How Total Performance Scores of Medicare Value-Based Purchasing Program Hospitals Change Over Time

Background and Purpose: The Medicare Value-Based Purchasing (VBP) program established performance-based financial incentives for hospitals. We hypothesized that total performance scores (TPS) would vary by hospital type. Methods: Value-Based Purchasing reports were collected from 2015 to 2017 and merged with the Centers for Medicare and Medicaid Services (CMS) Impact File data. A total of 3,005 hospitals were grouped into physician-owned surgical hospitals (POSH), accountable care organizations (ACO), Kaiser, Vizient, and General hospitals. Longitudinal linear mixed-effects models compared temporal differences of TPS and secondary composite outcome, process, patient satisfaction, safety, and cost efficiency measures between hospital types. Results: Total performance scores decreased across all hospital types (p < .001). Physician-owned surgical hospitals had the highest TPS (59.9), followed by Kaiser (49.2), ACO (36.7), General (34.8), and Vizient (30.7) (p < .001). Hospital types differed significantly in size, geography, mean case-mix index, Medicare patient discharges, percent Medicare days to inpatient days, Disproportionate Share Hospital payments, and uncompensated care per claim. Scores improved in 84% of POSH and 14.6% of Kaiser hospitals using score reallocations. Conclusion: In comparison with General hospitals, the TPS was higher for POSH and Kaiser and lower for Vizient in part due to weighting reallocation and individual domain scores. Implications: Centers for Medicare and Medicaid Services scoring system changes have not addressed the methodological biases favoring certain hospital types.