Risk scores do not adjust for aggressive, evidence-based changes in percutaneous coronary intervention practice patterns

ABSTRACT  Aim: Public reporting of procedural outcomes leads to risk averse behavior because physicians do not believe the scores account for patient risk. We investigated the effects of more aggressive percutaneous coronary intervention (PCI) practice on risk-adjusted mortality. Methods & results: 8935 PCI were performed. Risk adjustment was performed with the New York State PCI risk score. The cohort was divided into two eras based on programs implemented to promote more aggressive care. Between eras, overall adjusted mortality ratios rose from 0.66 to 0.90 (observed/predicted, p = 0.02), despite evidence supporting consistent procedural quality. Conclusion: Evidence-based changes in PCI practice were associated with worsening risk-adjusted procedural mortality. These data are consistent with physician beliefs regarding risk-adjusted outcome measures.