Introduction to Healthcare Informatics, Second Edition
Chapter 5:
Data and Information
© 2017 American Health Information Management Association
© 2017 American Health Information Management Association
Objectives
Describe the types of data, as well as the relationship between data and information
Practice using health information standards
Create a data map for a given scenario
Compare and contrast the methods of data collection
Demonstrate the use of data quality practices
Design appropriate data presentations
Support and evaluate EHRs, HIEs, and RECs
© 2017 American Health Information Management Association
Data Basics
Singular or plural data
Numerical or alphabetical
Categorical or discrete
Nominal or ordinal
Continuous
Ratio or interval
© 2017 American Health Information Management Association
Questions to Ask about Data or Information
For which purpose or use was the data originally collected?
Are there any data standards related to the data?
How was the data collected?
What is the quality of the data?
How will the data be analyzed to produce information?
How can the data or information be presented so that it is meaningful to users?
© 2017 American Health Information Management Association
Data Format
Raw data
unformatted combination of text, symbols, and words
E.g. “Green, 25”
Formatted database field
Last name = Green
Age in Years = 25
© 2017 American Health Information Management Association
Data Standard Development
Ad hoc standards
De facto standards
Government mandate
Consensus standards
LOINC
UHDDS
Microsoft Windows
ICD
© 2017 American Health Information Management Association
Caitlin Wilson (CW) – Not sure what the examples on the right are supposed to line up with. Can these be removed or put in a different location?
ISO Standards Development Process
Proposal
Preparatory
Committee
Enquiry
Approval
Publication
© 2017 American Health Information Management Association
US Health Information Technology Standards
© 2017 American Health Information Management Association
Health IT Policy Committee (sets overall health IT policy)
Health IT Standards Committee (recommends standards to meet health IT policy)
Health Insurance Portability and Accountability Act (informs policy and standards)
Data Set Standards
Vital statistics
London Bills of Mortality
Immunization registries
Health Plan Employer Data and Information Set
Continuity of Care Document (CCD) (ASTM Standard 2005)
© 2017 American Health Information Management Association
Classification, Code Set, and Terminology Standards
Classification
Code set
Terminology
Concepts
© 2017 American Health Information Management Association
Data Mapping
Definition
Creation of a map
Purpose
Source
Target
Forward maps
Reverse maps
Relationship
Equivalence
© 2017 American Health Information Management Association
Transaction Standards
Data-interchange standards
Transaction set
Types of transaction standards
Accredited Standards Committee X12 5010
National Committee on Prescription Drug Programs
Health Level 7
© 2017 American Health Information Management Association
Electronic Health Record Standards and Certification
Certification Commission on Health Information Technology
Health Information Technology for Economic and Clinical Health Act mandated
ONC-Authorized Testing and Certification Bodies
National Institute for Standards and Technology
© 2017 American Health Information Management Association
Data Collection Considerations
Why is the organization entering these data?
What data need to be entered?
Who will enter the data?
What will be the most efficient method of data entry for each type of user?
Under what circumstances will users enter data?
What use does the organization expect to make of the data?
© 2017 American Health Information Management Association
Data Entry
Structured data entry
Easily computable
Controlled values
Inflexible
Potential loss of meaning
Unstructured data entry
Expressivity
Post-hoc text processing
Natural language processing
Narrative-text string
Concept identification
Negation and temporal issues
© 2017 American Health Information Management Association
Data Measurement
Nominal data
Ordinal data
Interval data
Ratio data
© 2017 American Health Information Management Association
Management and Governance
Data management
Data governance
Discover
Design
Enable
Maintain
Archive or retire
© 2017 American Health Information Management Association
Data Quality
Data Quality Management Model
Application
Collection
Warehousing
Analysis
© 2017 American Health Information Management Association
Data Quality Characteristics
Accuracy
Accessibility
Comprehensiveness
Consistency
Currency
Definition
Granularity
Precision
Relevance
Timeliness
© 2017 American Health Information Management Association
| DATA QUALITY CHARACTERISTIC AND DEFINITION | HEALTHCARE EXAMPLE |
| Accuracy—The extent to which the data are free of identifiable errors | The data element gender is completed for all patients and a random check of 500 records performed annually revealed only one demographic data element in conflict with the documentation. |
| Accessibility—Data items that are easily obtainable and legal to access with strong protections and controls built into the process | All persons with access to the EHR have the ability to search the master patient index and the search function is designed such that the wrong patient is rarely accessed. |
| Comprehensiveness—All required data items are included; ensures that the entire scope of the data is collected with intentional limitations documented | Providers may decide that recording external cause data is not useful. If so, their data dictionary for the diagnoses data elements would need to include this as an intentional limitation. |
