levels are controlled by the number of
Reorder Point (ROP) with EOQ Ordering
Average daily demandAverage lead timeStd dev demand during lead time
Service levelincrement
Stockout riskz associated with service levelAverage demand during lead timeSafety stockReorder point
d =L =L =
SL =SL =
dL =SS =
ROP =
7.647
11.5
0.95
0.051.64
53.4818.972.4
UnitsDaysUnits
UnitsUnitsUnits
0.0 20.0 40.0
Prob
abili
ty
Reorder Point
60.0 80.0 100.0
Daily demand
Daily demand
ROP
EXHIBIT 13.8VVH 95 Percent
Service Level Reorder Point
Chapter 13: Supply Chain Management 363
kanbans in the system and inventory is “waste” in a Lean system, organizations try to decrease the number of kanbans as much as possible.
Material Requirements Planning and Enterprise Resources PlanningMaterial requirements planning (MRP) systems were first employed by manufacturing organizations in the 1960s when computers became commer-cially available. These systems were used to manage and control the purchase and production of dependent-demand items.
A simple example illustrates the logic of MRP (exhibit 13.9). A table manufacturer knows (or forecasts) that 50 tables, consisting of a top and four legs, will be demanded five weeks in the future. The manufacturer also knows that producing a table takes one week if both the legs and the top are avail-able. Lead time is two weeks for table legs and three weeks for tabletops. From this information, MRP helps the manufacturer determine that, to produce 50 tables in week 5, it needs to have on hand 50 tabletops and 200 table legs in week 4. The company also needs to order 200 table legs in week 2 and 50 tabletops in week 1.
The same type of logic can be applied in healthcare for dependent-demand items. For example, if the demand for a particular type of surgery is known or can be forecast, supplies related to this type of surgery can be ordered on the basis of MRP-type logic.
Enterprise resources planning (ERP) evolved from the relatively simple MRP systems as computing power grew and software applications became more sophisticated. ERP-type systems in healthcare today are found throughout the entire organization and include finance, accounting, human resources, patient records, and many more functions in addition to inventory management and control.
Material requirements planning (MRP)A computer system designed to manage the purchase and control of dependent-demand items.
Enterprise resources planning (ERP)Global information systems that help individuals and groups manage the entire organization, including accounting, operations, and human resources.
Ordertable tops
Ordertable legs
1W 5432kee
EXHIBIT 13.9Material Requirements Planning Logic
Healthcare Operat ions Management364
Procurement and Vendor Relationship Management
Analyzing and improving the processes used for procurement can result in signifi-cant savings for an organization. Technology can be used to not only streamline processes but also improve data reliability, accuracy, and visibility. Streamlining procurement processes can also reduce associated labor costs. Electronic pro-curement (e-procurement) is one example of how technology can be used to increase procurement efficiency. The ease of obtaining product information, the reduced time associated with the procurement process, and the increased use of a limited number of suppliers can significantly reduce costs as well.
In addition to basic procurement data, information about supplier reli-ability can be maintained in these systems to allow organizations to make informed choices about vendors, assuming these entities track and regularly review supplier performance. For example, one vendor may be inexpensive but extremely unreliable, whereas another may be slightly more expensive but more reliable and faster. Conducting an analysis can help an organization determine that using the slightly more expensive vendor is prudent because the amount of SS held or the need to expedite shipments may be reduced.
Value-based standardization can be used to reduce both the number of different items and the quantity of those items held. Focusing on high-use or high-cost items can leverage the benefits of standardization and reduce the number of suppliers to the organization. Holding fewer supplies and engaging fewer suppliers can result in both labor and material cost savings.
One effective means of ensuring supply availability and reducing internal labor is outsourcing. Distributors can break orders down by point of use—for example, the emergency department, the dietary department—and deliver directly to that point as needed rather than having the organization’s person-nel perform that function. In addition, the use of prepackaged supply packs or surgical carts can reduce the amount of in-house labor needed to organize these supplies and ensure that the correct supplies are available when needed. Vendor-managed inventory is another way to outsource some of the work involved with procurement, with automated supply carts or cabinets and point-of-use systems enabling this type of inventory. Participation in group purchas-ing organizations leverages increased order quantities, thereby reducing costs.
Finally, disintermediation is a way to improve supply chain management. Reducing the number of organizations in the chain can result in reduced costs and improvements in speed and reliability.
Strategic View
Most important of all the discussion related to the supply chain, effective SCM requires a strategic systems analysis and design. This strategic view enables
Chapter 13: Supply Chain Management 365
systems solutions rather than individual solutions—an important distinction, as best-practice solutions can be standardized across an entire organization rather than applied haphazardly or incorrectly. A strategic design enables system-level integration, allowing for improved decision making throughout the organization.
Successful SCM initiatives require the same elements as Six Sigma, Lean, and the Baldrige criteria:
• Top management support and collaboration, including time and money• Employee buy-in, including clinician support and frontline empowerment• Evaluation of the structure and staffing of the supply chain to ensure
that it supports the desired improvements and that all relevant functions are represented in a meaningful way (cross-functional teams may be the best way to ensure this adherence)
• Process analysis and improvement, including a thorough and complete understanding of existing systems, processes, and protocols (through process mapping) and their improvement
• Collection and analysis of relevant, accurate data and metrics to determine areas of improvement, means of improvement, and whether improvement is achieved
• Evaluation of technology-enabled solutions in terms of both costs and benefits
• Training in the use of new technologies and techniques, which is essential for broad application and use in the organization
• Internal awareness programs to highlight the need for and benefits of strategic SCM
• Improved inventory management through enhanced understanding of the system-level consequences of unofficial inventory, JIT systems, and inventory tracking systems
• Enhancement of vendor partnerships through information sharing and the investigation and determination of mutually beneficial solutions
• Performance tracking of vendors to determine the best vendors to involve in the SCM process
• Periodic education and continuous support by the organization for a systemwide view of the supply chain
• Pursuit of continuous improvement of the system rather than of individual departments or organizations in that system
Conclusion
In the past, healthcare organizations did not focus on SCM issues; today, increasing cost pressures drive them to examine and optimize their supply
Healthcare Operat ions Management366
chains. The ideas and tools presented in this chapter help the healthcare supply chain professional achieve improvements and thereby lower costs.
Discussion Questions
1. Why is SCM important to healthcare organizations?2. List some inventory items found in your organization. Which of these
might be classified as A, B, or C items? Why? How would you manage these items differently depending on their classification?
3. Think of an item for which your organization carries SS. Why is SS needed for this item? Can the amount of SS needed be reduced? How?
4. Describe the ERP system(s) found in your organization. How could it be improved?
Exercises
1. Using the materials available at the book’s companion website, investigate and summarize commercially available software solutions for healthcare organizations. Using the forecasting template found on the
companion website, forecast total US healthcare expenditures for 2017 with SMA, WMA, SES, trend-adjusted exponential smoothing, and linear trend models. a. Which model do you believe offers the best forecast?b. Do you see any problems with your model?c. Repeat this exercise for hospital care, physician services, other
professional services, dental services, home health care, and prescription drugs. According to your findings, do any one of these areas drive the increase in healthcare expense?
2. Using the Excel inventory template found on the companion website if you choose, and starting with an Excel spreadsheet including data for this problem, which also is available on the companion website, prepare the following exercise.
Hospital purchasing agent Abby Smith needs to order examination gloves. Currently, she orders 1,000 boxes of gloves whenever she thinks a need for the item exists. Abby has heard that a better way is available to do her job and wants to use EOQ to determine how much to order and when. She collects the following information.
On the web at ache.org/books/OpsManagement3
On the web at ache.org/books/OpsManagement3
Chapter 13: Supply Chain Management 367
Cost of gloves: $4.00/box
Carrying costs: 33%, or $ /box
Cost of ordering: $150/order
Lead time: 10 days
Annual demand: 10,000 boxes/year
a. What quantity should Abby order? Prove that your order quantity is “better” than Abby’s by graphing ordering costs, holding costs, and total costs for 1,000, 1,500, and 2,000 boxes.
b. How often should Abby place the order? Approximately how much time (in days) will elapse between orders?
c. Assuming that Abby is not worried about SS, when should she place her order? Draw another graph to illustrate why she needs to place her order at that particular point.
d. Abby is concerned that the reorder point she determined is wrong because demand for gloves varies. She gathers the following usage information:
Period (10 days each) Demand
1 274
2 274
3 284
4 274
5 254
6 264
7 264
8 284
9 274
10 294
11 274
12 284
13 264
14 274
Average 274
Healthcare Operat ions Management368
a. Abby decides she will be happy if the probability of a stockout is 5 percent. How much SS should Abby carry?
b. If Abby were to set up a two-bin system for gloves, how many boxes of gloves would be in each bin?
References
Box, G. E. P., and G. M. Jenkins. 1976. Time Series Analysis: Forecasting and Control, 2nd edition. San Francisco: Holden-Day.
Chao, L. 2016. “Trinity Health Will Centralize Control of Its Medical Supply Chain.” Wall Street Journal. Published March 10. www.wsj.com/articles/trinity-health-will-centralize-control-of-its-medical-supply-chain-1457643981.
. 2015. “Hospitals Take High-Tech Approach to Supply Chain.” Wall Street Journal. Published October 21. www.wsj.com/articles/hospitals-take-high- tech-approach-to-supply-chain-1445353371.
Duffy, M. 2009. “Is Supply Chain the Cure for Rising Healthcare Costs?” Supply Chain Management Review 13 (6): 28–35.
Femia, J., and A. Marshall. 2012. Vilfredo Pareto: Beyond Disciplinary Boundaries. New York: Routledge.
Harris, F. W. 1913. “How Many Parts to Make at Once.” Factory, the Magazine of Manage-ment 10 (2): 135–36, 152.
Johnson, C., and C. Teplitz. 2009. “Applying Collaborative Contracting to the Supply Chain Department of a Regional Health Care Provider.” Journal of Applied Business Research 25 (2): 41–50.
McKone-Sweet, K., P. Hamilton, and S. Willis. 2005. “The Ailing Healthcare Supply Chain: A Prescription for Change.” Journal of Supply Chain Management 41 (1): 4–17.
Radianse. 2016. “Radianse Return on Investment.” Accessed October 10. www.radianse.com/resources/radianse-roi/.
Wright, C. M. 2007. “Where’s My Defibrillator? More Effectively Tracking Hospital Assets.” APICS 17 (1): 28–33.
CHAPTER
369
IMPROVING FINANCIAL PERFORMANCE WITH OPERATIONS MANAGEMENT
Operations Management in Action
Vidant Health’s main hospital is located in Green-ville, North Carolina, but it has seven regional hospitals—some at least two hours distant. The census at the regional hospitals had been declin-ing, while at the main hospital it was increasing. This shift resulted in an issue of matching the needed staff at each hospital to the actual census. To resolve this dilemma, the system’s leadership consulted with other organizations to find a tool that could help deploy staff members where they were needed the most throughout the system. The result was Vidant FlexWork, a broad-based application system that addresses both clinical and administrative staffing for every hospital in the system.
“Employees looking for additional shifts to work enter the online Vidant FlexWork portal.” Immediately, “special intervention” positions appear. These are urgent openings for which staff members receive incentive points should they accept one. Those points translate to a standard-ized plan that offers items ranging from gift cards for national retailers to major appliances. “We have balanced the incentive points with how much it would cost us to hire additional people from an external agency, so the program works to our ben-efit,” says Lynn Lanier, vice president of finance and operations at Vidant Health.
After the special intervention positions, openings across the system are listed, which match a predetermined set of criteria provided by
14OVE RVI EW: TH E F I NANCIAL PR E SS U R E FO R CHANGE
Better Tools for Improving Financial PerformanceBecause Medicare is one of the largest sources of fund-
ing in the US healthcare system, its payment policies are
adopted by many other payers. Each year, the Medicare
Payment Advisory Commission (MedPAC) recommends
payment policy changes to the Centers for Medicare &
Medicaid Services (CMS) and the US Congress. In these
reports, MedPAC goes to great lengths to examine Medicare
beneficiaries’ access to care, the number of hospitals going
into and out of business, and whether hospitals can make
a profit on Medicare revenues.
In response to numerous hospital executives
complaining that Medicare payment is insufficient, which
results in cost shifting to private payers, the 2016 MedPAC
report analyzed Medicare costs and margins for all hospi-
tals in the United States (MedPAC 2016):
In 2014, hospitals’ aggregate Medicare margin
was –5.8 percent. However, a set of relatively
efficient hospitals [was] able to break even
on Medicare while performing well on quality
metrics. In addition, hospitals’ marginal profits
under Medicare were positive 10 percent; thus,
hospitals with excess capacity had a financial
incentive to serve more Medicare patients.
Under current law, payment rates are projected
to decline from 2014 to 2016 due to a $3 billion
decline in uncompensated care payments and
(continued)
Healthcare Operat ions Management370
each employee, including skills, expertise and desired location. “In addition to the cost savings associated with the FlexWork portal, this tool has helped us think more like a system as opposed to a collection of hos-pitals,” says Lanier. “When we cross-pollinate our employees this way, our quality message and initiatives are strength-ened.” On average, 20 percent of the shifts posted are awarded to non-home-unit employees, meaning that someone whose primary job is not in that unit has worked in a sister unit in the same or another hospital, or a different unit altogether. And Vidant Health is seeing those numbers consistently rise over time.
Since its implementation in 2008, FlexWork has saved, on average, between $5 million and $9 million annually. When FlexWork first went live, the system had five regional hospitals outside of its flagship facility. Now in ten hospitals, FlexWork has made the integration of those facilities even more effective. “When several of the newer hospitals came into our system, their productivity measures weren’t where we would have liked them, and that caused them to operate at a higher cost,” says Lanier.
Source: Excerpted and adapted from May (2013).
Making Ends Meet on Medicare and the Pressure of Narrow Networks
A number of forces have historically worked together to increase the total costs of care beyond inflation:
• The increasing incidence of chronic disease• An aging population• New diagnostic and treatment technologies
OVE RVI EW (Continued)
other policy changes (by law, uncompen-
sated care payments decline when the
share of the population that is insured
increases). We project hospitals’ aggre-
gate Medicare margin for 2016 will be
about –9 percent.
MedPAC’s implicit conclusion is that
because some hospitals do well with Medicare
payment levels, all others should be able to thrive
at these payment levels as well.
This policy direction is woven throughout
the Affordable Care Act (ACA), and the goal of policy-
makers in the United States is to stabilize or reduce
the growth of healthcare costs until it equals the
general rate of inflation.
Chapter 14: Improving F inancial Performance with Operat ions Management 371
• The increasing complexity of billing and payment systems• A provider payment system (fee-for-service) that encourages the use of
healthcare services
Today’s healthcare executive is therefore caught between two intense environmental pressures: the need to reduce costs in the face of continuing inflationary pressures and the expectation of little new revenue. This chapter provides a road map to stable or improved financial performance through the use of operations management tools presented in the preceding chapters of this book.
Specifically, this chapter
• defines improved financial performance,• describes a systems view of reducing costs and increasing revenues that
takes into consideration the new value purchasing methodologies used for payment for services,
• details how the operations tools described in this book can be used to optimize costs and revenue for each of these payment methodologies, and
• provides a case example of one hospital that has improved its operations enough to generate a positive margin on Medicare revenues.
Definition of Financial ImprovementAlthough this textbook is not primarily about financial management, a num-ber of measures are generally accepted as indicators of the successful financial performance of a healthcare enterprise. (For a more comprehensive view, refer to Gapenski and Reiter [2016].)
From a balance sheet perspective, three indicators are frequently used to assess an organization’s performance (Cleverley and Cleverley 2010):
• Cash on hand• Percentage of debt financed• Age of plant
Three key indicators of financial health on the income statement are the following:
• Revenue (growth or decline)• Margin• Costs (per unit of service)
Healthcare Operat ions Management372
Because revenue growth per service is likely to increase slowly (in Medi-care’s case, it may actually decline), healthcare executives must focus on col-lecting all available revenue while reducing costs. The approaches described in this chapter can achieve these goals. Furthermore, in addition to achieving these financial goals, the use of operations management tools almost always results in stable or improved clinical quality and patient satisfaction.
A Systems Approach to Financial ManagementMeeting financial goals is part of most healthcare managers’ job descriptions, yet many organizations lack a comprehensive approach to supporting the manager in achieving these goals. Without this type of framework, managers are often required to take measures that may provide immediate results but foster long-term problems. Some examples include
• adopting across-the-board expense reductions,• eliminating overtime without changing any processes,• using less expensive supplies without changes in the supply chain,• tolerating queuing and long waits for service, and• outsourcing key activities without having quality monitoring systems in
place.
