This MEPS public use release is one in a series of data releases drawn from the 1997-2000 Medical Expenditure Panel Survey (MEPS) Household Component (HC). This release contains four Person Round Plan (PRPL) files, one for each year 1997 (File 1) through 2000 (File 4). The HC-047 files contains records for persons insured through establishments providing hospital/physician, medigap, dental or prescription medication coverage and includes variables pertaining to managed care and satisfaction with plan coverage. These are designed to facilitate research on the sometimes complex and dynamic relationships between consumers and their private insurance. Released as ASCII files with SAS, SPSS, and STATA programming statements, and in SAS Transport format, the PRPL files are not person-level files and linking a PRPL file to a MEPS full year person-level file (e.g., HC-038 for year 2000) requires users making analytic decisions based on understanding the complexity of the PRPL file.
SPSS and STATA Programming Statements have been added.
Update #2: 12/27/10
The value of COVTYPIN (single or family coverage) was incorrectly set for a small number of plans. These cases have all four of the following conditions: (1) every coresiding family member’s coverage ceased before the end of the reference period (interview date or December 31st, so that EVALCOVR≠1), (2) the policyholder was a coresiding family member, (3) the plan did not cover anyone residing elsewhere, and (4) at least one coresiding dependent had coverage during the reference period. In these cases, COVTYPIN was erroneously set to single. Beginning in 2010, if coverage ceased for every coresiding family member before the end of the reference period, COVTYPIN is set to family if there were two or more persons covered any time during the reference period. Users can reset the values for these cases in prior years with the available data. An example in SAS follows:
proc sort data=h111 out=prpl; by eprsidx; run; data eprs(keep=eprsidx numin numever); set prpl; by eprsidx; retain numin numever; array stat{24} status1-status24; if first.eprsidx then do; numin=0; numever=0; end; if substr(dupersid,6,3) ne '901' then do; do i=1 to 24; if stat{i}=1 then covered=1; end; if covered=1 then numever=numever+1; if evalcovr=1 then numin=numin+1; end; if last.eprsidx then output eprs; run; data prplx(drop=numin numever covtypin rename=(covtypix=covtypin)); merge prpl eprs; by eprsidx; if numin=0 and covtypin=1 and numever>1 then covtypix=2; else covtypix=covtypin; label covtypix='COVERAGE: 1=SINGLE, 2=FAMILY'; run;
Update #1: 09/30/03
The variable LTCINS contains household reports of long-term care insurance, but estimates of the prevalence of long-term care insurance do not benchmark well to data from the insurance industry. The variable LTCINS should not be used and will n be released in future files.
*The PDF version of the codebook is recommended for printing; the HTML version is database driven and lets you navigate quickly to details on each variable.
**Right-click on the data file link, then select Save Target As or Save Link As to download the file.
Back
Email Updates
