Concordance between self-reported race ethnicity and that recorded in a veteran affairs electronic medical record

Abstract

Background Using information from electronic health records (EHRs) to examine racial/ethnic health disparities is increasingly common. This study examines the degree of concordance between administratively recorded race/ethnicity and the criterion standard of self-reported race/ethnicity at a tertiary care Veterans Affairs Medical Center (VAMC) in North Carolina. Methods We compared self-reported race among 204 respondents to a cross-sectional mailed survey of patients with diabetes conducted in 2006-2007 to the race/ethnicity recorded in the EHR. Concordance was defined as the percent agreement between self-reported and administratively-reported race. Results The overall response rate to the survey was 68.9% (204 of 296). Of the 204 respondents, 32 (15.7%) reported a different race/ethnicity from the race/ethnicity reported in the EHR. Misclassification resulted from either the patient reporting a race/ethnicity and having the information missing in the EHR (9.3% of respondents) or the EHR having a different race/ethnicity listed than reported by the patient (6.3% of respondents). Limitations This study was conducted at one VAMC. Conclusions While we found misclassification of race/ethnicity in the EHR, the level of discordance is smaller than previously reported in the Veterans Health Administration. Despite this, efforts still need to be made to ensure correct information is included in the EHR.