As costs of survey data collection rise, researchers increasingly look to administrative data sources as supplements or substitutes. For those interested in health disparities, electronic health records (EHRs) is one such source. An advantage of EHRs is up-to-date detailed information on patient visits, lab reports, diagnoses, medications, and the like for a very large N. However, challenges arise when trying to relate this information to appropriate denominators, especially if the goal is to describe racial and ethnic health disparities. The present study sheds new light on this problem by linking EHRs from a large integrated health delivery system in North Carolina to microdata from the American Community Survey (ACS) to consider: (a) racial and ethnic differences in the assignment of Protected Identification Keys (PIKs), anonymized identifiers used by the Census Bureau to link datasets; (b) racial and ethnic differences in the success of linking to persons with PIK assignments included in the ACS 2001-2017; and (c) differences in the racial and ethnic identification of persons in appearing in both datasets. Results are discussed in terms of what they imply for the description of health disparities based on EHR-derived data.
Barbara Entwisle is a Kenan Distinguished Professor of Sociology at the University of North Carolina at Chapel Hill and Fellow of the Carolina Population Center.
This talk is co-sponsored by the UW Center for Studies in Demography & Ecology.