A. Ramirez1, G. Stukenborg1, B. Turrentine1, R. Jones1 1University Of Virginia,Charlottesville, VA, USA
Introduction:
To address Medicare’s escalating costs, decreasing quality, and transparency the US Congress enacted the Patient Protection and Affordable Care Act of 2010 (ACA). Section 3001 of ACA established the Hospital Value-Based Purchasing Program (VBP) to measure the value of healthcare provided by participating hospitals. The VPB established quality indicators for processes, outcomes, patient satisfaction and cost per Medicare beneficiary to estimate value (value=quality/cost) of care provided. The domains above were used to calculate the Total Performance Score (TPS) allowing CMS to rank hospitals. For 2015, the Center for Medicare and Medicaid Services (CMS) withheld 1.5% of each Medicare hospitals’ anticipated annual payment to establish a financial framework for redistribution based upon TPS rank. High quality hospitals are rewarded and lower quality hospitals are punished. Review of the top 100 TPS included 27 physician-owned specialty hospitals (POSH) suggesting possible stratification by hospital type/business model.
Methods:
We reviewed the February 2015 VBP database including hospital name, address, unadjusted and adjusted process, outcome, patient satisfaction, cost, and total performance scores. We chose to compare POSH with all hospitals and POSH with another hospital type, University Hospital Consortium (UHC) members. We used the general linear model to estimate the TPS reported for each hospital as a function of hospital category and categorized by POSH and UHC membership. The statistical significance of the association between TPS and hospital category was assessed using the F test statistic at the threshold of p < 0.05.
Results:
The dataset included 3,089 hospitals with TPS ranging from 92.86 to 6.6 and a mean of 41.7. Of these 92 were identified as POSH and 111 were UHC hospitals. The mean TPS for POSH (64.43) was significantly higher than all other hospitals. Results from the general linear model indicate the estimated mean differences in TPS was 23.45 points higher for POSH (p < 0.0001) compared to all other hospitals. The mean TPS for UHC hospitals (36.89) was significantly lower than all other hospitals. The mean difference in TPS was -4.95 points lower for UHC hospitals (p < 0.0001) compared to all other hospitals.
Conclusions:
The Medicare VBP scoring method can effectively sort participating hospitals. Over time this methodology should improve, particularly with the development of progressively better quality indicators. Application of VBP should incentivize quality improvement and decrease healthcare cost. The observation that POSH had higher than average TPS and UHC hospitals had lower than average TPS requires further study. POSH include predominantly surgical centers while UHC hospitals represent a more heterogeneous patient population with large numbers of chronic disease. The observed differences comment on the need to further explore the impact of healthcare business models on outcomes.