A fair individualized polysocial risk score for identifying increased social risk in type 2 diabetes

• Published in: Nature communications Vol. 15; no. 1; pp. 8653 - 11
• Main Authors: Huang, Yu, Guo, Jingchuan, Donahoo, William T., Lee, Yao An, Fan, Zhengkang, Lu, Ying, Chen, Wei-Han, Tang, Huilin, Bilello, Lori, Saguil, Aaron A., Rosenberg, Eric, Shenkman, Elizabeth A., Bian, Jiang
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 05-10-2024; Nature Publishing Group; Nature Portfolio
• Subjects: 692/163/2743/137/773; 692/308/409; 692/700/459/1748; Adult; Aged; Diabetes; Diabetes mellitus (non-insulin dependent); Diabetes Mellitus, Type 2 - epidemiology; Electronic Health Records; Electronic medical records; Ethnicity; Explainable artificial intelligence; Female; Florida - epidemiology; Hospitalization; Hospitalization - statistics & numerical data; Humanities and Social Sciences; Humans; Machine Learning; Male; Middle Aged; Minority & ethnic groups; multidisciplinary; Optimization; Racial Groups; Risk; Risk Assessment - methods; Risk Factors; Science; Science (multidisciplinary); Social Determinants of Health; Social discrimination learning; Florida
• ISSN: 2041-1723; 2041-1723
• DOI: 10.1038/s41467-024-52960-9

Racial and ethnic minorities bear a disproportionate burden of type 2 diabetes (T2D) and its complications, with social determinants of health (SDoH) recognized as key drivers of these disparities. Implementing efficient and effective social needs management strategies is crucial. We propose a machine learning analytic pipeline to calculate the individualized polysocial risk score (iPsRS), which can identify T2D patients at high social risk for hospitalization, incorporating explainable AI techniques and algorithmic fairness optimization. We use electronic health records (EHR) data from T2D patients in the University of Florida Health Integrated Data Repository, incorporating both contextual SDoH (e.g., neighborhood deprivation) and person-level SDoH (e.g., housing instability). After fairness optimization across racial and ethnic groups, the iPsRS achieved a C statistic of 0.71 in predicting 1-year hospitalization. Our iPsRS can fairly and accurately screen patients with T2D who are at increased social risk for hospitalization.
Racial and ethnic minorities bear a disproportionate burden of type 2 diabetes (T2D) and its complications, with social determinants of health (SDoH) recognized as key drivers of these disparities. Here, the authors developed an individualized polysocial risk score (iPsRS) to screen for unmet social needs essential to hospitalization risk in patients with Type 2 Diabetes, incorporating fairness optimization and explainable AI.