Abstract 1286: Socioeconomic Impacts of a Proposal for Hospital Pay-for-Performance
Under pay-for-performance (P4P), the Centers for Medicare and Medicaid Services (CMS) propose to define performance as a blend of attainment (absolute performance score) and improvement (increase in performance score). Hospitals would be rewarded either by attaining a sufficiently high score or improving a sufficient amount. We quantify the likely impact of this proposal on hospitals in socioeconomically disadvantaged areas of the country, where performance is historically poor, but improvement may be achievable. Using a national hospital sample from 2004–2007, we identified institutions as locationally “disadvantaged” along 5 local dimensions:
few high school graduates; and
few with college education in the local workforce.
Composite process of care scores were generated for acute myocardial infarction (AMI) and congestive heart failure (CHF). For each condition, a blended score (scaled 0–10) was generated for each hospital, using the CMS approach. On all 5 dimensions, attainment favored the most locationally advantaged hospitals, with the greatest mean score difference being 3.1 points for AMI (4.9 for highest quartile college educated vs. 1.8 for lowest quartile college educated in the workforce, p<0.001). In most, but not all, categories, improvement again favored locationally advantaged hospitals, but to a lesser degree, with the greatest mean score difference being 0.9 points for CHF (4.4 for hospitals in counties with no clinician shortage vs. 3.5 for counties with a clinician shortage, p=0.027). Scores that blended attainment and improvement favored locationally advantaged hospitals substantially, with the greatest mean blended score difference being 1.7 points for AMI (5.5 for highest quartile college educated vs. 3.8 for lowest quartile college educated, p<0.001). These patterns generally held in multivariate analysis controlling for individual hospital characteristics. With a basis in attainment and improvement, the CMS blended approach will disproportionately reward hospitals in socioeconomically advantaged locations. Modifications may be needed to ensure that P4P does not transfer funds from disadvantaged to advantaged areas of the US.