| Consistency—The extent to which the healthcare data are reliable and the same across applications | Within the EHR data such as allergies must be consistently displayed within different applications or screens to prevent confusion. |
| Currency—The extent to which data are up-to-date; a datum value is up-to-date if it is current for a specific point in time, and it is outdated if it was current at a preceding time but incorrect at a later time | Patient age is generally current when the care is delivered. If age is not reentered the software needs to be have functionality to automatically update the age when appropriate. |
Data Quality Characteristics
© 2017 American Health Information Management Association
| DATA QUALITY CHARACTERISTIC AND DEFINITION | HEALTHCARE EXAMPLE |
| Definition—The specific meaning of a healthcare-related data element | Address as a data element label can mean the street address or it can mean the entire address to include city, state, and zip code. |
| Granularity—The level of detail at which the attributes and values of healthcare data are defined | Adult weights are usually only recorded in pounds, possibly tenths of a pound. Newborn weights must be recorded in terms of ounces for accuracy. |
| Precision—Data values should be strictly stated to support the purpose | Diagnosis Related Group values are carried out to four digits behind the decimal. It would be inaccurate to have the system only use two digits behind the decimal. |
| Relevance—The extent to which healthcare-related data are useful for the purposes for which they were collected | Recording a primary diagnosis for hospital inpatients would be irrelevant because coding guidelines mandate collection of the principal diagnosis, which can be entirely different. |
| Timeliness—Concept of data quality that involves whether the data is up-to-date and available within a useful time frame; timeliness is determined by the manner and context in which the data are being used | It would be inappropriate for blood pressure readings from intensive-care unit (ICU) monitors to only be updated each hour. |
Data Quality Characteristics
© 2017 American Health Information Management Association
Data Quality Assessment and Management Process
Who are the data consumers?
What are the needs of the data consumers?
What are the required features and quality characteristics?
How well do our current information products meet the needs and requirements?
Where are the gaps and how important are they?
© 2017 American Health Information Management Association
Data Consumer Needs Assessment Tool
© 2017 American Health Information Management Association
Data Analysis
Understanding the data
Data dictionary
Methods of collection
Cleaning the data
Identify errors
Descriptive statistics
Categorical data
Use of crosstabs
Determine correct values or impute
If uncorrectable delete the record
© 2017 American Health Information Management Association
Data Cleaning—Continuous Data
© 2017 American Health Information Management Association
Data Cleaning—Categorical Data
© 2017 American Health Information Management Association
Data Cleaning—Crosstab
© 2017 American Health Information Management Association
Data Analysis
Analyzing the data
Goals and objectives
Level of analysis of the study
Limitations of the data
Tools available
Analysis needs
Use of the results
© 2017 American Health Information Management Association
Data Analysis Evaluation
Review new data elements from data set merges
Review new derived or computed data elements
Verification of analysis timeframe and any data subject to change over time
Are statistical analyses correct?
Do counts, sums, averages make sense?
© 2017 American Health Information Management Association
Imputation
Method for dealing with missing data
Remove the record from data analysis
Coded missing data
E.g. -999 for missing data
Imputation
Replace missing values with median of data set or mean of points near the missing data
© 2017 American Health Information Management Association
Clear Data Presentation
Words—concise and accurate
Tables
The table should be a logical unit that is self-explanatory and stands on its own
The source of the data in the table should be specified
Headings for rows and columns should be understandable
Blank cells should contain a zero or a dash
Formatting for headings and cell contents should be consistent so that the eye is not confused (Horton 2013, 508)
© 2017 American Health Information Management Association
Caitlin Wilson (CW) – Update reference here and on remaining slides
Charts and Graphs
Guiding Principles
Distortion—The representation of numbers or percentages should be proportional to the quantities represented.
Proportion and scale—Graphs should emphasize the horizontal and be greater in length than height. A general rule is that the y-axis (height) be three-quarters the x-axis (length) of the graph.
Abbreviations—Any abbreviations used should be spelled out for clarity.
Color —Color should be used as appropriate to the use of the graph. If the chart is going to be printed will it be printed in black and white or color?
Text—The font and use of capitalization needs to be considered carefully. The use of all capital letters can sometimes be difficult to read (Horton 2013, 510)
© 2017 American Health Information Management Association
Data Visualization
See shape of data
Find skewed data and outliers
Explore trends
See missing data
© 2017 American Health Information Management Association
Summary
Data is used to create information
Both are essential to an effective health care industry
Health informatics professionals need to be able to perform accurate data analyses
© 2017 American Health Information Management Association
