Evaluating Percutaneous Coronary Intervention Safety, Quality, and Appropriateness Across Michigan Using Blinded Cross-Institutional Peer Review.

Abstract

Background: Several quality improvement initiatives have focused on the quality gap in percutaneous coronary intervention (PCI), yet significant variations in quality persist. Our objective was to use a novel blinded peer review system to evaluate PCI quality, safety, and appropriateness across Michigan.

Methods: Single-vessel PCI cases were randomly selected from the Blue Cross Blue Shield of Michigan Cardiovascular Consortium registry across Michigan (2018-2020), and anonymized angiograms and pertinent case records were uploaded to a secure server. Cases were reviewed by blinded interventional cardiologists internal and external to the institution, using a standardized peer review form and rated on procedural quality, safety, and appropriateness. We compared appropriateness ratings between reviewers and registry-based appropriateness criteria.

Results: We conducted 1627 independent peer reviews of 961 cases; 23.7% of cases were for non-ST-segment-elevation myocardial infarction, and 36.4% were for ST-segment-elevation myocardial infarction. The majority (96.4%) of reviewers rated angiogram quality as excellent or adequate. Reviewers noted a complication or suboptimal result in 11.1% of reviews; 44.0% of these were deemed avoidable. Most PCI procedures were considered appropriate or may be appropriate, (87.1%) by all those reviewing. Reviewers were less likely to categorize PCI cases as appropriate compared with registry-based appropriate use criteria definitions (73.1% versus 93.3%). The percentage of cases rated as both appropriate/may be appropriate and technically competent ranged from 76.7% to 100% across sites.

Conclusions: While the overall quality and appropriateness of PCI in Michigan are high, key opportunities to improve care were identified. Additional studies are needed to assess the utility of expanding this approach across the United States.