A more effective and longer-lasting methodology than the above mea-sures is a systems approach to financial management (see exhibit 14.1). First, expenses are divided into those directly related to revenue generation and those considered overhead. Because multiple payment methodologies are in place today and for the foreseeable future, revenue is further divided into these various models. Each category can be addressed with the techniques described in this chapter.
Reduction in overhead expenses is more straightforward than in revenue-related expenses, and therefore more general techniques can be used. Revenue can be improved and optimized by growing service lines and optimizing the revenue cycle.
Expenses Directly Related to RevenueAll expenses directly related to revenue should be classified into six payment methodologies:
• Fee-for-service• Bundled• Shared savings• Full capitation
Chapter 14: Improving F inancial Performance with Operat ions Management 373
EXH
IBIT
14.
1A
Syst
ems
Appr
oach
to F
inan
cial
Man
agem
ent
Star
t
From
star
t
Div
ide
expe
nses
Cate
gori
zepa
ymen
t
Ove
rhea
d
Rev
enue
pr
oduc
ing
Fee-
for-
serv
ice
Bun
dled
(e.g
., D
RG
)
Shar
ed s
avin
gs(e
.g.,
ACO
)
Full
capi
tati
on(e
.g.,
heal
th p
lan)
and
glob
al b
udge
ts
Use
dat
a an
alys
isto
iden
tify
mar
ket
segm
ent g
row
thop
port
unit
ies
Char
ter p
roje
cts
to g
row
iden
tifie
dse
rvic
e
Use
LSS
tom
eet q
ualit
ygo
als
Use
EB
M to
ensu
re a
ppro
pria
teut
iliza
tion
Use
LSS
and
SCM
to re
duce
cos
t per
serv
ice
Use
LSS
and
SCM
to re
duce
cos
t per
depa
rtm
ent
RED
UCE
DCO
STS/
paym
ent
met
hod
and
IMPR
OV
ED
QU
ALI
TY
BO
NU
SES
RED
UCE
DCO
STS
for
over
head
INCR
EASE
DV
OLU
ME
AN
DR
EVEN
UE
For f
inal
fina
ncia
lpe
rfor
man
ce m
onit
orin
g
Mon
itor
wit
hba
lanc
edsc
orec
ard
Qua
lity
bonu
ses—
pena
ltie
s
Use
LSS
toop
tim
ize
cust
omer
expe
rien
ce
Use
LSS
toop
tim
ize
reve
nue
cycl
e
To re
venu
epr
oces
s
From
reve
nue
proc
ess To
expe
nse
proc
ess
=D
o da
ta a
naly
sis—
char
ter p
roje
ct
EXH
IBIT
14.
1A
Syst
ems
Appr
oach
to F
inan
cial
Man
agem
ent
Not
e: A
CO =
acc
ount
able
car
e or
gani
zati
on; D
RG
= d
iagn
osis
-rel
ated
gro
up; E
BM
= e
vide
nce-
base
d m
edic
ine;
LS
S =
Lea
n Si
x Si
gma;
SCM
= s
uppl
y ch
ain
mna
gem
ent
Healthcare Operat ions Management374
• Quality bonuses or penalties• Global payments
These payment areas are discussed in detail later in the section.Next, projects are chartered with a focus on each specific payment meth-
odology and operating unit(s). The first step in the project is to collect data on the current state of service delivery and determine where variance occurs in resources used and outcomes achieved. The tools of process improvement, supply chain management, and schedule optimization are then applied to reduce variance and improve outcomes. This approach reduces costs and, in many instances, increases throughput.
Fee-for-ServiceThe most atomic-level area of cost control is individual fee-for-service. Although the delivery of each service contains a variety of components (personnel, sup-plies, overhead), the “fee” is created to represent an identifiable service that is understandable by the providers and payers. Examples include services such as an office visit and a laboratory test.
Activity-based costing (ABC) is a tool that can be used to deconstruct the billing service unit and identify opportunities for cost reductions. Gapenski and Reiter (2016, chapter 7) provide a useful example of using ABC to analyze the clinic visit.
ABC follows five steps:
1. Identify the relevant activities.2. Determine the total cost of each activity, including direct and indirect
costs.3. Determine the cost drivers for the activity.4. Collect activity data for each service.5. Calculate the total cost of the service by aggregating activity costs.
For example, Gapenski and Reiter (2016) assume that the total annual cost of patient check-in, consisting of clerical labor (direct costs) plus space and other overhead costs (indirect costs), are $50,000 to support 10,000 visits per year. This calculation yields an allocation rate of $5 per visit. Similar calculations are made for each component of the office visit, and the allocation rate is then determined for each activity (exhibit 14.2). Once the allocation rates are determined, the total activity costs for each service can be calculated (exhibit 14.3).
Each cost element can be optimized with the tools described in this book. Exhibit 14.4 provides examples.
Activity-based costing (ABC)A cost allocation model that assigns a cost to each activity in an organizational unit and then totals the cost for the unit on the basis of the actual consumption of each activity.
Chapter 14: Improving F inancial Performance with Operat ions Management 375
Act
ivit
y D
ata
Act
ivit
yA
nnua
l Cos
tsCo
st D
rive
rS
ervi
ce X
Ser
vice
YTo
tal
Allo
cati
on
Rat
e
Chec
k-in
$ 50
,000
Visi
ts5,
000
5,00
010
,000
$ 5.
00
Asse
ssm
ent
75,0
00M
inut
es p
er s
ervi
ce5
1075
,000
1.00
Dia
gnos
is25
0,00
0M
inut
es p
er s
ervi
ce10
1512
5,00
02.
00
Trea
tmen
t45
0,00
0M
inut
es p
er s
ervi
ce10
2015
0,00
03.
00
Pres
crip
tion
2,50
0N
umbe
r of d
rugs
0.5
212
,500
0.20
Chec
k-ou
t50
,000
Visi
ts5,
000
5,00
010
,000
5.00
Bill
ing
15
0,00
0N
umbe
r of b
ills
per v
isit
12
15,0
0010
.00
Tota
l cos
ts$1
,027
,500
Sour
ce:
Repr
inte
d fr
om G
apen
ski a
nd R
eite
r (20
16).
EXH
IBIT
14.
2AB
C Ill
ustr
atio
n: In
itia
l Dat
a an
d Al
loca
tion
Rat
e Ca
lcul
atio
n
Healthcare Operat ions Management376
Ser
vice
XS
ervi
ce Y
Act
ivit
yCo
st D
rive
rR
ate
Cons
umpt
ion
Cost
Cons
umpt
ion
Cost
Chec
k-in
Visi
t$
5.00
1$
5.00
1$
5.00
Asse
ssm
ent
Num
ber o
f min
utes
1.00
55.
0010
10.0
0
Dia
gnos
isN
umbe
r of m
inut
es2.
0010
20.0
015
30.0
0
Trea
tmen
tN
umbe
r of m
inut
es3.
0010
30.0
020
60.0
0
Pres
crip
tion
Num
ber o
f dru
gs0.
200.
50.
102
0.40
Chec
k-ou
tVi
sits
5.00
15.
001
5.0
Bill
ing
Num
ber o
f bill
s10
.00
1 1
0.00
2
20.0
0
Co
st p
er s
ervi
ce$7
5.10
$130
.40
Sour
ce:
Repr
inte
d fr
om G
apen
ski a
nd R
eite
r (20
16).
EXH
IBIT
14.
3AB
C Ill
ustr
atio
n: F
inal
Agg
rega
tion
of A
ctiv
ity
Cost
s pe
r Vis
it
Chapter 14: Improving F inancial Performance with Operat ions Management 377
After each activity in a service is analyzed and improved, the total ser-vice cost can also be optimized by using Six Sigma and Lean techniques, as described in chapters 9 and 10, respectively. Chapter 11 outlines a number of specific techniques to optimize throughput in a clinic (hence reducing people cost per visit), and chapter 13 provides a number of supply chain management techniques to reduce supply costs. As costs are reduced at the fee-for-service level, costs at all other levels decrease as well.
Bundled PaymentsVarious fees are frequently bundled together and paid as one amount. The intent of bundling is to give the provider an incentive to minimize costs inside the bundle. Examples of bundled payments in hospitals include the following:
• Per diem. All payments for a day in a hospital are paid at one rate.• Medicare prospective payment. All payments for a stay in the hospital are
paid at one rate that is adjusted for the complexity of the admission by the diagnosis-related group (DRG) system.
• Medicare bundled payment. All payments for an episode of care are paid at one rate adjusted for complexity.
To optimize the cost structure of bundled payments, the underlying fee-for-service costs must be targeted and improved. Because hospitals have created and maintain thousands of individual fees in a document known as the chargemaster, an analysis project should be undertaken to identify which fees to target. Criteria for targeting may include the following:
Activity Improvement Tools Opportunity
Check-in Process improvement (Lean and Six Sigma, simulation, etc.) automation
Strong
Assessment Process improvement Low
Diagnosis Evidence-based medicine Medium
Treatment Evidence-based medicine Medium
Prescription Supply chain management Strong
Check-out Process improvement, automation Strong
Billing Data mining and analysis, process improvement
Strong
EXHIBIT 14.4Use of Operations Improvement Tools to Reduce Costs
Healthcare Operat ions Management378
• High volume• High cost compared to benchmarks from other organizations• High use in bundled payments where costs are highly variable
After reducing the costs for individual services, the tools of evidence-based medicine (EBM) can now be applied. They are particularly useful for optimizing costs in bundled payment models, as these protocols reflect the shared wisdom of many clinical studies on the most efficient and effective approach to a particular condition. Chapter 3 outlines contemporary approaches to the use of EBM and the power of clinical decision support systems to sup-port its implementation. Tracking physicians’ variance in their application of EBM provides another useful opportunity for cost reduction.
Shared SavingsThe next higher level of payment is the shared savings model, which is most prominently featured in the ACA as the accountable care organization (ACO). In the initial shared savings model, reimbursement was still made via fee-for-service or bundled payments. However, patients were attributed to the ACO on the basis of their use of primary care providers (e.g., 50 percent of their primary care was provided by an ACO’s primary care team). Costs for all patients were then summed for a period, and if these total costs were less than a target set by the payer, the savings were shared with both providers and payer.
An advantage of the ACO model is that it permits a variety of providers to form new systems of care to deliver services to Medicare beneficiaries. CMS continues to refine this model to increase provider participation and moderate healthcare costs for Medicare beneficiaries. The most current information on CMS ACO models can be found on the “Accountable Care Organizations” page of the CMS (2015) website.
Success in the shared savings model requires new data systems to track patients from a longitudinal perspective beyond each episode of service to ensure that when higher-than-expected costs occur, case managers can intervene. The goal of management in this model is to stay within expected expenses per patient per month while achieving quality benchmarks. This type of challenge is well suited for tools of Six Sigma such as the following:
• Run and control charts• Pareto diagrams• Cause-and-effect diagrams• Scatter plots• Regression analysis• Benchmarking
Shared savings modelA model of healthcare delivery that includes an organized system of delivery, accountability for the quality and costs of services, and a sharing of savings with the payer for these services.
Chapter 14: Improving F inancial Performance with Operat ions Management 379
Six Sigma is detailed in depth in chapter 9. In addition, the full suite of analytical tools discussed in chapter 8 can be used for this task. The tools of EBM—including chronic disease management, the medical home, compara-tive effectiveness research, and electronic health records with clinical decision support—are also important for implementing the shared savings model.
Full CapitationThe highest level of payment is full capitation. This type of arrangement with a payer should only be accepted if the organization has had experience and success with the shared savings model.
If an organization has successfully implemented a one-sided ACO-type organization and has a stable provider base and market, it may transition to a two-sided ACO, a fully state-certified health plan, or a partnership with an existing health plan to receive full capitation. In this model, the savings or loss per member per month is fully borne by the provider organization.
The key to success in this model is to reduce the use of expensive resources, which can be achieved through disciplined attention to improving systems of care. One of the most successful examples is Group Health Coopera-tive in Seattle. Its CEO outlines the following initiatives and their outcomes (Vaida 2011):
• Implement healthcare home—10 percent drop in inpatient admissions, 20 percent decline in emergency room use
• Implement shared decision making for surgery on basis of EBM findings—12 percent drop in elective surgeries
• Develop new systems to prevent readmissions of Medicare patients through EBM and process improvement—7 percent decline in readmission rate
Quality Bonuses or PenaltiesChapter 3 reviews a number of current and anticipated value purchasing mea-sures. The policy emphasis has shifted from paying for volume to paying for value. Because these new payment systems are complex and frequently changing, establishing process improvement teams (chapter 5) and using balanced score-card techniques (chapter 4) are important for healthcare leaders in monitoring results. These project teams can use all the tools of process improvement (Lean, Six Sigma, process simulation) to change procedures for improved results. Examples of teams include the following (Healthcare.gov 2012):
• Readmission reductions• Length-of-stay management• Hospital-acquired infection and condition reductions
Full capitationA methodology in which providers are paid a monthly fee for each patient who receives care in their system.
Healthcare Operat ions Management380
• Joint Commission core measures• Publicly reported quality measures
Monitoring the results of comparative effectiveness research is impor-tant to ensure that the provider is using the most current EBM. The Agency for Healthcare Research and Quality has provided the Effective Health Care Program website (http://effectivehealthcare.ahrq.gov) as an easily accessible guide to the newest discoveries.
Global PaymentsThe ACA contains a mandate for a demonstration to evaluate the use of global budgets for hospital payments. In this model, the hospital negotiates one annual payment budget for its services and must keep its costs under this budget—regardless of patient volume or acuity. The global payment model is common in many countries other than the United States. The model has the advantage of predictability for both the payers and providers, and it substantially reduces overhead costs for billing systems. However, increases in patient demand or new technology cannot be easily or quickly accommodated, and in some cases a delay results in queuing for elective services such as hip replacement.
All of the cost management tools contained in this book are useful to suc-ceed in this environment. However, the following can carry the largest impact:
• Balanced scorecard strategy maps and reporting• Analytics, benchmarking, and statistical tools to identify opportunities
for cost reductions• Process improvement with Lean and Six Sigma, with a special emphasis
on services that develop queues• Scheduling and capacity management• Supply chain management
Overhead ExpensesAll costs not directly related to revenue are overhead. A number of both general and specific tools can be used to reduce overhead expenses.
Process ImprovementAll of the process improvement tools discussed in this book (chapters 9 through 11) can also be applied to administrative processes in overhead departments. Examples include hiring new employees, conducting marketing campaigns, and processing patient complaints.
Chapter 14: Improving F inancial Performance with Operat ions Management 381
Consolidated ActivitiesMany “miscellaneous” expenses are spread through all departments with no individual in charge of managing their costs. These items can include travel, consulting, and dues fees. By centralizing management costs, savings can be achieved through bidding and the selection of a prime vendor. The various tools of project management, including earned value analysis, can be useful in holding vendors accountable for results and costs—especially for consulting contracts.
Staffing LayersAs organizations grow, close attention should be paid to the layers of manage-ment. Symptoms of overlayering include many departmental assistant managers and a proliferation of administrative assistants. These layers can be avoided through the crisp use of strategy maps and scorecards, which are closely linked to the organization’s data warehouse.
Meetings, Reports, and Automation Tools“Why do I need to go to these meetings? I have real work to do.” This is a familiar complaint from many healthcare workers—especially clinicians. Meet-ings should be minimized and the discipline of good meeting management maintained at all times (see chapter 5). One step in good meeting management is the evaluation of the meeting itself (usually at the end), and one question that should always be asked is, Do we need this meeting in the future?
Historically, many organizations have relied on paper reports that are sent to “management.” These reports should be either automated and sent via e-mail or moved to electronic scorecards. The five whys of Lean are useful in evaluating reports:
1. Why am I getting this report?2. Why do you think I need these numbers?3. Why can’t I use an exception report?4. Why can’t these exceptions be part of a scorecard with an andon
indicator (red, yellow, blue)?5. Why can’t the scorecard include a follow-up task with assigned
accountability?
As desktop computing, networks, and database design have matured, many automation tools have been developed to improve office and clerical productivity. Web conferences now are a reasonable substitute for face-to-face meetings and can save significant travel time and expense. Calendaring tools
Healthcare Operat ions Management382
allow individuals to efficiently schedule meetings without the aid of assistants. Blogs, social media, and texting are other tools that can be used, albeit with care regarding security and other issues, to improve the productivity and con-nectivity of managers.
