Determinants of CMS HCC Risk Scores, Discharge to Community, and Preventable Readmissions in Home Healthcare: Policy and Practice Implications

Abstract

Home healthcare has become a vital component of the U.S. healthcare system, particularly for aging populations and individuals with disabilities. Yet, substantial disparities exist in patient outcomes and risk assessments, raising concerns about the equity and accuracy of performance metrics used by the Centers for Medicare & Medicaid Services (CMS). This dissertation investigates the determinants of three critical CMS metrics—Hierarchical Condition Category (HCC) risk scores, Discharge to Community (DTC) rates, and Potentially Preventable Readmissions (PPR)—to uncover how demographic, socioeconomic, and organizational factors shape home healthcare outcomes. The study integrates CMS data, U.S. Census Bureau demographics, and Home Health Compare datasets spanning from 2016 to 2024. Using interpretable machine learning models and statistical tests, the analysis identifies key predictors of risk and performance at the county level for both disabled and aged/dual-eligible Medicare beneficiaries. Random Forest, Decision Tree, and XGBoost models reveal that race (particularly African American population density), household income, unemployment rate, caregiver availability, and home health agency (HHA) density are significant predictors of higher HCC risk scores. Similarly, agency-level characteristics such as visit volume, Medicare spending per episode, and quality ratings are shown to strongly influence DTC and PPR outcomes, beyond clinical adherence alone. Further, the research evaluates post-pandemic trends, showing that HHAs continue to face systemic challenges including workforce shortages, increased hospital utilization, and declining functional outcomes, with statistically significant performance shifts detected using Mann-Whitney U tests and Cliff’s Delta. The findings have significant policy and practice implications. They highlight the need to refine CMS risk adjustment methodologies to account for structural and demographic disparities. The study also advocates for expanding insurance coverage, caregiver support, and targeted funding for HHAs in underserved communities. By offering a data-driven roadmap for improving equity, quality, and efficiency in home healthcare, this research contributes to the ongoing effort to reform Medicare policy and ensure better care for the nation’s most vulnerable populations.