MEPS Medical Expenditure Panel Survey - onetomapanalytics/Meta_Data GitHub Wiki

MEPS - Medical Expenditure Panel Survey

General description

  1. Database primary purpose - Provide nationally representative estimates of health care use, expenditures, sources of payment, and health insurance coverage for the U.S. civilian non-institutionalized population. The MEPS Household Component (HC) also provides estimates of respondents’ health status, demographic and socio-economic characteristics, employment, access to care, and satisfaction with health care. Estimates can be produced for individuals, families, and selected population subgroups
  2. Overall data type - Demographics, health outcomes, use of medical care services, charges, and payments
  3. Dataset type - Cross-sectional (except for the Panel 17 longitudinal data file, which can be used to analyze changes over a two-year period 2012-2013)
  4. Data source - Survey
  5. Data level - Patient level
  6. Geographic location of the data collection sites - United States
  7. Sponsor, manager, or home institution - Agency for Healthcare Research and Quality (AHRQ)
  8. Date range - 2013 - 2014
  9. Geolocation data - Census region
  10. Dates - Month and year of birth and of the reference period
  11. Longitudinal tracking - Track patients within MEPS files through a person identifier unique within the data year
  12. Financial variables - Charges and expenditure
  13. Clinical areas of interest - All
  14. Number of records - The following files - which can be merged with each other and with other 2013 MEPS HC data files - are available:
  • Full-Year Consolidated Data files: 36,940 persons
  • Full-Year Population Characteristics files: 34,875 persons
  • Medical Conditions files: 123,875 medical condition records
  • Person Round Plan files: 58,947 records
  • Longitudinal Data files: 17,923 persons
  • Prescribed Medicines files: 327,557 prescribed medicine records
  • Dental visits: 28,547 event records
  • Other Medical Expenses files: 7,421 other medical expenditure records
  • Hospital Inpatient Stays files: 2,862 hospital inpatient stay records
  • Emergency Room Visits files: 7,510 emergency room visits records
  • Outpatient Visits files: 13,145 outpatient event records
  • Office-Based Medical Provider Visits files: 165,764 office-based provider event records
  • Home Health files: 5,982 home health records
  • NHEA-Aligned MEPS Projected Expenditure Data Files: 37,418 respondents
  1. Variables that are uniquely present in this dataset - MEPS is the most complete source of data on the cost and use of health care and health insurance coverage.
  2. Other - MEPS has two key components at the moment: the Household Component and the Insurance Component. The Household Component collects information from individual households and their members, augmented by information from medical providers. The Insurance Component is a distinct employer survey that provides information on employer-based health insurance. MEPS also includes a Medical Provider Component (MPC), which comprises hospitals, physicians, home health care providers, and pharmacies that MEPS-HC respondents indicated. Its goal is to supplement and/or replace information provided by MEPS-HC respondents. MEPS only included a Nursing Home Component (NHC) in 1996, which collected information from a national sample of nursing homes and residents on the characteristics of the facilities and services offered; expenditures and sources of payment on an individual resident level; and resident characteristics such as functional limitation, cognitive impairment, age, income, and insurance coverage. The NHC also gathered information on the availability and utilization of community-based services prior to nursing home admission. For reasons of confidentiality, NHC data are only available at the AHRQ Data Center or one of the Federal Statistical Research Data Centers. The Nationwide Center for Health Statistics (NCHS) provides information on the NCHS National Nursing Home Survey (NNHS), a continual series of national sample surveys of nursing facilities, residents, and staff conducted in 1973-74, 1977, 1985, 1995, 1997, and 1999. Please, see the section "Data dictionary" for files present in the Onetomap project.

Applicable methods

  1. Association methods, such as logistic regression models (1, 2, 3), multivariable linear regression (4), binomial regression analysis (5), generalized linear model (6)
  2. Cluster analysis (7)
  3. Cost-effectiveness analysis (8)
  4. Difference in differences (9, 10)
  5. Exploratory analysis (11, 12)
  6. Factor analysis (13, 14)
  7. Inferential tests (11)
  8. Interrupted time series (15, 16)
  9. Machine learning (17, 18)
  10. Propensity scores (19, 20, 21)
  11. Sensitivity analysis (13, 22, 23)

High-impact designs

  • Characterize medication use by and within medication class(es) (24)

  • Evaluate racial/ethnic disparities in screenning (25), hospitalization and mortality (26)

  • Estimate national disability-associated health care expenditures and analyze spending by insurance and service categories (27)

  • Investigate the association between weight status and pain treatment (28)

  • Assess the impact of Medicare on access to and affordability of health care (29, 30, 31)

  • Describe a robust method for relative risk estimation for vaccine effectiveness (32)

  • Examine the association between mental health comorbidities and patients' satisfaction with physicians (33)

  • Estimate the average incremental health care expenditures associated with habitual long and short duration of sleep (34)

  • Examine sex differences in pharmaceutical treatment and the potential consequences of pharmaceutical side effects on distress and depression (35)

  • Describe the proportion of patients in contact with a primary care physician (36)

  • Demonstrate the impact of medical costs and quality of life losses of other diseases in the life years gained on the cost-effectiveness of cancer screening (8)

  • Compare the quality of chronic disease management received by patients with and without dementia (37)

  • Examine the association of medication prescriptions with non-US-born status (38)

  • Determine the association between self-employment and lack of health insurance coverage (39)

  • Investigate the services that contribute to higher costs and utilization (40)

  • Compare the healthcare expenditures associated with multimorbidity versus no multimorbidity (41)

  • Examine the relationship between macroeconomic fluctuations and medical care usage (42)

Data dictionary

To access MEPS data dictionaries, click here

Variable categories

  1. Geographic variables (i.e., census region)
  2. Demographic variables (e.g., age, sex, race, ethnicity, marital status, educational attainment, language and English proficiency, foreign-born status, military service, poverty status)
  3. Income and tax filing variables (e.g., filing status, dependents, food stamps, family income)
  4. Person-level priority condition variables (e.g., perceived health and mental health status, diagnosis, pregnancy)
  5. Health status variables (e.g., assistive devices, physical and cognitive limitations, time since last screening)
  6. Disability days variables (e.g., missed work due to ill)
  7. Access to care variables (e.g., language spoken, born in the U.S., type of provider, unable to get medical and dental care)
  8. Employment variables (e.g., status, number of jobs, hourly wage)
  9. Health insurance variables (e.g., covered by TRICARE, CHAMPVA, Medicare, SCHIP, or other public or private insurance)
  10. Utilization, expenditure, and source of payment variables (e.g., amount paid, household reported total charge, total expenditure for event, total facility charge)
  11. Prescribed medicines (e.g., date patient first used medicine, quantity prescribed)
  12. Visits (e.g., dental, inpatient, outpatient, emergency room, office-based medical provider, home health)

Linkage to other datasets

  • Linkages can be established with the previous year’s NHIS providing additional data for longitudinal analytic purposes through the NHIS - MEPS link files