Newbie - need help understanding MEPS data sets


I am intern who has limited data experience just trying to understand how to get through these MEPS data sets. My goal is to calculate the total average cost for childhood asthma in 2000. When extracting data I select - year, exptot, mepsidrd, age, and asthma. I’m getting confused on the rounds of data and whether or not its the same people in the separate year rounds or the same. It confuses me because the MEPS ID doesn’t match in the other round. Honestly this could all sound very un experienced but could use some serious help.

It seems like you have two questions: one about a specific calculation and another about MEPSID generally.

First, calculating the average total medical expenditures related to asthma for children in 2000 is actually more complicated than you might imagine, and currently not possible using only MEPS data from IPUMS.

IPUMS Health Surveys offers the full year consolidated (FYC) file from the MEPS household component (MEPS-HC). This reports summary information about expenditures and expenditures by provider type. You could use these data to calculate the average medical expenditures for children who were ever diagnosed with asthma. However, to explicitly associate expenditures with certain conditions and/or events (e.g., prescription costs related to asthma or any expenditure for care related to asthma), you need to leverage a separate file (the CLINK file); these files are not included in IPUMS MEPS currently. For more information, you may be interested in this diagram describing the relationship between MEPS public use files.

I am not sure if I understand what you mean by MEPSID doesn’t match in the other round, so am sharing a bit more information that may be helpful. However, please follow up by responding to this post or emailing with further questions. MEPS is a short panel survey; people are interviewed 5 times over the course of 2 years. The IPUMS MEPS variable MEPSID combines the dwelling unit identifier, person identifier, and panel. You can use MEPSID to identify the same person in multiple rounds of data (they will have 3 round records per year in a hierarchical file). I am sharing a bit more information about the MEPS panel design for reference. It would be helpful to know what type of data structure you are using in your extract; IPUMS MEPS allows users to request data structured in different formats.