Inconsistent correlation between procedural volume and publicly reported outcomes in adult cardiac operations

BACKGROUND: Statewide public reporting provides transparent surgical outcomes. The objective of this study is to evaluate the correlation between procedural volume and publicly reported outcomes following adult cardiac surgery. METHODS: The Pennsylvania Health Care Cost Containment Council (PHC4) statewide public reporting databases were analyzed. Isolated coronary artery bypass grafting (CABG), isolated valve surgery, and CABG plus valve surgery performed between 2014 and 2016 were included. The primary outcomes were operative mortality and 30-day readmission. Expected operative mortality and 30-day readmission were calculated using the risk models developed by PHC4. Observed-to-expected (OE) ratios were correlated with procedural volume using weighted linear regression analysis. RESULTS: The study included 29 578 operations (16 641 isolated CABGs, 8618 isolated valves, and 4319 CABG plus valves) performed by 182 surgeons at 60 hospitals. The expected risk of operative mortality for surgeons was 1.5%, 1.8%, and 4.3%, and for hospitals 1.5%, 1.7%, and 4.3% for isolated CABGs, isolated valves, and CABG plus valves, respectively. Expected 30-day readmission for surgeons and hospitals was 10.3%, 13.4%, and 14.4% and 10.2%, 13.2%, and 14.3% for the same operations, respectively. There was an inconsistent correlation between surgeon and hospital volume and OE operative mortality or 30-day readmission for any of the index operations. CONCLUSION: In this study of 29 578 index adult cardiac operations there is not a consistent association between surgeon or hospital volume and mortality or readmission for publicly reported outcomes. These data suggest that volume is not a reliable predictor of surgeon or hospital level OE outcomes in publicly reported data.