Abstract 3895: The Impact of Pay for Performance (P4P) on Quality of Care in Acute Coronary Syndromes: Initial External Evaluation of the Center for Medicare and Medicaid Services P4P Pilot
Background: Pay for Performance (P4P) has been promoted as a powerful tool to improve quality; yet there are limited data supporting its effectiveness and/or potential unintended consequences. In 2003, the Center for Medicare and Medicaid Services (CMS) launched a P4P pilot project to measure and reward care in several areas including myocardial infarction (MI) care. Using data from the CRUSADE initiative, we compared trends in the care and outcomes of 41,443 patients with non-ST-segment elevation MI at P4P vs non-P4P centers.
Methods: P4P hospitals (n=57) were matched 1:2 to non-P4P centers (n=114) based on their baseline composite ACC/AHA guideline adherence score and facility characteristics. From 2003 to 2006, temporal trends in care were examined for 3 categories:
financially rewarded CMS metrics (aspirin, beta blockers, angiotensin-converting enzyme inhibitors, smoking cessation counseling);
non-financially rewarded guideline recommendations (ECG <10 min, heparin, glycoprotein IIb/IIIa inhibitors, cath <48 hrs, d/c Statin, d/c clopidogrel), and
in-hospital mortality rates.
Aggregate adherence rates for CMS and non-CMS measures were calculated. Differences in trends at P4P vs non-P4P sites were evaluated using a repeated variance measure model with time as a covariate.
Results: Aggregate CMS-rewarded performance metrics as well as non-rewarded but indicated care processes both improved significantly over time at P4P hospitals (both p<0.001). However, the rate of improvement over time was not significantly different between P4P vs non-P4P hospitals for either the CMS-rewarded measures (p=0.196) or for the non-rewarded care processes (p=0.489). Mortality rates also improved over time and to a similar degree at P4P vs non-P4P sites (p=0.240).
Conclusion: This study is one of the first to evaluate CMS’s P4P pilot program. Among centers participating in a voluntary quality improvement program, we could not show that P4P had a significant incremental impact on MI care quality or patient outcomes. We also did not find evidence that P4P had an adverse impact on improvement in care processes not subject to financial incentives. Further studies of P4P are needed to determine its optimal role in quality improvement initiatives.