A Computational Framework for Tailored Preventive Care Recommendations Using Electronic Health Records

Most healthcare systems worldwide are designed to be reactive. According to the U.S. Centers for Disease Control and Prevention (CDC), 90% of the nation's $3.3 trillion annual healthcare expenditures are attributed to individuals with chronic and mental health conditions. Thus, preventing diseases is essential to improving public health and managing escalating healthcare costs. However, the preventive care clinical decision support (CDS) modules in most electronic health record (EHR) systems primarily rely on basic criteria such as age, gender, and screening intervals. This "one-size-fits-all" approach fails to provide personalized recommendations that consider patient-specific risk factors, such as family history, social history, ethnicity, and chronic conditions. Social and environmental determinants of health are increasingly recognized as significant contributors to disease, disability, and mortality in the U.S., yet limited research has been conducted on leveraging artificial intelligence (AI) to analyze the U.S. Preventive Services Task Force (USPSTF) guidelines and EHR data for generating personalized preventive care recommendations. This research develops a computational framework that integrates information from preventive care guidelines with patient-specific data, including risk factors extracted from EHRs, to generate personalized recommendations. The system will offer tailored recommendations along with justifications grounded in both the EHR data and preventive care guidelines.