Facility and Capital CostsThe acquisition and deployment of capital is beyond the scope of this text-book. However, evaluating the use of facilities can offer a significant oppor-tunity for cost reduction. Clinical space use optimization is best exercised with the patient flow improvement tools in chapter 11. In addition, storage space can be minimized by the effective application of the Lean tool known as the five Ss.
Administrative space should be evaluated to discern whether employees need to be on-site. Many organizations have developed effective work-at-home policies for employees with high-speed Internet access. A half-step toward completely working at home is hoteling. In this model, the employee works most of her time at home but comes to the office one or two days per week. When she is at the office, she is given a workspace that is assigned the same way hotel rooms are managed. Hoteling can save up to 80 percent of the space otherwise required for these employees.
Prioritized Departmental ActivitiesThe most aggressive cost-reduction technique in a department is to eliminate an existing function. A useful approach is to create a cost/importance chart, as shown in exhibit 14.5. The location of each function dictates whether it may be eliminated.
The vertical axis is the importance of a function to accomplishing a department’s mission.
The horizontal axis is the cost of the function. ABC is a useful costing method for these determinations, as most overhead budgets lump costs into basic expense types (e.g., personnel, supplies, services, miscellaneous). Once the chart is complete, managers may target high-cost, low-importance func-tions for reduction or elimination (function D in exhibit 14.5).
RevenueThe primary focus of this chapter is on cost reduction, but opportunities also exist for improving revenue through the use of operations management tools. The general challenge of increased revenue is also addressed in another book from Health Administration Press, Introduction to the Financial Management of Healthcare Organizations (Nowicki 2014); here, we highlight two promi-nent methods.
Chapter 14: Improving F inancial Performance with Operat ions Management 383
Optimized Revenue CycleBecause of the complexity of the US healthcare reimbursement system, many of the analytical tools in this book can be applied to optimizing the cycle of billing and collection of charges. Six Sigma in particular is a useful approach, as its goal is to reduce the variability of outcomes in processes. This method can be translated in a revenue cycle to minimizing the variance in the eventual payment amount and receipt time for the same service.
Current challenges of the revenue cycle include the following:
• Consumer-directed healthcare and the need to collect high deductibles from patients
• Pay-for-performance systems (see chapter 3)
To succeed in this environment, Buysman (2010) suggests the follow-ing five strategies:
• Provide real-time information to claims processors through data warehouses.
• Use exception-based workflow, in which processes are mapped and automated and only exceptions are handled by staff.
• Provide real-time feedback to clinicians about the impact on reimbursement of care choices (e.g., medication choice).
• Automate most tasks in the revenue cycle.• Enhance online functionality for customers (i.e., let the consumer do
the work—see chapter 11).
Linking All Cost and Revenue Models TogetherBecause many opportunities exist for financial improvement, prioritizing improvement efforts is useful. To assist in this task, build a financial model
Revenue cycleGenerating charges, issuing bills, and managing payments and receivables for a defined period.
Cost
Function B
Function A
Function C
Function D
Impo
rtan
ce
EXHIBIT 14.5Cost/Importance Chart
Healthcare Operat ions Management384
to understand the impact of various improvement projects. Exhibit 14.6 is a simplified financial model for a medium-sized hospital.
As discussed, the revenue is split into its various components and associ-ated costs determined on the basis of a ratio of costs to charges. The baseline improvement column in exhibit 14.6 shows possible percentage improvements in each segment. Cells D5 to D10 (highlighted) are the variables that affect the bottom line—cell 37 (highlighted at bottom of exhibit). The user can manually test various improvement strategies to assess their impact on the bottom line.
Solver is a powerful tool for determining the optimal mix of strategies. However, its parameters must be set to make sure its recommended improve-ment percentages are achievable.
Exhibit 14.7 shows the results of a Solver run with pure fee-for-service savings allowed up to 30 percent. The Solver results suggest that this 10 percent increase in improvement from baseline is more important than either the DRG or bundled savings. This outcome is logical because these payment bundles are built out of the fee-for-service costs.
Case Example: Benefis Health System1 In 2008, Medicare and other public reimbursements for Benefis Health Sys-tem, based in Great Falls, Montana, covered just 75 percent of its costs for this patient segment. In that context, the system was losing money on a significant portion of its business and subsidizing those losses with the remaining 25 percent of patients, who held commercial insurance or were self-pay patients. Benefis understood that this business model was not sustainable going forward and set a goal to break even on Medicare payments.
The first cost-reduction area targeted in the Medicare breakeven efforts related to productivity improvements. Efforts included benchmarking pro-ductivity, reducing staff overtime and readjusting staffing mix, managing through attrition, and increasing automation capabilities. These efforts accounted for 41 percent of the cost savings. The next largest savings area, at 33 percent overall, resulted from renegotiating contracts with vendors and consultants, canceling unnecessary contracts and changing vendors when cancelation was not feasible. Improved efficiency in supply chain accounted for 15 percent of the savings; steps included eliminating unused inventory and unnecessary maintenance processes and adopting energy-efficient pro-cesses. The remaining cost savings comprised a number of different areas and initiatives. When it came to cost savings, the system’s motto was, “Leave no rock unturned.”
The march toward Medicare breakeven began in 2009. Staff examined every cost of the organization, and they were persistent in identifying opportu-nities for cost reduction. A key component of this effort was a strong no-layoff
Chapter 14: Improving F inancial Performance with Operat ions Management 385
IMPROVEMENTS
BaselineAmount
($)Cost/
charge Solver Baseline Notes
Revenue
Pure fee-for-service
50 0.4 20% 20% Improvement reduces cost per service
DRG 100 0.8 10% 10% Improvement reduces number of services
Bundled 10 0.9 5% 5% Improvement reduces number of services
Shared savings
20 0.9 5% 5% Improvement reduces number of services
Capitation 40 0.95 3% 3% Improvement reduces number of services
Overhead 50 1% 1% Improvement reduces direct costs
Model
Revenue
Pure fee-for-service
50.0 50.0
DRG 100.0 90.0
Bundled 10.0 9.5
Shared savings
20.0 20.0
Capitation 40.0 40.0
Total 220.0 209.5 207.3
Costs
Pure fee-for-service
20.0 16.0
DRG 80.0 57.6
Bundled 9.0 6.8
Shared savings
18.0 13.0
Capitation 38.0 27.4
Overhead 50.0 49,5
Total 215.0 170.3 170.3
Net 5.0 39.2 39.2
EXHIBIT 14.6Hospital Financial Model
Healthcare Operat ions Management386
policy. Some individuals changed jobs, but no one lost a job. By 2012, not only were Medicare reimbursements and expenses aligned, but the systems produced a slight operating margin on the Medicare business.
Conclusion
We have placed this chapter near the end of this book, as we feel that cost containment is the predominant challenge for healthcare executives in the United States and throughout the world. As medical technology improves, cost pressures will only increase. And yet we know that the widespread use of the principles discussed in this chapter, supported by the application of the tools and techniques discussed throughout this book, can stabilize or reduce healthcare inflation. Furthermore, it can be done, as we have demonstrated throughout the book with our “Operations Management in Action” examples. Cost management with improved financial performance is an achievable goal for organizations willing to engage with discipline and energy.
Discussion Questions
1. Why do other payers use Medicare as the benchmark for payment? What are other options?
EXHIBIT 14.7Hospital
Financial Model Using Solver to Prioritize
Financial Improvement
Projects
Chapter 14: Improving F inancial Performance with Operat ions Management 387
2. How important is it to involve physicians in financial improvement efforts? What is the best strategy for physician engagement?
3. Compare and contrast the following three organizational approaches to financial management using operations management tools:a. A centralized department that has experts (Six Sigma black belts) on
staff who charter and lead projects throughout an organizationb. A centralized department that only conducts training in process
improvement and maintains the project management office; all projects are led by line staff who have been trained in process improvement tools
c. The use of consultants to lead process improvement projects
Exercises
1. Use the financial model discussed on page 385 (available on the book’s companion website) to find alternative priorities for financial improvement.
2. Develop a project charter for a bundled payment financial improvement project (see chapter 5).
3. Develop a strategy map to implement a healthcare home (see chapters 3 and 4).
Note
1. This section is excerpted and adapted from Goodnow (2015). Used with permission.
References
Buysman, L. 2010. “Five Things to Look For in a Next-Generation Revenue Cycle Manage-ment System.” Healthcare Financial Management 64 (8): 40–43.
Centers for Medicare & Medicaid Services (CMS). 2015. “Accountable Care Organizations (ACO).” Modified January 6. www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/ACO/index.html?redirect=/aco/.
Cleverley, W. O., and J. O. Cleverley. 2010. “Critical Financial Questions for Healthcare Executives.” Presentation at American College of Healthcare Executives Congress on Healthcare Leadership, Chicago, March.
On the web at ache.org/books/OpsManagement3
Healthcare Operat ions Management388
Gapenski, L. C., and K. L. Reiter. 2016. Healthcare Finance: An Introduction to Accounting and Financial Management, 6th edition. Chicago: Health Administration Press.
Goodnow, J. H. 2015. “Achieving Medicare Breakeven.” Healthcare Executive 30 (2): 76–79.Healthcare.gov. 2012. “Comparing Care Providers. Tools to Help You Assess the Quality
of Care You’re Getting.” Healthcare.gov. Accessed May 15. www.healthcare.gov/compare/index.html.
May, E. L. 2013. “Balancing the Cost Structure: Innovative Approaches for Long-Term Savings and High-Quality Care.” Healthcare Executive 28 (3): 10–18.
Medicare Payment Advisory Commission (MedPAC). 2016. Report to the Congress: Medi-care Payment Policy. Published March 15. www.medpac.gov/docs/default-source/reports/march-2016-report-to-the-congress-medicare-payment-policy.pdf?sfvrsn=0.
Nowicki, M. 2014. Introduction to the Financial Management of Healthcare Organizations, 6th edition. Chicago: Health Administration Press.
Vaida, B. 2011. “How Group Health Is Holding Costs Down: A KHN Interview with CEO Scott Armstrong.” Kaiser Health News. Published February 14. http://khn.org/news/group-health-scott-armstrong-coops/.
PART
VPUTTING IT ALL TOGETHER FOR
OPERATIONAL EXCELLENCE
CHAPTER
391
HOLDING THE GAINS
A strategy for holding the gains must be developed at the beginning of any operations improvement effort. HR planning, managerial accounting, and control systems are the keys to maintain-ing the gains. Although this book is not focused primarily on HR or finance, these functions are essential to sustaining the improvements achieved. Staff from these support departments should be engaged at the beginning of operations improve-ment activities and invited to be part of project teams if possible.
Approaches to Holding Gains
In this section, we introduce the three approaches; later in the chapter, they are discussed in the context of the general algorithm for using operations manage-ment tools. More extensive information related to these functional areas can be found in Human Resources in Healthcare: Managing for Success (Fried and Fot-tler 2015) and Healthcare Finance: An Introduction to Accounting and Finan-cial Management (Gapenski and Reiter 2016).
Human Resources PlanningMany of the project management and process improvement tools described in this book can bring major change in the work lives of a healthcare organization’s
15OVE RVI EW
This chapter concludes the book and integrates its concepts. The
chapter includes
• three strategies to maintain the gains in operational
improvement projects: human resources (HR) planning,
managerial accounting, and control systems;
• an algorithm that assists practitioners in choosing and
applying the tools, techniques, and methods described in
this book;
• an examination of how Vincent Valley Hospital and Health
System (VVH) uses the tools for operational excellence; and
• a look at an optimized healthcare delivery system of the
future.
The preceding chapters present an integrated approach
to achieving operational excellence. First, strategy execution and
change management systems must be well developed. Next, the
balanced scorecard and formal project management techniques
are effective methods to employ in these key organizational
challenges.
The third step is the application of quantitative tools to
meet those challenges, such as state-of-the-art data collection
and analytics tools and problem-solving and decision-making
techniques. Processes and scheduling systems can be improved
with Six Sigma and Lean. Supply chain techniques help maximize
value and minimize costs in operations.
The final step in achieving healthcare operations excel-
lence is to hold the gains. Staff and leadership energy is usually
high when an initiative is introduced, at the start of a large project,
or at the beginning of an effort to solve a problem. However, as
time passes, new priorities emerge, team members change, and
operations can drift back to unsatisfactory levels.
Healthcare Operat ions Management392
employees. Their use will occur in the process of an employee’s work, with the hope that they make that work productive and fulfilling. Some of the more powerful tools, such as Lean and Six Sigma, can provide major pro-ductivity gains—in some cases, 30 to 60 percent increases can be achieved. A clinical process improvement project may significantly alter the tasks that fill an employee’s workday. In this environment, a disciplined plan for employee redeployment or retraining is essential. Many healthcare organizations fail at this critical step, as they lack processes to capture and maintain gains in productivity and quality improvement. Although this connection may not be readily apparent, the presence of such a plan aids in an organization’s ability to hold its improvement gains.
As part of the executive function of a healthcare organization, the HR department serves as a strategic partner in making effective and long-lasting change. During each annual planning cycle, strategic projects to further the goals of the organization are identified. Many of these initiatives become part of the balanced scorecard. At this point, the HR department should be included to undertake planning to place the right person in the right job at the right time. This process is shown in exhibit 15.1.
Projectidentified
Decreasestaffing?
Plan for maintaining staff
Retrain and poolor redeploy
Eliminate vacantposition
FTEsneededin other
department?
Vacantposition?
Lay off
No
Yes
Yes
Yes
No
No
Note: FTE = full-time equivalent.
EXHIBIT 15.1Process for HR
Planning
Chapter 15: Holding the G ains 393
The HR staff need to estimate the impact of each project or initiative that will be undertaken during the year. If the project has a goal of providing more service with the same number of staff members, the HR task will be to maintain this staffing level. Broader HR planning can now occur, such as tracking the availability of workers for these positions in external labor pools or identifying and training existing employees to fill these roles if turnover occurs.
If, on the other hand, the likely outcome of a project will be to reduce staff in a department, ensuring clarity about the next steps is important. If unfilled positions are no longer needed, the most prudent step is to eliminate them. However, if the position is currently filled, existing employees need to be transferred to different departments in need of full-time equivalents. If no openings exist in other departments, these employees may become part of a pool of employees used to fill temporary shortages inside the organization. Retraining for other open positions is also an option if the displaced employee has related skills. Because they have just participated in process improvement projects, these staff members may also receive additional training in process improvement tools and be assigned to other departments to aid in their projects.
If none of these options is feasible, the last action available to the man-ager is to lay off the employee. Executing projects that will clearly result in job loss is difficult—getting employees to redesign themselves out of a job is almost impossible. However, layoffs can generally be avoided in healthcare, as labor shortages are widespread. In addition, most projects identified should be of the first type, those that will increase throughput with existing staff, as these tend to be the most critical for improved patient access and increases in the quality of clinical care.
The HR planning function should be ongoing and comprehensive, and a well-communicated plan for employee reassignment and replacement should be in place. By identifying all potential projects during the annual planning cycle, the HR department can develop an organization-wide staffing plan. Without this critical function, many of the gains in operating improvements will be lost.
Managerial AccountingThe second key tool for holding the gains is the use of managerial account-ing (Gapenski and Reiter 2016, part III). In contrast to financial accounting, which is used to prepare financial statements (the past), managerial accounting focuses on the future. Managerial accounting can be used to anticipate the profitability of a project intended to improve patient flow or model the revenue gains from a clinical pay-for-performance (P4P) contract. Even projects that appear to have no financial impact can benefit from managerial accounting. For example, a project to reduce hospital-acquired infections may not only provide improvements in the quality of care but also reduce the length of stay for a number of patients and therefore increase the hospital’s profitability.
Managerial accountingThe field of accounting that focuses primarily on subunit (i.e., departmental) data used internally for managerial decision making.
Healthcare Operat ions Management394
Managerial accounting is a primary analytical tool to reduce costs and increase revenue, as discussed in chapter 14.
Having a member of the finance staff engaged with operations improve-ment efforts is useful. This team member should perform an initial analysis of the expected financial results for a project and monitor the financial model throughout the project. She should also ensure that the financial effects of an individual project flow through to the financial results for the entire organiza-tion. The use of predictive analytics and business modeling (chapter 8) can help in evaluating the risks and rewards associated with various projects or decisions.
The first step in managerial accounting is to understand an operating unit’s revenue source and how it changes with a change in operations. For example, capitation revenue may flow to a primary care clinic; in this case, a reduction in the volume of services will result in a profitability gain. However, if the revenue source for the clinic is fee-for-service payments, the reduction in volume will result in a revenue loss.
