1. National Data on Health Literate Care: An Introduction to Public Use Health Literacy Data from MEPS
Lan Liang Center for Financing, Access & Cost Trends
Cindy Brach
Center for Delivery, Organization, & Markets

2. MEPS OVERVIEW: Outline
What is MEPS
Components, Purpose, Design and Sample
Core and Supplemental Content
Supplemental CAPI sections & Questionnaires
MEPS SAQ and Its Health Literacy Data
CAHPS communication and health literacy items
How to Use MEPS Data
Data Files
Website and Web Tools
Data Center

3. What is MEPS

4. Medical Expenditure Panel Survey – Household Component
Nationally representative survey of 15,000 households
Provides national estimates of:
Health care use
Expenditures
Insurance coverage
Sources of payment
Access to care
Health care quality
MEPS is primarily designed to provide nationally representative data on the types of health care Americans use, how frequently they use them, how much is paid for the services and the sources of payments.
MEPS supports distributional estimates of expenditures. For example, in 2004 the top 1% of the population accounted for 23% of total expenditures, while the bottom 50% of the population accounted for only 3% of total expenditures.
MEPS also provides information on the types and cost of private health insurance available to and held by the U.S. population, and can be used to examine the association of demographics and expenditures and can track trends over time.

5. MEPS-HC Survey Design
Subsample of households participating in the previous year’s National Health Interview Survey
Representative of the civilian non-institutionalized population of the USA.
Collects data for 2 years of health care usage from each panel
5 in-person interviews over 2½ year period in English or Spanish
One respondent per household
May not be able to report accurately certain types of information (e.g., type of health plan, detailed event information, diagnoses)
Person- and family-level data collected

6. MEPS Panel Design: Data Reference Periods
2014
2015
2016 Q1 Q2 Q3 Q4
Panel 18
Round 3
Round 4
Round 5
Panel 19
Round 1
Round 2
Panel 20
Panel 21
Sample Size
N = 33,162
N = 33,893
N =33,259
N is equal to the number of people with a positive person weight on the file.

7. MEPS-HC SAMPLE SIZES Sample size limitations preclude some analyses
Year
Number of families
Number of persons
2016
13,587
33,259
2015
13,800
33,893
2014
13,421
33,162
2013
13,936
35,068
2012
14,763
37,182
2011
13,449
33,622
2010
12,445
31,228
2009
13,875
34,920
2008
12,316
31,262
2007
11,615
29,370
2006
12,811
32,577
2005
12,810
32,320
2004
13,018
32,737
2003
12,860
32,681
2002
14,828
37,418
2001
12,852
32,122
2000
9,515
23,839
1999
9,345
23,565
1998
9,023
22,953
1997
13,087
32,636
1996
8,655
21,571
Sample size limitations preclude some analyses

8. MEPS-HC Core Content Collected in every round (5x) Demographics
Charges and Payments
Medical Conditions
Employment
Health Status
Health Insurance
Utilization

9. Core Interview Content
Demographics
Age
Sex
Race and Ethnicity
Language and English Proficiency
Marital Status
Foreign Born Status
Student Status and Educational Attainment
Military Service and Honorable Discharge
Income and Poverty Status

10. Core Interview Content (cont’d)
Charge and Payments (CP)
Tracks total charges and sources of payment for medical events reported in earlier sections
Conditions (CN)
Collects additional information about physical and mental health conditions
Employment (EM)
Covers questions about each person's employment or self-employment status

11. Core Interview Content (cont’d)
Health Insurance (HX)
Includes information re private and public health insurance
Health Status (HE)
Adult and child physical and mental health
For children includes immunizations, limitations to school attendance, and more.
Utilization
Use and expenditures for office- and hospital-based care, home health care, dental services, vision aids, and prescribed medicines
Data collected for each person at the event level (e.g., doctor visit, hospital stay)

12. Supplemental Sections
Sections asked in rounds 3 and 5:
Income
Preventive Care
Priority Conditions (panel 21 onwards in rounds 1, 3 and 5)
Assets (round 5 only)
Sections asked in rounds 2 and 4:
Access to care
Child Health and Preventive Care
Adult Self-Administered Questionnaire (SAQ)

13. Supplemental Sections Round 3 and 5 Sections
Income (IN)
Collects information about income and tax returns
Preventive Care (AP)
Collects information on any preventive care received (dental and physical check-ups, flu shots, and other preventative health exams)
Assets (AS)
Asks about household members' real estate, businesses, vehicles, investments, other assets, and debts

14. Round 2 and 4 Sections Access to Care (AC)
Identifies usual source of care (if any), difficulties getting care, and language assistance.
Child Preventive Health (CS)
Collects information on general health status, special health care needs, potential behavioral problems, accessibility to health care, preventative care, height, and weight of any child in the family.
Self-Administered Questionnaire (SAQ)
Paper and pen survey of 48 questions given each adult (18 years old and older)

15. What questions do you have?

16. MEPS SAQ and Its Health Literacy Data

17. Adult Self-Administered Questionnaire (SAQ)
Access to care, including specialist
3 measures of health status:
Short-Form 12 Version 2 (SF-12v2)
Kessler Index (K6) of non-specific psychological distress
Patient Health Questionnaire (PHQ-2 – mental health screener)
Communication: CAHPS® measures
Receipt of health literate care
Global rating
Receipt of preventive services
Opinions about health care

18. SAQ Limitations Questions are not asked out loud and must be read
Percentage completed by person other than intended respondent (about 11% in 2016)
Available only in English and Spanish
Response rate: between 88.7% and 94.2% from 2011 to 2014
Forms that are not collected in Rounds 2 and 4 are requested again in Rounds 3 and 5

