RESUMO
Background: Racial and ethnic minority groups and individuals facing social disadvantages, which often stem from their social determinants of health (SDoH), bear a disproportionate burden of type 2 diabetes (T2D) and its complications. It is crucial to implement effective social risk management strategies at the point of care. Objective: To develop an electronic health records (EHR)-based machine learning (ML) analytical pipeline to address unmet social needs associated with hospitalization risk in patients with T2D. Methods: We identified real-world patients with T2D from the EHR data from University of Florida (UF) Health Integrated Data Repository (IDR), incorporating both contextual SDoH (e.g., neighborhood deprivation) and individual-level SDoH (e.g., housing instability). The 2015-2020 data were used for training and validation and 2021-2022 data for independent testing. We developed a machine learning analytic pipeline, namely individualized polysocial risk score (iPsRS), to identify high social risk associated with hospitalizations in T2D patients, along with explainable AI (XAI) and fairness optimization. Results: The study cohort included 10,192 real-world patients with T2D, with a mean age of 59 years and 58% female. Of the cohort, 50% were non-Hispanic White, 39% were non-Hispanic Black, 6% were Hispanic, and 5% were other races/ethnicities. Our iPsRS, including both contextual and individual-level SDoH as input factors, achieved a C statistic of 0.72 in predicting 1-year hospitalization after fairness optimization across racial and ethnic groups. The iPsRS showed excellent utility for capturing individuals at high hospitalization risk because of SDoH, that is, the actual 1-year hospitalization rate in the top 5% of iPsRS was 28.1%, ~13 times as high as the bottom decile (2.2% for 1-year hospitalization rate). Conclusion: Our ML pipeline iPsRS can fairly and accurately screen for patients who have increased social risk leading to hospitalization in real word patients with T2D.
RESUMO
PURPOSE: To improve admissions process equity, the Uniformed Services University masked Medical College Admission Test (MCAT) scores at or above the 51st percentile to admissions committee members. This policy was aimed at improving admissions rates for applicants in 2 priority groups: those from races and ethnicities underrepresented in medicine (URM) and those from lower socioeconomic status, represented by first-generation college (FGC) graduates. METHOD: All applicants invited to interview were included: 1,624 applicants from admissions years 2014-2016 before MCAT score masking and 1,668 applicants from admissions years 2018-2020 during MCAT score masking. Logistic regression determined admissions likelihood before and during masking. Independent sample t tests compared average admissions committee scores for all applicants and for those in priority groups. Linear regression determined the weight of MCAT scores on admissions committee scores. RESULTS: Despite there being more priority group applicants during MCAT score masking, the admissions likelihood for an individual priority group applicant decreased during this period. URM applicants had an odds ratio of 0.513 for acceptance during MCAT score masking compared to before masking, and FGC applicants had an odds ratio of 0.695. Masking significantly reduced mean admissions committee scores, which decreased approximately twice as much for priority group applicants as for nonpriority group applicants (0.96 points vs 0.51 points). These score decreases were highest for priority group applicants with MCAT scores above the 67th percentile. Masking reduced the weight of MCAT scores; 10.9% of admissions committee score variance was explained by MCAT scores before masking and only 1.2% during masking. CONCLUSIONS: Despite known disparities in MCAT scores with respect to race, ethnicity, and socioeconomic status, admissions decisions in this study were more equitable when MCAT scores were included. While masking MCAT scores reduced the influence of the exam in admissions decisions, it also reduced admissions rates for URM and FGC applicants.