Evaluating many revenue sources in healthcare can be complex. For example, understanding inpatient hospital reimbursement via diagnosis-related group can be difficult, as some diagnoses pay substantially more than others. In addition, many rules affect net reimbursement to the hospital, so a com-prehensive analysis must be undertaken.
The trend toward consumer-directed healthcare and healthcare savings accounts means that the retail price of some services also affects net revenue. If a market-sensitive outpatient service is priced too high, net revenue may decline as consumer demand decreases.
Next, the costs for the operation must be identified and segmented into three categories: variable, fixed, and overhead. Variable costs are those that vary with the volume of the service; a good example is supplies used with a procedure. Fixed costs are those that do not vary with volume and include such items as space costs and equipment depreciation. Employee wages and benefits are usually designated as fixed costs, although they may be variable if the volume of services changes substantially and staffing levels are adjusted on the basis of volume.
The final cost category is overhead, which is allocated to each depart-ment or unit in an organization that generates revenue. This allocation pays for costs of departments that do not generate revenue. Knowing which overhead formulas are used to allocate costs is critical to understanding the impact of operational changes. For example, an overhead rate based on a percentage of revenue has a substantially different effect than one based on the square foot-age a department occupies.
The next step in the managerial accounting process is to conduct a cost-volume-profit (CVP) analysis. Exhibit 15.2 illustrates a CVP analysis of two outpatient services at VVH.
Cost-volume-profit (CVP) analysisA managerial accounting method used to evaluate the impact of cost and volume on profit in an organizational unit.
Chapter 15: Holding the G ains 395
In the first case, the service is backlogged and current profit (base case) is $7,000 per year. However, if a process improvement project is undertaken, the volume can be increased from 1,000 to 1,500 tests per year. If staffing and other fixed costs remain constant, the net profit is increased to $63,000 per year.
The second example shows a situation in which the service is operating at an annual loss of $28,000. In this case, the process improvement goal is to reduce fixed costs (staffing) with a slight increase in volume. The result is a $40,000 reduction in fixed cost, which yields a profit margin of $17,600. HR planning is critical in a project such as this to ensure a comfortable transition for displaced employees.
Control SystemThe final key to holding the gains is a control system. Control systems have two major components: measurement/reporting and monitoring/response.
Chapter 6 discusses many tools for data capture and analysis with an objective of finding and fixing problems. Many of the same tools should be deployed for continuous reporting of the results of operations improvement projects. Data collection systems for monitoring outcomes should be built into any operations improvement project from the beginning.
Backlogged Financial Loss
Base
Process Improvement
Project Base
Process Improvement
Project
Test volume 1,000 1,500 1,000 1,050
Revenue/test $150 $150 $150 $150
Total revenue $150,000 $225,000 $150,000 $157,500
Costs
Variable cost/unit $38 $38 $38 $38
Fixed costs $85,000 $85,000 $120,000 $80,000
Overhead $20,000 $20,000 $20,000 $20,000
Total cost $143,000 $162,000 $178,000 $139,900
Profit $7,000 $63,000 ($28,000) $17,600
Note: CVP = cost-volume-profit.
EXHIBIT 15.2Managerial Accounting: CVP Analysis
Healthcare Operat ions Management396
Once data collection is under way, results should be displayed both numerically and graphically. The run chart (chapter 9) is still one of the most effective tools for monitoring the performance of a process. Exhibit 15.3 illustrates a simple run chart for birthing center patient satisfaction, where a goal of greater than 90 percent satisfied patients has been set. This type of chart can show progress over time to ensure that the organization is moving toward its goals.
In addition to a robust data capture and reporting system, a plan for monitoring and response is critical. This plan should include identification of the individual or team responsible for the operation and a method for com-municating the reports to them. In some cases, these operations improvement activities are of such strategic importance that they become part of a depart-mental or organization-wide balanced scorecard.
A response procedure or plan should be developed to address situations in which a process fails to perform as it should. Jidoka and andon systems (chapter 10) can help organizations discover and correct problems with system performance. Control charts (chapter 9) can be used to identify out-of-control situations. Once an out-of-control situation is identified, action should be taken to determine the special or assignable cause and eliminate it.
50
60
70
80
90
100
July
August
September
October
November
December
January
Febru
ary
March April
May
June
Month
Perc
enta
ge S
atis
fied
Facilities
Clinical quality
High touch
EXHIBIT 15.3Run Chart
for Birthing Center Patient
Satisfaction
Chapter 15: Holding the G ains 397
Which Tools to Use: A General Algorithm
This book presents an array of techniques, tools, and methods to achieve operational excellence. How does the practitioner choose from this broad array? As in clinical care, a mix of art and science is involved in choosing the best approach.
A general algorithm for selecting tools is presented below, and the book’s companion website contains an automated and more detailed version. The framework for this detailed path through the logic (exhibit 15.4) is represented by a series of steps.
Step A. Issue FormulationFirst, formulate the issue you wish to address. Determine the current state and a desired state (e.g., competitors have taken 5 percent of our market share in obstetrics, and we want to recapture the market; the pediatric clinic lost $100,000 last year, and we want to break even next year; public rankings for our diabetes care place our clinic below the median, and we want to be in the top quartile). Framing the problem correctly is important to ensure that the outcome is the right solution to the right issue rather than the right answer to the wrong question; all relevant stakeholders should be consulted at this step.
A number of effective decision-making and problem-solving tools can be used to
• frame the question or problem,• analyze the problem and various solutions to the problem, and• implement those solutions.
The tools and techniques identified next provide a basis for tackling difficult, complicated problems.
• The decision-making process: a generic decision process used for any type of process improvement or problem solving (plan-do-check-act [PDCA], define-measure-analyze-improve-control [DMAIC], and project management all follow this same basic outline) – Framing: used to ensure that the correct problem or issue is being
addressed – Gathering intelligence: finding and organizing the information
needed to address the issue (data collection) – Coming to conclusions: determining the solution to the problem
(data analysis)
On the web at ache.org/books/OpsManagement3
Healthcare Operat ions Management398
EXH
IBIT
15.
4Al
gori
thm
for
Use
of t
he T
ools
, Tec
hniq
ues,
and
Met
hodo
logi
es in
Thi
s B
ook
NO
TE: M
odel
cre
ated
in M
icro
soft
Vis
io®.
Chap
ter
num
bers
are
in p
aren
thes
es.
Star
t
Issu
e fo
rmul
atio
n(6
)
Is th
e is
sue
stra
tegi
c?
Larg
epr
ojec
t?
End
No
Dat
a
Use
Six
Sig
ma
tool
s (9
)
Appl
y ba
sic
proc
ess
impr
ovem
ent t
ools
(11)
Dat
a
Use
SCM
(13)
Dev
elop
and
use
a
bala
nced
sco
reca
rd (4
)
Dat
a
Sche
dulin
g(1
2)
Yes
Yes
Yes
Focu
s on
impr
ovin
g qu
alit
yor
redu
cing
vari
atio
n?
Focu
s on
redu
cing
was
teor
impr
ovin
gflo
w?
Use
form
al p
roje
ctm
anag
emen
t app
roac
h (5
)
Yes
Dat
a co
llect
ion
and
anal
ysis
is u
sed
inco
njuc
tion
wit
h a
num
ber o
f ope
rati
ons
impr
ovem
ent t
ools
,an
d th
eir u
se is
indi
cate
d by
this
box
Dat
a co
llect
ion
and
anal
ysis
(7)
Focu
son
wha
t-if
issu
es?
Yes
Yes
Dat
a
Use
Lea
nto
ols
(10)
Yes
Yes
Focu
s on
supp
ly c
hain
No
Hol
d th
ega
ins
(15)
Dat
a
Use
anal
ytic
s(8
)
Focu
s on
sc
hedu
ling?
No
No
No
No
No
EXH
IBIT
15.
4Al
gori
thm
for U
sing
the
Tool
s, T
echn
ique
s, a
nd M
etho
dolo
gies
in T
his
Boo
k
Not
e: C
hapt
er n
umbe
rs a
re in
par
enth
eses
. Mod
el c
reat
ed in
Mic
roso
ft V
isio
. SCM
= s
uppl
y ch
ain
man
agem
ent.
Chapter 15: Holding the G ains 399
– Learning from feedback: ensuring that learning is not lost and that the solution actually works (holding the gains)
• Mapping tools
– Mind mapping: used to help formulate and understand the problem or issue
– Process mapping, activity mapping, and service blueprinting: used to “picture” the system and process steps
• Root-cause analysis (RCA) tools
– Five whys technique and fishbone diagrams: used to identify causes and root causes of problems to determine how to eliminate those problems
– Failure mode and effects analysis: a more detailed root cause–type analysis used to identify and plan for both possible and actual failures.
Step B. Strategic or Operational IssueNext, decide whether the issue is strategic (e.g., major resources and high-level staff will be involved) or part of ongoing operations.
If the issue is strategic, go to step C, balanced scorecard for strategic issues. If it is operational, go to step D, project management, or E, basic performance improvement tools, depending on the size and scope of possible solutions.
To effectively implement a major strategy, develop a balanced scorecard to link initiatives and measure progress.
Step C. Balanced Scorecard for Strategic IssuesTo effectively implement a major strategy, develop a balanced scorecard to link initiatives and measure progress. Elements of the balanced scorecard include the following:
• Strategy map—used to link initiatives or projects to achieve the desired state
• Four perspectives—ensures that initiatives and projects span the four main perspectives of the balanced scorecard, including financial, customer/patient, operations, and employee learning and growth
• Metrics—used to measure progress through leading (predictive) and lagging (results) indicators
If the balanced scorecard contains a major initiative, go to step D; otherwise, go to step E.
Healthcare Operat ions Management400
Step D. Project ManagementThe formal project management methodology should be used for initiatives that typically last longer than six months and involve a project team. Project management includes the following tools:
• Project charter—a document that outlines stakeholders, the project sponsor, the project mission and scope, a change process, expected results, and estimated resources required
• Work breakdown structure—a list of tasks for accomplishing the project goals, with assigned responsibilities and estimated durations and costs
• Schedule—a progression of tasks in order of precedence and linked by relationship, and the identification of the critical path that determines the overall duration of the project
• Change control—a method by which to formally monitor progress and make changes during the execution of a project
• Risk management—the identification of project risks and plans to mitigate each risk
If the project is primarily concerned with improving quality or reducing variation, use the project management technique and tools described in step F, quality and Six Sigma. If the operating issue is large enough for project man-agement and primarily concerned with eliminating waste or improving flow, go to step G, Lean. If the issue is related to evaluating and managing risk or analyzing and improving processes, go to step H, analytics. If the project is focused on supply chain issues, go to step I, supply chain management (SCM). If the project focus is not encompassed by Six Sigma, Lean, simulation, or SCM, return to step E and use the basic performance improvement tools in the larger project management system.
Step E. Basic Performance Improvement ToolsBasic performance improvement tools are used to improve and optimize a process. In addition to RCA, the following tools can be helpful in moving toward effective and efficient processes and systems.
• Optimization using linear programming—used to determine the optimal allocation of scarce resources.
• Theory of constraints (TOC)—five steps for identifying and managing constraints in the system:
1. Identify the constraint (or bottleneck).
2. Exploit the constraint by determining how to get the maximum performance out of the constraint without major system changes or capital improvements.
Chapter 15: Holding the G ains 401
3. Subordinate everything else to the constraint by synchronizing other nonbottleneck resources (or steps in the process) to match the output of the constraint.
4. Elevate the constraint by taking some step (e.g., capital expenditure, staffing increase) to increase the capacity of the constraining resource until it is no longer the constraint and another activity becomes the new constraint.
5. Repeat the process for the new constraint.• Force field analysis—used to identify and manage the forces working for
and against change (applicable to any change initiative, including TOC, Six Sigma, and Lean).
If these tools provide an optimal solution, go to step J, holding the gains. Sometimes the operating issues are so large that they will benefit from the formal project management discipline. In this case, go to step D. If the project is relatively small and focused on eliminating waste, go to step G, where the kaizen event tool can be used to achieve quick improvements.
Step F. Quality and Six SigmaThe focus of quality initiatives and the Six Sigma methodology is on improving quality, eliminating errors, and reducing variation.
• DMAIC—the five-step process improvement or problem-solving technique used in Six Sigma:
1. Define the problem or process (see step A, issue formulation).
2. Measure the current state of the process (see the section titled Data and Statistics later in this chapter).
3. Analyze the collected data to determine how to fix the problem or improve the process.
4. Improve the process or solve the problem.
5. Control to ensure that changes are embedded in the system (see step J).
Note that at any point in the process, looping back to a previous step may be necessary. Once the process is complete, start the loop again.
• Seven basic quality tools—in the DMAIC process, tools used to improve the process or solve the problem:
1. Fishbone diagram, for analyzing and illustrating the root causes of an effect.
2. Check sheet, a simple form used to collect data.
Healthcare Operat ions Management402
3. Histogram, a graph used to show frequency distributions.
4. Pareto diagram, a sorted histogram.
5. Flowchart, a process map.
6. Scatter plot, a graphic technique to analyze the relationship between two variables.
7. Run chart, a plot of a process characteristic in chronological sequence.
• Statistical process control—an ongoing measurement of process output characteristics for ensuring quality that enables the identification of a problem situation before an error occurs.
• Process capability—a measure of whether a process is capable of producing the desired output.
• Benchmarking—the determination of what is possible on the basis of what others are doing; used for comparison purposes and goal setting.
• Quality function deployment—used to match customer requirements (voice of the customer) with process capabilities, given that trade-offs must be made.
• Poka-yoke—mistake proofing.
Once these tools have produced satisfactory results, proceed to step J, holding the gains.
Step G. LeanLean initiatives are typically focused on eliminating waste and improving flow in the system or process.
• Kaizen philosophy—the five-step process improvement technique used in Lean:
1. Specify value by identifying activities that provide value from the customer’s perspective.
2. Map and improve the value stream by determining the sequence of activities or the current state of the process and the desired future state, and eliminating non-value-added steps and other waste.
3. Initiate flow, enabling the process to proceed as smoothly and quickly as possible.
4. Pull to enable the customer to trigger movement of products or services toward them.
Chapter 15: Holding the G ains 403
5. Build perfection by repeating the cycle to ensure a focus on continuous improvement.
• Value stream mapping—used to define the process and determine where waste is occurring.
• Takt time—a measure of time needed for the process on the basis of customer demand.
• Throughput time—a measure of the actual time needed in the process.• Five Ss—a technique to organize the workplace.• Spaghetti diagram—a mapping technique to show the movement of
customers (patients), workers, equipment, and so on.• Kaizen blitz or event—used to improve the process quickly, when
project management is not needed.• Standardized work—written documentation of the precise way in which
every step in a process should be performed; a way to ensure that activities are completed the same way every time in an efficient manner.
• Jidoka and andon—techniques or tools used to ensure that “things are done right the first time” to catch and correct errors.
• Kanban—a scheduling tool used to pull rather than push work.• Single-minute exchange of die—a technique to increase the speed of
changeover.• Heijunka—leveling production (or workload) so that the system or
process can flow without interruption.
Once these tools have produced satisfactory results, proceed to step J.
Step H. AnalyticsBig data and advanced analytics can be used to evaluate what-if situations. Usually, these data tools are less expensive or speedier than the cost or time needed to change the real system and evaluate the effects of those changes.
The analytics process approach consists of the following steps:
1. Develop an understanding of the data by using descriptive tools such as dashboards, key performance indicators, and scorecards. Use advanced software to perform data visualization.
2. Develop predictive models using statistical modeling and alternative data models.
3. Develop business solutions using prescriptive or analytical models. Use software tools to choose the best solution.
Once these tools have produced satisfactory results, proceed to step J.
Healthcare Operat ions Management404
Step I. Supply Chain ManagementSCM focuses on all of the processes involved in moving supplies and equip-ment from the manufacturer to their use in patient care areas. SCM is the management of all activities and processes related to both upstream vendors and downstream customers in the value chain. Effective and efficient manage-ment of the supply chain requires an understanding of all of the following:
• Tools for tracking and managing inventory• Forecasting• Inventory models• Inventory systems• Procurement and vendor relationship management• Strategic SCM
Once these tools have produced satisfactory results, proceed to step J.
