Validating self-reported mammography use in vulnerable communities: findings and recommendations.

Abstract

Most health surveys ask women whether they have had a recent mammogram, all of which report mammography use (past 2 years) at about 70% to 80% regardless of race or residence. We examined the potential extent of overreporting of mammography use in low income African-American and Latina women, and whether self-report inaccuracies might bias estimated associations between patient characteristics and mammography use. Using venue-based sampling in two poor communities on the west side of Chicago, we asked eligible women living in two west side communities of Chicago to complete a survey about breast health (n = 2,200) and to provide consent to view their medical record. Of the 1,909 women who screened eligible for medical record review, 1,566 consented (82%). We obtained medical records of all women who provided both permission and a valid local mammography facility (n = 1,221). We compared the self-reported responses from the survey with the imaging reports found in the medical record (documented). To account for missing data, we conducted multiple imputations for key demographic variables and report standard measures of accuracy. Although 73% of women self-reported a mammogram in the last 2 years, only 44% of self-reports were documented. Overreporting of mammography use was observed for all three ethnic groups. These results suggest considerable overestimation of prevalence of use in these vulnerable populations. Relying on known faulty self-reported mammography data as a measure of mammography use provides an overly optimistic picture of utilization, a problem that may be exacerbated in vulnerable minority communities.