19. CAHPS Communication Questions in MEPS
In the last 12 months, how often did doctors or other health providers:
Listen carefully to you?
Explain things in a way that was easy to understand?
Show respect for what you had to say?
Spend enough time with you?
Responses: Never, Sometimes, Usually, Always

20. CAHPS Communication Data and Measures
Data available since 2000
Poor communication composite reported in National Healthcare Quality and Disparities Report
Used as Healthy People measures
For HP 2030 will use poor composite measures instead of individual items
Used for payment incentives

21. Health Literacy Data and Measures
Data available from 2011
Measures of health literate care, NOT of respondent’s health literacy
3 Measures
2 of health literate communication with providers
1 of organizational health literacy
Responses: Never, Sometimes, Usually, Always

22. MEPS Measures of Health Literate Care
If received instructions about what to do about a specific illness or health condition:
How often were these instructions easy to understand?
How often did doctors or other health providers ask you to describe how you were going to follow these instructions? (initiated teach-back)
If asked to fill out or sign any forms at a doctor’s or other health provider’s office
How often were you offered help in filling out a form at the doctor’s or other health provider’s office?

23. Proportion of Adults Age 25 + Who Reported Receiving Health Literate Care
Instructions: 10% increase (p<.001)
Source: Liang & Brach 2017

24. Proportion of Adults Age 25 + Who Reported Receiving Health Literate Care (cont’d)
Teach-Back: 22% increase (p<.001)
Source: Liang & Brach 2017

25. Proportion of Adults Age 25 + Who Reported Receiving Health Literate Care (cont’d)
Forms: 16% increase (p<.1)
Source: Liang & Brach 2017

26. Regression Analysis Combined 2011-2014 data Independent variables:
Demographics (gender, age group, marital status, education, race/ethnicity, comfort speaking English)
Health status (perceived health status and perceived mental health status, taking five or more prescription drugs, body mass index categories, smoking status)
Location (U.S. Census region, and metropolitan statistical area)
Usual source of care (gender, race, specialty, and practice location of the usual source of care)
Year fixed effects

27. Regression Results (p<.001)
Gave easy-to-understand instructions
Asians less likely than White Non-Hispanics
Asked to teach-back
Less educated more likely than post-graduates
Asians, Non-Hispanic Blacks, & Hispanics more likely than White Non-Hispanics
Offered help with forms
Elderly (75+ ) more likely than year olds
Less than high school more likely than post-grads
Uncomfortable speaking English more likely
Lack of LEP findings – sample was constrained to people with a USC. Only half of LEP have a USC, but those almost all that do have a USC report language assistance (either a bilingual clinician or interpreter).
Source: Liang & Brach 2017

28. Potential Research Questions
Do providers who reportedly deliver health literate care differ from those who don’t?
Does gender or racial/ethnic concordance affect reports of health literate care?
Are reports of health literate care associated with lower out-of-pocket expenditures?
Are people without a usual source of care more or less likely to report health literate care?

29. Potential Research Questions Cont’d
Are reports of health literate care associated with the global provider rating?
Is there an association between reports of health literate care and health care utilization?
Hospitalizations?
Prescriptions medicines?
Preventive services?
Can a reliable composite measure of health literate care be constructed?

30. What questions do you have?

31. How to Use MEPS Data

32. MEPS-HC Data Files for Public Use (PUF’s): Levels
Person Level - detailed person information
Each record represents a person, has all of person’s demographics, health, income, expenses, etc.
Event Level - detailed event information
Each record represents an event, such as a hospital visit, has all details on conditions, expenditures, etc. for that visit
Condition Level - detailed condition information
Each record represents a condition, all details on that condition are on that record
Job Level - detailed job information
Each record represents a job and all details associated with it

33. MEPS-HC Data Files for Public Use (PUF): Types
Full-year Files – Contain expenditure and utilization data for the calendar year from several rounds of data collection
Consolidated Data File
Event File
Medical Conditions File
Jobs File
Person Round Plan Public Use File

34. MEPS-HC Data Files for Public Use (PUF): Supporting Documents
Documentation Files
Contain general information about MEPS
Contain survey information specific to each file
Discuss file variables
Contains variable-source crosswalk to link back to questionnaire items
Instructions on how to link files
File Codebooks
Contains names and location of all variables
Provides formatted frequencies for all variables
List both weighted and un-weighted estimates
Users Notes
SAS, SPSS, & STATA Programming Statements
Data file in SAS transport and ASCII formats
Sample code

35. MEPS Web Site https://meps.ahrq.gov/mepsweb/
Materials on the MEPS Web site
Micro Data Files- Public Use Files (PUFs)
Questionnaires: Core and Supplemental
Redesigned HC Summary Data Tables Platform
Medical utilization and expenditures
Demographic and socioeconomic characteristics
Health insurance coverage
Access to care and satisfaction with care
Prescribed medicine purchases
Publications
Statistical briefs
Methodology reports
Research findings

36. AHRQ Data Center (ADC) Purpose Location Access Statistical software
Provides researchers access to non-public use MEPS data (except directly identifiable information)
Location
Secure room
Access
Terminal connected to secure LAN
No internet connectivity
Statistical software
SAS, STATA, SUDAAN, R
Limited staff support by experts on:
Data
Confidentiality issues
Software

37. AHRQ Data Center: Procedures
Submit proposal to Data Center coordinator
Review within 2 weeks for feasibility, and data availability
Institutional Review Board (IRB) review required from users’ institute
Sign the Data Center agreement
Fee - $300 to cover technical assistance and simple file construction, waived for full time students
Run analysis – on or off-site – depending on the project

38. What questions do you have?