Step J. Holding the GainsUpon successful completion of operational improvements, the three tools introduced at the beginning of this chapter can be used to ensure that the changes endure:
• HR planning—a disciplined approach to using employees in new ways after an improvement project is completed
• Managerial accounting—a study of the expected financial consequences and gains after an operations improvement project has been implemented
• Control system—a set of tools to monitor the performance of a new process and methods to take corrective action if desired results are not achieved
Data and Statistics
All of the aforementioned tools, techniques, and methodologies require data and data analysis. Tools and techniques associated with data collection and analysis include the following:
• Data collection techniques—used to ensure that valid data are collected for further analysis
• Graphic display of data—used to “see” the data• Mathematical descriptions of data—used to compare sets of data and for
simulation
Chapter 15: Holding the G ains 405
• Statistical tests—used to determine whether differences in data are present
• Regression analyses—used to investigate and define relationships among variables
• Forecasting—used to predict future values of random variables
Operational Excellence
Many leading hospitals, medical groups, and health plans are using the tools and techniques contained in this book. However, these tools have not seen widespread use in healthcare, nor have they been as comprehensively applied as in other sectors of the economy. We have developed a scale for the application of these tools to gauge progress toward comprehensive operational excellence in healthcare.
Level 1No organized operations monitoring or improvement efforts are present at level 1. Quality efforts are aimed at compliance and the submission of data to regulating agencies.
Level 2At level 2, the organization has begun to use operations data for decision mak-ing. Pockets of process improvement activity occur where process mapping and PDCA or rapid prototyping are adopted. Evidence-based medicine (EBM) guidelines are used in some clinical activities.
Level 3Senior management has identified operations improvement efforts as a priority in level 3. The organization conducts operations improvement experiments, uses a disciplined project management methodology, and maintains a com-prehensive balanced scorecard. Some P4P bonuses are received from payers, and the organization obtains above-average scores on publicly reported quality measures.
Level 4A level 4 organization engages in multiple process improvement efforts using a combination of project management, Six Sigma, Lean, and simulation tools. It has trained a significant number of employees in the advanced use of these tools, and these individuals lead process improvement projects. EBM guidelines are comprehensively used, and all P4P bonuses are achieved.
Healthcare Operat ions Management406
Level 5Operational excellence is the primary strategic objective of an organization at level 5. The executive leadership team has embraced operational excellence as a key component of the organization’s strategic plan and demonstrates knowledge in all of its tools. Operations improvement efforts are under way in all departments, led by departmental staff who have been trained in advanced tools. The organization uses real-time simulation to control patient flow and operations. New EBM guidelines and best practices for administrative opera-tions are developed and published by this organization, which scores in the top 5 percent of any national ranking on quality and operational excellence.
A few leading organizations currently are at level 4, but most reside between levels 2 and 3. Our friends at VVH are at the top of level 3 and mov-ing toward level 4.
Vincent Valley Hospital and Health System Strives for Operational ExcellenceAs presented in chapter 3, VVH leadership believes it has a number of oppor-tunities to succeed with the Hospital Value-Based Purchasing program and has added the program as an initiative to its corporate balanced scorecard, as follows: “Conduct projects to optimize Medicare value-based purchasing to generate at least a 2 percent increase in inpatient revenue.”
VVH has reorganized its structure to combine a number of operations and quality activities into a new organization-wide department known as opera-tions management and quality.
One team is being created to target the following specific measures for improvement:
• Pneumonia patients assessed for and given a pneumococcal vaccination• Pneumonia patients whose initial emergency department blood culture
was performed prior to the administration of the first hospital dose of antibiotics
• Pneumonia patients given smoking cessation advice and counseling• Pneumonia patients given initial antibiotic(s) within six hours after
arrival• Pneumonia patients given the most appropriate initial antibiotic(s)• Pneumonia patients assessed for and given an influenza vaccination
The first step in the project is to identify this team and develop a proj-ect charter and schedule (chapter 5). Both the HR and finance departments are to be included in the project team to model financial consequences (new
Chapter 15: Holding the G ains 407
revenues, possible new costs, capital requirements) and the potential effect on staffing levels.
The project team begins by collecting data on current performance and summarizing them using visual and mathematical techniques to determine where performance does not meet goals (chapters 7 and 8). A process map is constructed and analyzed to determine where processes may be improved to achieve the desired results. Various Six Sigma tools (fishbone diagrams, check sheets, Pareto diagrams, and scatter plots) are employed to further analyze and improve the process (chapter 9).
The clinicians on the project team perform a careful analysis to deter-mine which areas of the treatment of patients at risk of pneumonia can be standardized and which need customization. The standard modules are then examined for both effectiveness and efficiency using value stream mapping (chapter 10).
Changes are identified, many of them requiring either a staffing adjust-ment or a change in VVH’s electronic health record. Because many options are available and the team is uncertain which will achieve the desired results, a decision tree (chapter 6) is constructed to identify the optimal process improvements. Finally, once the project team begins to implement these pro-cess improvements, the results are monitored with control charts (chapter 9).
The Healthcare Organization of the Future
A future healthcare organization operating at level 5 is illustrated in exhibit 15.5. This care delivery system will use many of the tools and techniques con-tained in this text. A demand prediction model will generate predictions of demand for inpatient and ambulatory care services. Because much of the care delivered in these sites will be through the use of EBM guidelines (chapter 3) that have optimized processes (chapters 7 to 10), the resource requirements can be predicted as well; these predictions will drive scheduling and supply chain systems.
A key component of this future system is a real-time operations monitor-ing and control system. This system uses simulation and modeling techniques to monitor, control, and optimize patient flow and diagnostic and treatment resources. Macro-level control systems such as the balanced scorecard (chapter 4) ensure that this system meets the organization’s strategic objectives. The result will be a finely tuned healthcare delivery system providing high-quality clinical care in the most efficient manner possible.
Healthcare Operat ions Management408
Conclusion
We hope that this text is helpful to you and your organization on your journey toward level 5 operational excellence. We are interested in your progress whether you are a new member of the health administration team, a seasoned depart-
ment head, or a physician leader—please use the e-mail addresses provided on the companion website to inform us of your successes, and let us know what we could do to make this a better text.
Because many of the tools discussed in this text are evolving, we will continuously update the companion website with revisions and additions; check it frequently. We, too, are striving to reach level 5.
Discussion Questions
1. Identify methods to reduce employees’ resistance to change during an operations improvement project.
2. What should be the key financial performance indicator used to analyze performance changes for hospitals? Clinics? Health plans? Public health agencies?
3. Describe tools (other than control charts) that can be used to ensure that processes achieve their desired results.
On the web at ache.org/books/OpsManagement3
Demand predictionsystem:
Volume—clinicalconditions
Ambulatorycare model—EBM based
Emergency andinpatient caremodel—EBM
based
Predicted resourceneeds:
• Facilities• Staff• Supplies
Real-time data
Real-time control
Staff schedulingsystem
Supply chainsystem
Real-time operationsmonitoring and control
Clinicaloperations
EXHIBIT 15.5An Optimized
Healthcare Delivery System
of the Future
Chapter 15: Holding the G ains 409
4. Describe the tools, methods, and techniques in this book that would be used to address the following operating issues:a. A hospital laboratory department provides results that are late and
frequently erroneous.b. A clinic’s web-based patient information system is not being used by
the expected number of patients.c. An ambulatory clinic is financially challenged but has a low staffing
ratio compared to that of similar clinics.
Case Study
VVH has a serious problem: A major strategic objective of the health system is to grow its ambulatory care network, but the organization faces a number of challenges in doing so. Although a new billing system was installed and various reimbursement maximization strategies were executed, total costs in the system exceed revenue, even as the clinic staff feel busy and backlog appointments have increased in number. Analysis of clinic data indicates a growing number of patients are canceling appointments or are no-shows.
In addition, a new group of multispecialty and primary care physicians has been created from the merger of three separate groups; this clinic is aggres-sively competing with VVH for privately insured patients. The new large clinic is making same-day clinic appointments available and heavily advertising them.
The board of VVH has asked the CEO to develop a plan to address this growing concern. The CEO begins by forming a small strategy team to lead improvement efforts; its first step is to assign the chief operating officer, chief financial officer, and medical director to direct the planning and finance staff on the improvement team.
VVH ultimately decides that it needs to increase the number of patients seen by clinicians and begins to implement advanced-access scheduling in its clinics. Because VVH believes in knowledge-based management and the shar-ing of improved methods of delivering health services, the organization has made its data and information available on the book’s companion website. VVH has invited students and practitioners to help the organization improve this system.
Case Study Questions1. Frame the original issue for VVH. Mind maps and RCA may be useful
here.2. How would you address the no-show and cancelation issues?3. Develop a project charter for one project associated with VVH’s problems.
On the web at ache.org/books/OpsManagement3
Healthcare Operat ions Management410
4. Develop a balanced scorecard for VVH’s clinics.5. If VVH were to focus on increasing throughput in the system, how
would you go about doing so? Be specific.
References
Fried, B. J., and M. D. Fottler. 2015. Human Resources in Healthcare: Managing for Success, 4th edition. Chicago: Health Administration Press.
Gapenski, L. C., and K. L. Reiter. 2016. Healthcare Finance: An Introduction to Accounting and Financial Management, 6th edition. Chicago: Health Administration Press.
411
GLOSSARY
Activity-based costing (ABC). A cost allocation model that assigns a cost to each activity in an organizational unit and then totals the cost for the unit on the basis of the actual consumption of each activity.
Advanced-access scheduling. A method of scheduling outpatient appointments that provides open time slots every day for seeing patients on the same day they request an appointment. Also known as same-day scheduling.
Agency for Healthcare Research and Quality (AHRQ). A federal agency that is part of the Department of Health and Human Services. It provides leadership and funding to identify and communicate the most effective methods to deliver high-quality healthcare in the United States.
Andon. A visual or audible signaling device used to indicate a problem in the process, typically used in conjunction with jidoka.
Balanced scorecard. A system of strategy links and reporting mechanisms that supports effective strategy execution.
Bayes’ theorem. A formula used to revise the calculation of conditional prob-ability as new information is obtained in the situation.
Business intelligence. The process of converting raw data through a variety of methods into information that can assist with decision making.
Capacity utilization. The percentage of time that a resource (worker, equip-ment, space, etc.) or process is actually busy producing or transforming output.
Care path. A sequence of best practices for healthcare staff to follow for a diag-nosis or procedure, designed to minimize waste and maximize quality of care.
Central limit theorem. A theory demonstrating that as the sample size from a population becomes sufficiently large, the sampling distribution of the means approaches normality, no matter the distribution of the original variable.
Coefficient of determination. The measure of how well a model fits the data.
Coefficient of variation (CV). A measure of variation in the data relative to the measure of central tendency in the data.
Confidence interval (CI). The probability that a population parameter falls between two values.
Consumer-directed healthcare. In general, the consumer (patient) is well in-formed about healthcare prices and quality and makes personal buying decisions
412 Glossary
on the basis of this information. The health savings account is frequently in-cluded as a key component of consumer-directed healthcare.
Contingency table. A tool used to examine the relationships between qualita-tive or categorical variables.
Continuous quality improvement (CQI). A comprehensive quality improve-ment and management system with three key components: planning, control, and improvement.
Control limits. Common variation limits that are ±3 standard deviations from the mean.
Correlation coefficient. A measure of the linear relationship between two variables.
Cost of quality. The costs associated with producing poor-quality goods and services, including tangible costs, such as scrap and rejects, and intangible costs, such as lost customer goodwill.
Cost-volume-profit (CVP) analysis. A managerial accounting method used to evaluate the impact of cost and volume on profit in an organizational unit.
Critical path method (CPM). The critical path is the longest course through a graph of linked tasks in a project. The critical path method is used to reduce the total time of a project by decreasing the duration of tasks on the critical path.
Cross-functional process map. A map that follows the flow of a process through the various departments of the organization using dashed lines to show the work being completed by a particular department or individual in the pro-cess. Also called swim lane process map.
Cycle time. The time required to accomplish a task in a system.
Decision analysis. A structured process for examining and evaluating decisions.
Decision tree. A graphical representation of the order of future and current events for how decisions are made.
Dot plot. A chart in which frequency is represented by a dot. Useful for dis-playing small data sets with positive values.
Economic order quantity (EOQ). An inventory model that indicates an opti-mal purchase quantity that will minimize total annual inventory costs.
Enterprise resources planning (ERP). Global information systems that help individuals and groups manage the entire organization, including accounting, operations, and human resources.
Evidence-based medicine (EBM). The conscientious and judicious use of the best current evidence in making decisions about the care of individual patients.
Failure mode and effects analysis (FMEA). A technique developed by the US military to identify the ways in which a process (or piece of equipment) might fail and to determine how best to mitigate those risks.
Fishbone diagram. A graphical technique used to display the relationship be-tween the potential causes of a problem and the effect created by the problem. Sometimes called Ishikawa diagram.
413Glossary
Five whys technique. A technique that uses a series of logical questions to find the root cause of a problem.
Force field analysis. A graphical technique that demonstrates all the forces for and against making a key change.
Full capitation. A methodology in which providers are paid a monthly fee for each patient who receives care in their system.
Gantt chart. A scheduling tool that lists project tasks, with bars indicating start and end dates for each task.
Health savings account (HSA). A personal monetary account that can only be used for healthcare expenses. The funds are not taxed, and the balance can be rolled over from year to year. HSAs are normally used with high-deductible health insurance plans.
Heijunka. The process of eliminating variations in volume and variety of pro-duction to reduce waste.
Histogram. A graph summarizing discrete or continuous data. Histograms visually display how much variation exists in the data.
Hypothesis testing. The process of testing a statistical distribution parameter against that of another distribution parameter to assess if statistical differences exist in the data.
Institute of Medicine (IOM). The healthcare arm of the National Academy of Sciences; an independent, nonprofit organization providing unbiased and authoritative advice to decision makers and the public.
ISO 9000. A series of process standards developed by the International Orga-nization for Standardization to give organizations guidelines for developing and maintaining effective quality systems.
Jidoka. The ability to prevent defects by stopping a process when an error occurs.
Just-in-time (JIT). An inventory management system designed to improve ef-ficiency and reduce waste. Part of Lean manufacturing.
Kaizen. Continuous improvement based on the beliefs that everything can be improved and that incremental changes result in an enhanced system.
Kaizen event. A focused, short-term project aimed at improving a particular process.
Kanban. A visual signal that triggers the movement of inventory or product in a system.
Knowledge hierarchy. The foundation of knowledge-based management, com-posed of five categories of learning: data, information, knowledge, understand-ing, and wisdom.
Lagging indicator. A performance measurement that assesses the outcome of existing actions.
Leading indicator. A performance measurement that predicts the future and is specific to an initiative or organizational strategy. Also called performance driver.
414 Glossary
Linear programming. A mathematical technique used to find the optimal solu-tion to a linear problem given a set of constrained resources.
Little’s law. The relationship between the arrival rate to a system, the time an item (e.g., a patient) spends in the system, and the number of items in a system.
Malcolm Baldrige National Quality Award. An annual award established by the US Congress in 1987 to recognize organizations in the United States for their achievements in quality.
Managerial accounting. The field of accounting that focuses primarily on sub-unit (i.e., departmental) data used internally for managerial decision making.
Material requirements planning (MRP). A computer system designed to manage the purchase and control of dependent-demand items.
Mind mapping. A nonlinear technique used to develop thoughts and ideas by placing pictures or phrases on a map to show logical connections.
Mitigation plan. A set of tasks intended to reduce or eliminate the effect of risk in a project.
Network diagram. A scheduling tool that connects tasks in order of precedence.
Observed probability. The number of times an event occurred divided by the total number of trials.
Optimization. A technique used to determine the ideal allocation of limited re-sources (such as people, money, or equipment) given a desired goal. Also called mathematical programming.
Pareto diagram. A rank-ordered frequency chart that indicates the number of times a particular item occurs in a situation.
Pareto principle. Developed by Italian economist Vilfredo Pareto in 1906 on the basis of his observation that 80 percent of the wealth in Italy was owned by 20 percent of the population.
Patient care microsystem. The level of healthcare delivery that includes provid-ers, technology, and treatment processes.
Patient-centered medical home (PCMH). Care that is accessible, continu-ous, comprehensive, family centered, coordinated, compassionate, and culturally effective.
Plan-do-check-act (PDCA). A core process improvement tool with four ele-ments: Plan a change to a process; enact the change; check to make sure it is working as expected; and act to make sure the change is sustainable. PDCA functions as a continuous cycle and, as such, is sometimes referred to as the Deming wheel.
Poka-yoke. A mechanism that prevents mistakes or makes them immediately obvious to prevent adverse outcomes.
Practical significance. The differences in the parameters of two data sets are large enough to be meaningful for the person or organization studying the situ-ation, whether or not they are statistically significant.
