Enhanced family history-based algorithms increase the identification of individuals meeting criteria for genetic testing of hereditary cancer syndromes but would not reduce disparities on their own

• DOI: 10.1016/j.jbi.2023.104568
• Publication Date: 2023-12-09T13:50:40Z
• Authors: Richard L. Bradshaw; Kensaku Kawamoto; Jemar R. Bather; Melody S. Goodman; Wendy K. Kohlmann; Daniel Chavez-Yenter; Molly Volkmar; Rachel Monahan; Kimberly A. Kaphi...; Guilherme Del Fiol
• Keywords: Identification; Disadvantaged

This study aimed to 1) investigate algorithm enhancements for identifying patients eligible genetic testing of hereditary cancer syndromes using family history data from electronic health records (EHRs); and 2) assess their impact on relative differences across sex, race, ethnicity, language preference. The used EHR a tertiary academic medical center. A baseline rule-base algorithm, relying structured (structured data; SD), was enhanced natural processing (NLP) component relaxed criteria (partial match [PM]). identification rates were analyzed considering Among 120,007 aged 25-60, detection rate found all groups the SD (all P<0.001). Both increased rates; NLP led 1.9% increase (PM) an 18.5% (both Combining with PM yielded 20.4% (P<0.001). Similar increases observed within subgroups. Relative persisted most categories algorithms, disproportionately higher who are White, Female, non-Hispanic, whose preferred is English. Algorithm syndromes, regardless However, in persisted, emphasizing need additional strategies reduce disparities such as addressing underlying biases information selectively applying disadvantaged populations. Systematic assessment performance population subgroups should be incorporated into development processes.