MEPS respondents underreport emergency department and office visits, but accurately report hospital stays
Research Activities, January 2010, No. 353
People underreport certain types of health care use when responding to the Medical Expenditure Panel Survey (MEPS) of the Agency for Healthcare Research and Quality (AHRQ), reveals a new study. MEPS collects data from U.S. households on their use of hospital and office-based services, prescription drugs, and other health-related services. AHRQ researchers Samuel H. Zuvekas, Ph.D., and Gary L. Olin, Ph.D. (retired), identified Medicare-insured individuals who participated in the MEPS during 2001 to 2003. Each individual was matched to their Medicare enrollment and claims data. This allowed the researchers to compare household-reported information with actual services delivered and outlined in claims to Medicare.
Overall, households accurately reported any hospital stays during the time period, as well as specifying the number of nights spent in the hospital. Reporting was less accurate, however, when it came to information on emergency department (ED) and office visits. ED visits were underreported by one-third and office visits by 19 percent on average. Underreporting varied by income, education, health status, and race/ethnicity. However, most of this variation was small relative to the overall gaps in reporting. That is, underreporting affected all groups, so that relative comparisons between groups are largely unaffected in most empirical analyses using the MEPS.
More details are in "Validating household reports of health care use in the Medical Expenditure Panel Survey," by Drs. Zuvekas and Olin, in the October 2009 HSR: Health Services Research 44(5), pp. 1679-1700. Reprints (AHRQ Publication No. 09-R083) are available from the AHRQ Publications Clearinghouse.