415Glossary
Prevention quality indicator (PQI). A set of measures that can be used with hospital discharge data to identify patients whose hospitalizations or complica-tions might have been avoided with the use of evidence-based ambulatory care.
Process capability. A measure of how well a process can produce output that meets desired standards or specifications.
Process map. A graphic depiction of a process showing the sequence of events, including tasks, decisions, and other activities from inputs to outputs. A process map is a type of flowchart.
Program evaluation and review technique (PERT). A graphic technique to link and analyze all tasks within a project; the resulting graph helps optimize the project’s schedule.
Public reporting. A statement of healthcare quality made by hospitals, long-term care facilities, and clinics. May also include patient satisfaction and provider charges.
Quality function deployment (QFD). A technique that translates customer requirements to specific product or process requirements.
Queue discipline. In queuing theory, the method by which customers are selected from the queue to be served.
Queuing theory. The mathematical study of wait lines.
Range (r) chart. Measures process performance of sample ranges for continu-ous data.
RASIC. A chart delineating all project team members’ roles for each task in a project. The acronym comes from the members’ roles: responsible, approval, support, informed, consult.
Revenue cycle. Generating charges, issuing bills, and managing payments and receivables for a defined period.
Risk adjustment. Raising or lowering fees paid to providers on the basis of fac-tors that may increase medical costs, such as age, sex, or illness.
Risk management. Within a project, the identification of possible events that, if realized, will affect the execution of the project and a plan to mitigate these events.
Rolled throughput yield (RTY). The probability that a unit (of product or service) will pass through all process steps free of defects.
Root-cause analysis (RCA). A generic term describing structured, step-by-step techniques for problem solving.
Rough-cut capacity planning. The process of converting the overall produc-tion plan into capacity needs for key resources.
Scatter plot. A graph displaying two variables that indicates whether they are related, how strongly they are related, and the direction of the relationship.
Scientific management. A disciplined approach to studying a system or pro-cess and then using data to optimize it to achieve improved efficiency and effectiveness.
416 Glossary
Sensitivity analysis. A tool that examines the impact of independently changing input variables to see their effect on the output of a model.
Sequencing rules. Heuristic rules that indicate the order in which jobs are pro-cessed from a queue. Also known as queuing priority.
Service blueprinting. A style of process mapping that separates actions into onstage (visible to the customer) and backstage (not visible to the customer) activities.
Service level. The probability of having an item on hand when needed.
Shared savings model. A model of healthcare delivery that includes an orga-nized system of delivery, accountability for the quality and costs of services, and a sharing of savings with the payer for these services.
Shewhart’s rule. An outlier exists in bell-shaped data if a data point is greater than three standard deviations from the mean.
Simple linear regression. An equation that relates two variables using a slope and an intercept in a linear fashion.
Single exponential smoothing (SES). A simple forecasting model that smooths data in a time series to predict the future.
Spaghetti diagram. A visual representation of the movement or travel of mate-rials, employees, or customers.
Stakeholder. Anyone who has a vested interest in the outcome of a project, in-cluding (but not limited to) employees, customers, users, partner organizations, project sponsors, and the project manager.
Standard deviation. A measurement of variation around the mean.
Standardized work. Documentation of the precise way in which every step in a process should be completed.
Statement of work (SOW). A detailed set of tasks, expected outcomes, dates, and costs of a project undertaken by an external contractor.
Statistical process control (SPC). A scientific approach to controlling the per-formance of a process by measuring the process outputs and then using statisti-cal tools to determine whether this process is meeting expected performance.
Statistical significance. The differences in two parameters of two data sets are large enough to reject the null hypothesis using hypothesis testing.
Strategy map. A set of initiatives that are graphically linked by if–then state-ments to describe an organization’s strategy.
Supply chain management. The management of all supplier, vendor, and dis-tribution activities related to the production of value to end consumers.
Systems thinking. A view of reality that emphasizes the relationships and inter-actions of each part of the system to all of the other parts.
Taguchi methods. Approaches to quality whereby product development fo-cuses on “perfect” rather than on conformance to specifications.
417Glossary
Takt time. The speed with which customers must be served to satisfy demand for the service.
Theoretical probability. The number of times an event will occur divided by the total number of possible outcomes.
Theory of constraints (TOC). The idea that every organization and process is subject to at least one constraint that limits its movement toward or achieve-ment of its goal.
Throughput time. The time required for an item to complete the entire pro-cess, including waiting time and transport time.
Total quality management (TQM). A management philosophy or program aimed at ensuring quality—defined as customer satisfaction—by focusing on it throughout the organization and for each product or service life cycle.
Toyota Production System (TPS). A quality improvement system developed by Toyota Motor Corporation for its automobile manufacturing lines. TPS has broad applicability beyond auto manufacturing and is now commonly known as Lean manufacturing.
Transformation. The process of converting a variable by linear regression into a format that is more readily usable.
Trend-adjusted exponential smoothing. An extension of a single exponential smoothing model that accounts for a trend when smoothing the data.
Tukey’s rule. An outlier exists in a skewed data set if a data point is greater than Q1 − one step or Q3 + one step, where one step = 1.5 × IQR.
Type I (α) error. The probability of rejecting the null hypothesis when it is true.
Type II (β) error. The probability of accepting the null hypothesis when it is false.
Value proposition. A marketing term summarizing the relative cost, features, and quality of a service or good.
Value purchasing. A system using payment as a means to reward providers who publicly report results and achieve high levels of clinical care. Also known as value-based purchasing.
Value stream map. An overview of how a system transforms supplies into fin-ished goods for the customer.
Variance. A statistical term that indicates how much a measurement varies around the mean.
Work breakdown structure (WBS). A list of the tasks that need to be accom-plished, their relationship to each other, and the resources required for a project to meet its goals.
X-bar chart. Measures process performance of sample means for continuous data.
419419
INDEX
Note: Italicized page locators refer to figures or tables in exhibits.
ABC classification system, 347Accountable care organizations
(ACOs), 378Activity-based costing (ABC): final
aggregation of activity costs per visit, 376; initial data and allocation rate calculation, 375; steps in, 374
Additive property of probability, 180–81, 182
Administrative space, evaluating, 382Advanced-access scheduling, 283,
323, 337–41; for an operating and market advantage, 337; benefits of, 324; fears about, 340–41; going live, 339; heijunka and, 273–74; implementing, 337–39; metrics for evaluating, 339–40
Adverse events, 4Affordable Care Act (ACA), 5, 81,
370; accountable care organiza-tions and, 378; global payments and, 380; healthy lifestyles and, 8; innovation centers and, 125; on mission of PCORI, 53; opera-tional issues with health insurance exchanges and, 97–99; passage of, 3; strategy execution and, 72; sys-tems of care and, 63; value-based purchasing and, 57
Agency for Healthcare Research and Quality (AHRQ), 4, 6, 23, 41, 54; Effective Health Care Program, 47, 380; patient-centered medical home defined by, 51; prevention
quality indicators, 49, 50; public reporting findings of, 54–55
Agile project management, 124, 124–25
Allegheny General Hospital (Pitts-burgh), 135
ALLHAT study (NIH), 203Allina Health (Minnesota), 204American Association of Health Plans,
6American Medical Association, 6, 47American Productivity and Quality
Center, 244American Recovery and Reinvestment
Act (ARRA), 53, 58, 204America’s Health Insurance Plans, 6,
47Analytical tools, 6, 153–61; decision
analysis, 157–61; optimization, 153–57
Analytics, 7–8, 40, 59, 204, 403. See also Data analytics; Healthcare analytics
Analytics department: key purpose of, 212
Analytics technology: sophisticated, 205
Andon, 270, 396, 403Arena simulation software, 297Arrival rate, 288Artificial intelligence, 215Artificial variance, 304–5Assembly lines, 23Assignable (or special) variation, 233
420 Index
Automation tools, 381–82Autoregressive integrated moving
average (ARIMA) models, 351–52Averaging methods, 349–52; autore-
gressive integrated moving aver-age models, 351–52; exponential smoothing, 350; simple moving average, 349–50; trend, seasonal, and cyclical models, 350–51; weighted moving average, 350
Backlog: advanced-access scheduling and, 339
Back orders, supply chain and, 355Bailey-Welch rule, 335Balanced scorecard, 380; balance in,
74–75; construction of targets, 93; customer perspective and market segmentation, 78–80; defined, 73; displaying results, 90; elements of, 76, 399; feedback and strategic learning, 90, 92; financial perspec-tive of, 77–78, 78; four perspec-tives of, 74–75, 75; in healthcare, 75–76; implementation of, 89–90; internal business process perspec-tive, 80–82; learning/growing perspective, 83–84; links, 89; mission/vision and, 77; modifica-tions of, 92–93; perspectives in, 75; project management and, 101; strategic alignment and, 85–86; strategic management systems and, 76; strategy maps and, 75, 82, 86; targets, resources, initiatives, and budgets, 89–90; template sample, 91
Balancing feedback, 10, 11Baldrige Award. See Malcolm Baldrige
National Quality AwardBar coding, 346, 348Bar graphs, 210, 210Baseline plan, 117Batalden, Paul B., 29Bayes’ theorem, 184, 185
Benchmarking, 244, 378, 380, 402Benefis Health System (Montana),
cost reduction case example, 384Best practices: identifying and replicat-
ing, 291–92Big data, 6; analytics and, 7–8; predic-
tive models and analysis of, 205; three Vs of, 40
BJC HeathCare, 38Black belts, Six Sigma infrastructure,
227, 228Blended balanced scorecard–strategy
mapping approach, 59Blitzes. See Kaizen eventsBlock appointment model, 334Bottlenecks, 291, 311Bundled payments, 377–78Burwell, Sylvia Mathews, 57Business intelligence reports, 206Buzan, Tony, 138
Calendaring tools, 381–82Cambridge Health Alliance Whidden
Hospital (Massachusetts): process improvement and patient flow at, 281
Capacity: matching to demand, 290, 323; predicting, advanced-access scheduling and, 338
Capacity of a process, 286–87Capacity utilization, 287; defined,
141, 142; maximizing, 142–43Capture and reporting system, 396Care paths, 270Catalyst for Payment Reform, 57Cause-and-effect diagrams, 140,
146–48, 378; example, 147; pro-cess type, 148; typical categories in, 146
c-charts, 233Center for Medicare & Medicaid
Innovation: Bundled Payments for Care Improvement Initiative, 57
Centers for Disease Control and Pre-vention (CDC), 167, 207
421Index
Centers for Medicare & Medicaid Services (CMS), 4, 36, 54, 63, 98, 369; accountable care organi-zations information, 378; Acute Care Episode Demonstration, 57; Merit-Based Incentive Payment System, 82; regulatory compliance measures and, 204
Centra Health: advanced-access imple-mentation at, 274
Central limit theorem, 185–86, 233Central tendency, measures of, 174–75Cerner Corporation, 204Change control, 117–18, 400Checklist Manifesto, The (Gawande),
270–71Check sheets, 140, 170–71, 304, 401,
407; use in quality management and Six Sigma, 232, 232
Chemotherapy: linkages within health-care system, 12
Chronic care model (CCM), 51Chronic disease management, 50–53;
chronic care model, 51; patient-centered medical homes, 51–53; shared savings model, 379
Clinical decision support systems, 59–61, 60
Clinical microsystems, 8–9Clinical practice guidelines: barriers to
patients’ compliance with, 47–48; evidence-based medicine and, 46–48
Clinical space optimization, 382Clinical systems, 8–9Cloud storage, 204Clustering, 215; Medicare data, 216;
methodology, 215CMS. See Centers for Medicare &
Medicaid ServicesCoefficient of determination (r2), 194,
194Coefficient of variation (CV), 177Cognitive computing systems, 217,
218
Common cause variation, 29, 233Commonwealth Fund, 125Communications plan: scope creation
and, 118Comparative effectiveness research:
infrastructure required for, 7; pri-orities in, 53
Competing on Analytics: The New Sci-ence of Winning (Davenport and Harris), 203
Conditional probability, 182–85Confidence interval (CI), 185–87Conformance quality, 222Consumer-directed healthcare, 8,
394Contingency plans, 289Contingency tables, 183, 183Continuous improvement: kaizen
philosophy of, 259; Six Sigma and mind-set of, 226
Continuous quality improvement (CQI), 34, 256
Control charts, 140, 233, 378, 396Control limits, 233Control systems, 395–96Correlation coefficient (r): defined,
194; problems with, 195–96Cost and revenue models, linking
together, 383–84Cost-effective process improvement:
enabling, 282Cost/importance chart, 382, 383Cost of quality, 223–25; defined, 223;
four parts in, 224Cost reduction: evidence-based medi-
cine and, 49–50Cost-reimbursement contracts, 121Cost-volume-profit (CVP) analysis,
394, 395Council of Supply Chain Management
Professionals, 38Critical path: establishing, 290; slack
and, 115Critical path method (CPM), 26, 27,
98, 115
422 Index
Critical pathway, identifying, 290Critical ratio, 331“Critical to quality” characteristics
(CTQs), 230Crosby, Philip B., 35Cross-functional process maps, 143Crossing the Quality Chasm: A New
Health System for the 21st Century: (IOM), 5, 223
CTQs. See “Critical to quality” charac-teristics (CTQs)
Customer measures, 79Customer perspective: balanced score-
cards and, 74, 75, 78–80; perfor-mance metrics from, 79
Custom patient care, 48, 48–49Cycle time, 263, 288
Dashboards, 212–14; key performance indicators, 213; metrics, 212–13; reports, 214; scorecards, 213
Data: goal of, 205; increase in, 204; in knowledge hierarchy, 20, 21; math-ematical descriptions of, 174–78, 404; visualization techniques, 171–74
Data analysis, 169–99Data analytics, 205–9; descriptive ana-
lytics, 206, 214; predictive analyt-ics, 206–7, 209, 214; prescriptive analytics, 209, 214
Data collection: goal of, 168Data mining: cognitive computing for,
217; for discovery, 214–17Data visualization tools, 209–14; bar
graphs, 210, 210; dashboards, 212–14; histograms, 212; line graphs, 210, 211; map functional-ity, 210–11; scatter plots, 212
Data warehousing and management, 346
Date constraints, slack and, 115Davenport, Thomas, 203Decision analysis, 157–61
Decision-making: analytical tools, 153–61; barriers, 136, 137; bril-liant, ten barriers to, and key ele-ments related to, 137; framework, 136, 136–38; mapping, 138–43; measures of process performance, 141–43; problem identification tools, 143, 145–53
Decision Traps: The Ten Barriers to Brilliant Decision-Making and How to Overcome Them (Russo and Shoemaker), 136
Decision trees, 157, 207, 208; con-struction of, 158; HMO vaccina-tion program, 158, 158–61, 159, 160
Deeming authority, 36Defects per million opportunities
(DPMOs), 225, 226, 238Define-measure-analyze-improve-
control (DMAIC) cycle, 225, 229, 229–32, 250, 276, 397, 401
Delays: feedback and, 11Demand: dependent, 355; indepen-
dent, 355; matching capacity to, 290, 323; predicting, advanced-access scheduling and, 338
