From 2002-2009 there are a handful (six or fewer) of respondents in each year whose SAQ eligibility status does not match the stated universe. For most of these cases, the discrepancy does not affect analyses of the data because these respondents are assigned an SAQ weight value of zero (SAQWEIGHT) (please see the SAQ Eligibility and SAQ Weight section below for more details). However, there are two cases where respondents were marked as ineligible (SAQELIG value of one) and have SAQ weights of zero, but appear to meet all of the SAQ eligibility requirements: MEPSID 0242290117 in 2002 and MEPSID 8480110213 in 2008.

Counterintuitively, a respondent can have an SAQELIG value of two ("Eligible for SAQ, has SAQ data") but have an SAQ weight value of zero. Respondents are only assigned an SAQELIG value of two and non-zero SAQWEIGHT if they meet all of the eligibility criteria, are key (see KEYNESS), and were in scope on the interview date associated with round 2/4 (see INSCOPRD). In all cases where respondents are assigned an SAQ weight value of zero all SAQ variables except for SAQELIG were coded as "inapplicable" in the original MEPS Full Year Consolidated data file delivered by AHRQ.