Demand forecasting, 349–54; aver-aging methods, 349–52; model development and evaluation, 352; VVH diaper demand forecasting, 352, 353, 354
Deming, W. Edwards, 27, 28–32, 34, 35; adaptation of the 14 points for medical service, 29–31
Deming System of Profound Knowl-edge, 31–32
Departmental activities, prioritized, 382
Dependent demand, 355Deployment champions, Six Sigma
infrastructure, 227, 228–29Descartes, René, 22Descriptive analytics, 206, 214
423Index
Diabetes care: chronic care model and, 51
Diagnosis-related groups (DRGs), 153–56, 154, 156
Discrete event simulation (DES), 297, 298, 299–301
Disease management: predictive mod-els and, 207
Disintermediation, 364Disruptive innovation, 126Division of labor, 22–23DMAIC. See Define-measure-analyze-
improve-control (DMAIC) cycleDonabedian, Avedis, 9, 32–34, 35Dot plots, 173, 173“Drip rate,” in Lean system, 263Duke University Health System, 49Duplicate activities: eliminating, 288
Earliest due date (EDD), 331, 332, 334
Early finish date, 115EBM. See Evidence-based medicineEbola virus, 167–68Economic order quantity (EOQ)
model, 354–58; cost curves, 357, 358; inventory order cycle, 356, 357
EHRs. See Electronic health records80/20 rule, 172Electronic health records (EHRs),
7–8, 9, 204, 291; clinical decision support systems and, 59, 60; popu-lation health and, 205; shared sav-ings model and, 379; text mining of, 215; unintended consequences and, 58
Electronic medication orders, 348Electronic procurement (e-procure-
ment), 364Empirical probability, 178Empiricism, 21Employees: balanced scorecard
and, 75, 75; laying off, 393;
redeployment or retraining of, 392, 393; skills/abilities, 83
Engineering a Learning System (IOM), 221
Enterprise resources planning (ERP), 363
Environment: delivery of care and, 9Epic Systems Corporation, 204Equal variance t-test, 189–90Errors: eliminating, 282–83Evidence-based medicine (EBM), 6–7,
17; barriers to, 47–48; bundled payment models and, 378; care paths and, 270; chronic disease management and, 50–53; com-parative effectiveness research and, 53–54; consistent application of, 46; cost reduction and, 49–50; criticisms of, 48; defined, 6, 45; financial gains from, 49–50; future of, 62–63; guidelines of, 46–48; operational excellence and, 405, 406; standard and custom patient care, 48, 48–49; tools to expand use of, 54–59; wider adoption of, 9
Excellence in healthcare: major areas of expertise related to, 3
Exploration in Quality Assessment and Monitoring, 33
External operational metrics: today and into the future, 82–83
Facility and capital costs, 382–83Failure mode and effects analysis
(FMEA), 230, 245; defined, 149; patient falls, example, 151; steps for, 149–50
Feasibility analysis, 104Feedback: definition of, 10; reinforc-
ing/balancing, systems with, 10, 11
Fee-for-service (FFS), 374, 377; advantage with, 59; problems with, 57
424 Index
Feigenbaum, Armand V., 35Financial accounting: managerial
accounting versus, 393Financial improvement, defined,
371–72Financial management: improvement
of, 369–87; improvement tools, 374, 377; systems approach to, 372–80, 373
Financial perspective: balanced score-cards and, 77–78; performance metrics from, 78
Financial reports, 73, 73Financial stakeholders: balanced score-
card and, 74, 75First come, first served (FCFS), 331,
332Fishbone diagrams, 304, 401, 407;
defined, 146; use in quality man-agement and Six Sigma, 232, 232
5 Million Lives Campaign, 34Five Ss, Lean and, 264–65, 266, 305,
382, 403Five whys technique, 145–46Fixed costs, 394Fixed order quantity with safety stock
(SS) model, 359–61, 362Fixed-price contract, 121Fixed time period with safety stock
(SS) model, 361Flowcharts, 286, 304; creating, steps
for, 140–41; standard symbols, 142; use in quality management and Six Sigma, 232, 232
Force field analysis, 162, 162–63, 163, 401
Ford, Henry, 23Forecasting, 349, 401Formal change mechanism, purpose
of, 117–18Four perspectives, 399Framing, 138F-test, 196Full capitation, 379Futurescan: Healthcare Trends and
Implications 2016–2021, 12
Galileo, Galilei, 22Gantt, Henry, 26Gantt charts, 26, 113, 114Gawande, Atul, 270Gilbreth, Frank, 25–26Gilbreth, Lillian, 25–26Global payments, 380Goal, The (Goldratt and Cox), 150Graphic display of data, 404Graphic tools, 169–74; check sheets,
170–71; dot plots, 173, 173; his-tograms, 171–72; mapping, 170; Pareto diagrams, 172–73; scatter plots, 173–74, 174
Graphs, 210, 210Green belts, Six Sigma infrastructure,
227, 228Gross domestic product (GDP):
health spending projections and, 4Group Health Cooperative (Seattle),
379
Hadoop software, database system, 40, 215
Harris, F. W., 354Harris, Jeanne, 203“Harvesting the low-hanging fruit,”
288Hawthorne studies, 32Healthcare: balanced scorecard in,
75–76Healthcare, systems view of, 8–11, 9;
clinical system, 8–9; reinforcing and balancing feedback in, 10; sys-tem stability and change, 10–11
Healthcare analytics, 203–18. See also Data analytics; data mining for dis-covery, 214–17; data visualization, 209–14; defining, 203–4
Healthcare Benchmarks and Quality Improvement, 244
Healthcare Finance: An Introduction to Accounting and Financial Man-agement (Gapenski and Reiter), 391
HealthCare.gov, 98, 99
425Index
Health Care Homes initiative (Minne-sota), 52–53
Healthcare leaders: complex world of, 72, 73
Healthcare organizations: of the future, 407, 408; strategy execu-tion and, 72
Healthcare Quality Book: Vision, Strat-egy and Tools (Joshi et al.), 4
Healthcare savings accounts (HSAs), 8, 394
Healthcare spending: growth projec-tions for, 4
Health Catalyst (Minnesota), 204Health insurance exchanges, 78HealthPartners (Minnesota): Six
Sigma Clostridium difficile study, 221–22
Heijunka, 273–74, 289, 403Hennepin County Medical Center
(HCMC), 207High-Tech Digital Imaging (HTDI):
actual versus trend in utilization of, 61; benefits with, 61
Histograms, 140, 171, 171–72, 175, 212, 304, 402; defined, 171; use in quality management and Six Sigma, 232, 232
Holding (carrying) costs, 355Holding the gains, approaches to,
391–409; control system, 395–96, 404; human resources planning, 391–93, 404; managerial account-ing, 393–95, 404; tools to use: general algorithm, 397–404
Homeostasis, 11Hospital census: rough-cut capacity
planning and, 324–26Hospital Compare, 54Hospital financial model, 385Hospitals: bundled payments in,
examples, 377Hoteling, 382Human resources (HR) planning,
391–93; ongoing and comprehen-sive, 393; process for, 392
Human Resources in Healthcare: Man-aging for Success (Fried and Fot-tler), 391
Hypothesis testing, 187–92
IBM Watson Analytics, 217; descrip-tion of, 217; opening page screen-shot, 218
Idle time, 288If Japan Can . . . Why Can’t We?, 28Income statement: financial health
indicators on, 371–72Independent demand, 355Individual appointment model, 334Industrial Revolution, 22Informatics systems: maturing of, 6Information: in knowledge hierarchy,
20, 21Information feedback: embedding,
290Information technology (IT): neces-
sary, 84; patient flow and investing in, 284
Innovation centers, 59, 125, 125–26Innovation process, 81Institute for Clinical Systems Improve-
ment (ICSI), 60Institute for Healthcare Improvement
(IHI), 41, 274; Triple Aim, 223Institute of Medicine (IOM), 7, 28;
clinical practice guidelines defined by, 47; Crossing the Quality Chasm, 5, 223; Engineering a Learning System, 221; To Err Is Human, 21
Integrated patient care, 48, 49Intermountain Healthcare (IHC), 49Internal business process perspective,
80–82International normalized ratio (INR),
255International Organization for Stan-
dardization (ISO), 35Internet, 291Internet of Things, 6Inventory, 288; classification systems,
347–48; defined, 347; theory of
426 Index
constraints and, 152; tracking sys-tems, 348–49
Ishikawa, Kaoru, 35, 232ISO. See International Organization
for StandardizationISO 9000, 35–36, 38, 225
Jidoka, 270, 396, 403Job loss, 393Job/operational scheduling, 330–34Joint Commission, The, 4, 149Juran, Joseph M., 27, 32, 34, 35, 167;
quality trilogy, 32, 33Juran’s Quality Handbook, 32Just-in-time (JIT), 37. See also Lean;
inventory systems, 362–63; pro-duction, 256
Kaizen, 83, 259, 276, 402–3Kaizen events, 83, 265, 267–69, 303,
305, 403Kanban, 37, 271, 271–72, 272, 362–
63, 403Kant, Immanuel, 21, 22Kaplan, Robert, 59, 83Key performance indicators: dash-
board visualizations and, 212, 213Key process indicators (KPIs), 167Knowledge-based management
(KBM), 20–21Knowledge hierarchy, 20, 20–21
Labor shortages, widespread, 393Lagging indicators, 85Late finish date, 115Layoffs, 393Leadership, 3; Six Sigma, 226–27;
skills, 128Leading indicators, 86Lead time, 355Lean, 25, 37–38, 284, 304, 379, 380,
400; andon, 270, 403; cycle time, 263; development of, 256; five Ss, 264–65, 266, 305, 382, 403; heijunka, 273–74, 403; human
resources planning and, 392; jidoka, 270, 403; kaizen, 83, 259, 276, 402–3; kaizen event or blitz, 265, 267–69, 305, 403; kanban, 271, 271–72, 272, 362–63, 403; merging of Six Sigma programs and, 274–76, 275; muda, 257; operational excellence and, 405; overview, 255; philosophy of, 257; process improvement and, 305; rapid changeover, 272–73; report evaluation and, 381; service cost optimization and, 377; standardized work, 269–70, 305, 403; successful SCM initiatives and, 365; takt time, 261, 263, 264, 305, 403; through-put time, 263, 264, 403; tools, 257; value stream mapping, 259–61, 305, 403; waste, types of, 257–58
Lean Production House, 256, 257Leapfrog Group, 23Learning/growing perspective, 83Length of stay (LOS): IT investments
and impact on, 284Lewin, Kurt, 162Linear optimization problems, 153Linear programming, 327, 400; exam-
ple, 153–56Linear regression, 351. See also Simple
linear regressionLine graphs, 210, 211Little’s law, 294–95, 338Litvak, Eugene, 283Load balancing (or load leveling), 289Localizing Care to High-Volume Cen-
ters (AHRQ), 23Locke, John, 21, 22
MacColl Center for Health Care Innovation, 51
Machine That Changed the World, The (Womack, Jones, and Roos), 38, 256
Malcolm Baldrige National Quality Award, 36–37, 38, 71–72; criteria,
427Index
successful SCM initiatives and, 225, 365
Management: traditional theory of, 73, 73
Management tools: failure of, reasons for, 73–74
Managerial accounting, 393–95; CVP analysis, 394, 395; financial accounting versus, 393; steps in, 394–95
Maps and mapping, 170, 170, 399; functionality, 210–11, 211; value stream, 259–61
Margin, financial, 371Massachusetts General Hospital: care
path for CABG surgery, 270Mass production, 22–23Master black belts, Six Sigma infra-
structure, 227, 228Master production scheduling (MPS),
330Material requirements planning
(MRP), 363, 363Mathematical descriptions of data,
174–78, 404McDonald, Bob, 75Mean (average), 174–75Mean absolute deviation (MAD), 176,
352, 354Mean square error (MSE), 196, 352,
354Mean square regression (MSR), 196Measures of central tendency, 174–75Measures of variability, 176–78Median (average), 175Medicaid: creation of, 45Medical home. See also Patient-
centered medical home (PCMH): shared savings model and, 379
Medicare: breakeven efforts, Benefis Health System and, 384, 386; bundled payments, 377; creation of, 45; Hospital VBP program, 57; making ends meet on, 370–71; PGP Demonstration Project, 57;
Shared Saving Program, 57; value purchasing, 57
Medicare Payment Advisory Commis-sion (MedPAC), 63, 369–70
Medicare prospective payment, 377
Medication orders, electronic, 348Meetings, 381Merit-Based Incentive Payment Sys-
tem (MIPS), 82–83Metformin, 62Metrics, 167–68, 399; dashboard visu-
alizations and, 212–13; for evaluat-ing advanced access, 339–40
Microsoft Project software, 27, 108, 115, 117
Milestones, 110Mind mapping, 138, 139, 170, 170,
285, 399Minnesota State Fair: text mining at,
215–16, 217Mitigation plan, 120Mixed block-individual model,
334–35Mobile applications: primary care and,
6Mode (average), 175Model Hospital Statistical Form, 17Modularized patient care, 48, 49Monitoring and response plan, 396Motorola, 225Muda (waste), 257Multicare Health System: outcome
improvements in pneumonia care at, 45–46
Multiplicative property of probability, 180
Narrow networks: pressure of, 370–71
National Academy of Sciences, 7National Committee for Quality
Assurance, 4National Guideline Clearinghouse, 6,
47
428 Index
National Institutes of Health (NIH), 54, 203
National Quality Forum, 4, 55Network diagrams, 113, 113Neural networks, 207, 209New Economics for Industry, Govern-
ment, Education (Deming), 31Nightingale, Florence, 17–18Nonlinear optimization problems, 153Non-value-added activities: eliminat-
ing, 288Non-value-added time, 288Norton, David, 59, 83No-shows: reducing, 282Null hypothesis, 188Number of defects or errors, 288
Observed probability, 178–79Ohno, Taiichi, 37, 256, 258100,000 Lives Campaign, 34Operating expenses: theory of con-
straints and, 152Operational excellence: scale for,
405–6Operational perspective: performance
metrics from, 82Operational reports, 73, 73Operations: balanced scorecard and,
75, 75; complex healthcare delivery systems and, 3
Operations improvement tools: cost reduction with, 377
Operations management, 22; defined, 18; effective, framework for, 13; theory of constraints and, 152; value purchasing and, 59
Operations research, 26Optimal Outpatient Scheduling tool,
335Optimization, 153Optum Labs: diabetes example, 62Ordering (setup) costs, 355Organizational infrastructure: delivery
of care and, 9Organizational performance indica-
tors, 371
Outcome indicators, 85–86Outliers, 177–78Out of the Crisis (Deming), 29Outsourcing, 364Overhead costs, 394Overhead expenses, 380–82; consoli-
dated activities, 381; departmental activities, 382; facility and capital costs, 382; meetings, reports, and automation tools, 381–82; process improvement, 380; reduction in, 372; staffing layers, 381
Parallel processing, 289Pareto, Vilfredo, 347Pareto charts and diagrams, 140, 172,
172–73, 248, 304, 378, 402, 407; defined, 172; use in quality man-agement and Six Sigma, 232, 232
Pareto principle, 32, 172, 347Park Nicollet (Minnesota): Lean tools
and anticoagulant delivery system at, 255–56
Patient appointment scheduling mod-els, 334–35
Patient behavior models, 59Patient care microsystem: elements
of, 8Patient-centered medical home
(PCMH), 77; defined, 51; func-tions and attributes of, 51–53
Patient-Centered Outcomes Research Institute (PCORI): chronic disease management and, 54; mission of, 53
Patient flow, 282; improving, manage-ment solutions for, 283–84; IT investments and, 284; poor, causes of, 283
Patient Protection and Affordable Care Act. See Affordable Care Act (ACA)
Patient self-service, 291Patients-in-process, 288Pay for performance (P4P), 24, 57,
393; issues in, 55; methods of, 55;
429Index
Vincent Valley Hospital and Health System and, 63–64
Payment reform, 55, 56p-charts, 233PDCA. See plan-do-check-actPercent value added: Lean initiatives
and, 260–61, 264Per diem, 377Performance drivers, 85–86Performance improvement: important
events in, 19; philosophies, 34–38Performance metrics, 63Performance quality, 222Physician compensation: value pur-
chasing and, 63Physician Group Practice Demonstra-
tion, 57PinnacleHealth (Pennsylvania),
348–49Plan-do-check-act (PDCA), 28, 29,
35, 137, 229, 267, 281–82Point-of-use data entry and retrieval,
346Point-of-use systems, 348Poka-yoke, 245, 304, 402Population health, 3, 205; account-
ability and management, 6; predic-tive models and, 207
Post-sale service, 81–82Poudre Valley Health System (PVHS),
71–72Practical significance, 191–92Predictive analytics, 205, 206–7, 209,
214Predictive tools: decision trees, 207,
208; neural networks, 207, 209; regressions, 207
Prescriptive analytics, 209, 214Prevention quality indicators (PQIs),
49, 50Primary care: redesign of, 6Principles of Scientific Management
(Taylor), 22, 23–24Probability, 178–85; additive prop-
erty of, 180–81, 182; bounds on, 179–80; conditional, 182–85;
determination of, 178–79; multi-plicative property of, 180; proper-ties of, 179–85
Problem identification tools, 143, 145–53; cause-and-effect diagram, 146, 146–48, 147, 148; failure mode and effects analysis, 149–50; five whys technique, 145–46; root-cause analysis, 143–45; theory of constraints, 150–53
Problem types, 282–83Process capability, 402; common
measures of, 238; Six Sigma limits, 239; Six Sigma quality and, 238
Processes: describing, 285Process improvement, 81, 137, 281;
Lean, 305; overhead expenses and, 380; in practice, 304–18; Six Sigma, 304–5; VVH emergency department project, 305–18, 308, 310, 311, 313, 315, 316, 317, 318
Process improvement, approaches to, 284–92; overview, 284–85; problem definition, 285; process mapping, 285–86, 287; process measurements, 286–88; tools for process improvement, 288–92
Process improvement tools, 288–92; alternative process flow paths and contingency plans, 289; combine related activities, 289; critical path establishment, 290; eliminate duplicate activities, 288; eliminate non-value-added activities, 288; embedding information feedback and real-time control, 290; ensuring quality at the source, 290; identify-ing best practices, 291–92; letting patient do the work, 291; load bal-ancing, 289; matching capacity to demand, 290; parallel processing, 289; technology use, 291; theory of constraints application, 291
Process maps and mapping, 139–41, 140, 285–86, 399, 407; creating, steps for, 140–41; cross-functional,
430 Index
143; defined, 139; service blue-printing, 143, 145; steps in, 286; VVH emergency department example, 286, 287
Process measurements, 286–88Process owner: identifying, 285Process performance measures,
141–43Process-type cause-and-effect dia-
grams, 148, 148Procurement system: contracting,
121–22; selecting a vendor, 122; streamlining processes, 364
Product quality: eight dimensions of, 222
Program evaluation and review tech-nique (PERT), 26, 27, 98, 110, 115
Project charter, 100, 102–5, 108, 400; document elements, 105; fac-tors constraining execution of, 102
Project leadership: skills needed for, 128
Project management, 26–27, 400; agile, 124, 124–25; complete pro-cess of, 101; matrix, 102; overview, 97–98; tools, 107–8; when to use, 100
Project Management Book of Knowl-edge, 98
Project Management Institute (PMI), 98, 100
Project management office (PMO), 122–23
Project management software, 107–8Project manager, 103, 126, 127Project plan, 100Project(s): closure, 123; contracting,
121–22; control, 117–20; crash-ing, 116; definition of, 99–100; failures, 103; feasibility analysis, 104; with increased performance requirement and shortened schedule, 103; procurement sys-tem, 121–22; quality manage-ment, 120–21; risk management,
118–20; scheduling, 113–16; selection, 100–101; stakeholders, 103–4; team, 126–28; well- managed, 99–100
Project scope: document, 108–9; mathematic expression of, 102; relationship to performance, level, time, and cost, 103; statement, 100
Proportions, 190–91Public health initiatives: text mining
applied to, 215–16, 217Public reporting, 54, 57p-value of statistical significance test,
190
Quality: cost of, 223–25; defining, 222–23; introduction to, 27–34; at the source, ensuring, 290
Quality Assurance Project, 223Quality bonuses or penalties, 379–80Quality circles, 37Quality function deployment (QFD),
81, 240–43, 304, 402; defined, 240; house of quality, 240, 241; Riverview Clinic diabetes patients and, 242–43, 243, 244
Quality improvement: slow pace of, 4–5
Quality management, 120–21Quality measures: criticism of, 58Quality tools: additional, in process
improvement, 240–45; fundamen-tal, 140
Quality trilogy (Juran), 33Queue discipline, 293Queuing priority, 331Queuing system: simple, 292Queuing theory, 292–304; defined,
292; discrete event simulation and, 297; notation, 293; solutions, 293–95
Radio-frequency identification (RFID), 346, 348–49
Range, calculating, 176
431Index
Range (r) chart, 233Rapid changeover, 272–73Rapid process improvement workshop
(kaizen event), 83RASIC (responsible, approval, sup-
port, informed, and consult), 112, 112, 117
Rationalism, 21, 22Real-time control: embedding, 290Regression, 192, 207Regression analysis, 378, 405Regulatory environment: analytics
and, 204Reinforcing feedback, 10, 11Related activities: combining, 289Relative frequency, 178–79Reports, 214; evaluating, 381Request for information (RFI), 122Request for proposal (RFP), 122Resource leveling, 114Revenue, 371; expenses directly
related to, 372, 374; improving, 382–83
Revenue cycle, optimized, 383Risk adjustment, 54Risk management, 118–20, 400Risk mitigation plan, 120, 120Risk priority number (RPN), 149, 150Risk register, 120Riverview Clinic (VVH). See also
Vincent Valley Hospital and Health System (VVH): appoint-ment schedule, 335, 336; clinic timing issues and Lean, 263–64, 264; high-level process maps, 140; patient check-in process map, 141; process capability, 238–39; quality function deployment at, 242–43, 243, 244; Six Sigma generic drug project, 245–48, 246, 247, 248, 249; statistical process control, 233, 234, 235, 235–37, 237; urgent care staffing at, 326–30, 327, 328, 329
Robots, 6
Rolled throughput yield (RTY), 239–40, 240
Root-cause analysis (RCA), 143, 145, 149, 230, 285, 399
Rough-cut capacity planning: defined, 325
Run charts, 140, 232, 232, 304, 378, 396, 396, 402
Safety stock (SS) model: fixed order quantity with, 359–61; service level and, 360, 360; variable demand inventory order cycle with, 359
Scatter plots, 140, 173–74, 174, 212, 232, 232, 304, 378, 402, 407
Schedules and scheduling, 400; advanced-access, 337–41; com-pression of, 116; job/operational, 330–34; patient appointment models, 334–35; projects, 113–16; staff, 326–30
Scientific management, history of, 22–26
Scope creep, 103Scorecards, 213Second Street Family Practice
(Maine): scheduling management, 323
Seiketsu (standardize), 265Seiri (sort), 265Seiso (shine), 265Seiton (set in order), 265Senge, Peter, 10Sensitivity analysis, 156–57, 157Sequencing rules, 331Service blueprint, 143, 145Service level, 359Service lines: growing, 372Service quality: five dimensions of, 223Service time, 288Setup time, 288Shared savings model, 378–79Shewhart, Walter, 27–28, 32, 35Shewhart’s rule, 177Shingo, Shigeo, 37, 272
432 Index
Shitsuke (sustain), 265Shortage costs, 355Shortest processing time (SPT), 331,
332Shouldice Hospital (Toronto), 23Simple linear regression, 192–98;
assumptions of, 197; coefficients and, 194–96; defined, 192; inter-pretation of, 193–94; statistical measures of model fit, 196–97; transformations, 197–98
Simple moving average (SMA), 349–50
Simul8 simulation software, 297Simulation, 400; appointment sched-
uling models and rules, 335; dis-crete event simulation, 297–304; model development, 302; model validation, 302; output analysis, 301; queuing theory, 292–97
Single exponential smoothing (SES), 350
Single-minute exchange of die (SMED), 272, 273, 403
Six Sigma, 25, 35, 38, 225–40, 256, 284, 304, 379, 380, 383, 400; Clostridium difficile study, 221–22; culture, 226; defects per million opportunities (DPMOs), 225, 226; define-measure-analyze-improve-control (DMAIC) cycle, 225, 229, 229–32, 250, 276; development of, 225; fundamental philosophi-cal tenet of, 229; human resources planning and, 392; infrastructure, 227; leadership, 226–27; merg-ing of Lean and, 274–76, 275; operational excellence and, 405; organizational infrastructure and training, 227–29; primary function of, 304; process capability and, 238–39, 239; process improve-ment and, 304–5; process metrics, 230–31, 231; program themes, 225–26; quality tools, 232, 232,
304, 401–2, 407; Riverview Clinic generic drug project, 245–49, 246, 247, 248, 249; rolled throughput yield (RTY), 239–40; service cost optimization and, 377; shared sav-ings model and, 378–79; statistical process control (SPC), 233–38; strategy and measurement, 226; successful SCM initiatives and, 365
Slack, 115Slack time remaining (STR), 331Smith, Adam, 22, 23Social media, 382Solver, 384, 386Spaghetti diagrams, 265, 267, 305,
403Special cause variation, 28–29Specialization, 22–23Staffing layers, 381Stakeholders, 103–4Standard deviation, 177Standardized work, 269–70, 305, 403Standard patient care, 48, 48–49Statement of work (SOW), 121–22Statistical process control (SPC), 28,
233–37, 402; description of, 233; Riverview Clinic (VVH) vignette, 233, 234, 235, 235–37, 237
Statistical significance, 191–92, 192Statistical tests, 405Statistical thinking, 167, 168Stockouts, 355Storage space: minimizing, 382Strategic management systems: bal-
anced scorecards and, 76Strategic plans, 73, 73Strategic view, supply chain, 364–65Strategy execution: challenge of,
72–73Strategy maps, 75, 82, 86–89, 87, 88,
92Strengths, weaknesses, opportunities,
and threats (SWOT) analysis, 119Sum of squares error (SSE), 196Sum of squares regression (SSR), 196
433Index
Supply chain management (SCM), 38–39, 380, 400, 404; defined, 346; demand forecasting, 349–54; importance of, in healthcare, 345; inventory systems, 362–63; inventory tracking, 347–49; order amount and timing, 354–62; overhead expenses, 380–82; pro-curement and vendor relationship management, 364; strategic view, 364–65; supply chains, 6
Swim lane process map, 143, 144SWOT. See Strengths, weaknesses,
opportunities, and threats (SWOT) analysis
Systems improvement, 281Systems thinking, 39Systems view, of provision of services,
20
Tactical plan, 73Taguchi, Genichi, 243Taguchi methods, 240, 243–44Takt time, 261, 263, 264, 305, 403Taylor, Frederick, 22, 23–25, 26Teams: meetings and, 127–28; quality
bonuses or penalties and, 379–80; structure and authority, 127
Technology: analytics, 205Telemedicine, 6Texting, 382Text mining at the state fair (case
example), 215–16, 217Theoretical probability, 179Theory of constraints (TOC),
150–53, 284, 295; applying, 291; defined, 150; operations manage-ment and, 152; steps for, 150–51, 400–401
Theory of swift and even flow (TSEF), 282
Things-in-process, 288Throughput: theory of constraints
and, 152Throughput rate, 287
Throughput time, 141, 263, 264, 287, 295, 310, 403
Time-and-materials contract, 121Time fences, 330Time series analysis, 349Time series forecasting, 351To Err Is Human (IOM), 21, 28Tool selection, for improvement:
general algorithm, 397–404, 398; analytics, 403; balanced scorecard for strategic issues, 399; basic performance improvement tools, 400–401; holding the gains, 404; issue formulation, 397, 399; Lean, 402–3; project management, 400; quality and Six Sigma, 401–2; strategic or operational issue, 399; supply chain management, 404
Total quality management (TQM), 34, 35, 256
Toyoda, Sakichi, 270Toyota Group, 270Toyota Production System (TPS), 37,
135, 256, 258Transformations, 197–98Tree diagrams, 147–48; additive prop-
erty of probability, 182; Bayes’ theorem example, 185; ED wait time, 184; multiplicative property of probability, 180, 181
Trend-adjusted exponential smooth-ing technique (Holt), 350–51
Trinity Health (Michigan): supply chain management techniques at, 345–46
Triple Aim (IHI), 223t-test, 190, 196–97Tukey’s rule, 178Two-bin system, 348, 362Type I (α) error: clinic wait time
example, 188–89, 189; court sys-tem example, 188, 188; defined, 188
Type II (β) error: clinic wait time example, 188–89, 189; court
434 Index
system example, 188, 188; defined, 188
Understanding: in knowledge hierar-chy, 20, 21
UnitedHealth Center for Health Reform & Modernization, 57, 58
United States: opportunities for health system in, 6–8; six aims for health system in, 5, 5; systemic waste and healthcare in, 223–24
US Department of Health and Human Services (HHS), 6, 54
US Navy, 34
Value-added time, 288Value proposition: customers and,
79–80; defined, 79; Vincent Valley Hospital and Health System and, 80
Value purchasing (or value-based pur-chasing), 54, 57–59, 82; implica-tions for operations management, 59; Medicare and, 57; physician compensation and, 63; policy issues in, 58
Values-based standardization, 364Value stream mapping, 259–61, 305,
311, 403Variability, measures of, 176–78Variable costs, 394Variance, 176; artificial, 304–5; ubiq-
uity of, 167Variation: reducing, 282, 401Vendor relationship management, 364Venetian Arsenal, 23Venn diagram: multiplicative property
of probability, 180, 181Veterans Health Administration (Min-
nesota): sample 5S form, 266Vidant Health (North Carolina): Flex-
work portal, 369–70Vincent Valley Hospital and Health
System (VVH), 14. See also Riv-erview Clinic (VVH); ambulatory
care network growth, 409; bal-anced scorecard, 77; birthing center strategy map, 87; cause-and-effect diagram, 147, 148; census for, 324, 325, 326; CVP analysis of outpa-tient services at, 394, 395; diaper demand forecasting example, 352, 353, 354, 355; diaper order quan-tity example, 358–59, 361, 362; discrete event simulation software example, 297, 298, 299–304, 300, 301, 302, 303; emergency depart-ment strategy map, 88; force field analysis, 162–63, 163; improvement projects and associated training, 85; internal business processes, 83; kaizen event, 83, 268–69; labora-tory sequencing rules, 332, 332, 333, 334, 334; learning/growing perspective, 83; linear programming example, 153–56, 154, 156; mission and vision of, 77; operational excel-lence and, 406–7; pay for perfor-mance (P4P) and, 63–64; process improvement project: emergency department, 305–18, 308, 310, 311, 313, 315, 316, 317, 318; process mapping emergency department example, 285–86; project charter, 105, 106–7; queuing theory, 295–97; simulation, 297–304; strategy maps, 86–89; value proposition, 80; value stream mapping and birthing center at, 261, 262
Virginia Mason Medical Center: Patient Safety Alert System at, 271
Voice of the customer (VOC), 240Vorlicky, Loren, 29
Wagner, Edward, 51Waiting line theory. See Queuing
theoryWait time, 288Warehouse management, 349Warfarin, 255
435Index
Waste: Lean and types of, 257–58Web conferences, 381Weighted moving average (WMA),
350, 354Wellness, healthy lifestyle and, 8Winter’s triple exponential smoothed
model, 351Wisdom: in knowledge hierarchy, 20, 21Work-at-home policies, 382Work breakdown structure (WBS),
109–12, 111, 119, 400; defined, 109; general format for, 109
Work-in-process, 288Workloads: balancing, 289World Health Organization (WHO),
167, 168, 255
X-bar chart, 233, 236
Yellow belts, Six Sigma infrastructure, 227, 227
Zika virus, 167
437
ABOUT THE AUTHORS
Daniel B. McLaughlin is director of the Center for Health and Medical Affairs in the Opus College of Business at the University of St. Thomas, Minneapolis, Minnesota. He is active in teaching, research, and speaking at the university, with a special emphasis on healthcare operations and policy.
From 1984 to 1992, Mr. McLaughlin was administrator and CEO of Hennepin County Medical Center, the level I trauma center in Minneapo-lis. He was chair of the National Association of Public Hospitals and Health Systems and served on President Bill Clinton’s Task Force on Health Care Reform in 1993. In 2000, he helped establish and direct the National Institute of Health Policy at St. Thomas. He is the author of a number of textbooks and management guides published by Health Administration Press, including Make It Happen: Effective Execution in Healthcare Leadership and The Guide to Healthcare Reform: Readings and Commentary.
Mr. McLaughlin holds degrees in electrical engineering and healthcare administration from the University of Minnesota.
John R. Olson, PhD, is the research director for the Center of Innovation in the Business of Healthcare and program director for the business analytics program at the University of St. Thomas. He holds a doctorate in operations and supply chain management from the University of Nebraska and is a mas-ter black belt in Six Sigma and a Lean sensei. Over the past 10 years, he has consulted with several healthcare organizations to implement their continuous improvement programs, including Six Sigma and Lean initiatives.
Dr. Olson has published many articles and books in leading operations management journals and has consulted with numerous Fortune 500 companies as well as many firms in the public sector.
- Brief Contents
- Detailed Contents
- Preface
- Part I Introduction to Healthcare Operations
- Chapter 1 The Challenge and the Opportunity
- Chapter 2 History of Performance Improvement
- Chapter 3 Evidence-Based Medicine and Value-Based Purchasing
- Part II Setting Goals and Executing Strategy
- Chapter 4 Strategy and the Balanced Scorecard
- Chapter 5 Project Management
- Part III Performance Improvement Tools, Techniques, and Programs
- Chapter 6 Tools for Problem Solving and Decision Making
- Chapter 7 Statistical Thinking and Statistical Problem Solving
- Chapter 8 Healthcare Analytics
- Chapter 9 Quality Management – Focus on Six Sigma
- Chapter 10 The Lean Enterprise
- Part IV Applications to Contemporary Healthcare Operations Issues
- Chapter 11 Process Improvement and Patient Flow
- Chapter 12 Scheduling and Capacity Management
- Chapter 13 Supply Chain Management
- Chapter 14 Improving Financial Performance with Operations Management
- Part V Putting it all Together for Operational Excellence
- Chapter 15 Holding the Gains
- Glossary
- Index
- About the Authors
