No continuous relationship between Veterans Affairs hospital coronary artery bypass grafting surgical volume and operative mortality

Background. The purpose of this study was to determine whether risk-adjusted coronary artery bypass grafting mortality rates are significantly related to coronary artery bypass grafting surgical procedure volume within the Department of Veterans Affairs hospital system. Methods. From April 1987 to September 1992, expected mortality rates were calculated for 23,986 coronary artery bypass grafting procedures performed at 44 different Veterans Affairs hospitals. Results. This study found a statistically significant relationship between annual hospital coronary artery bypass grafting volume and observed mortality rates (p < 0.02). However, no statistically significant relationship between coronary artery bypass grafting volume and risk-adjusted operative mortality was found (p = 0.10). Using analysis of variance on hospital-level data, hospitals with 100 or less cases per year have higher observed to expected mortality ratios than hospitals performing more than 100 cases per year (p = 0.03). Using Poisson regression models, however, a volume threshold could not be found. Conclusions. These findings are consistent with the current Veterans Affairs policy requirements to periodically review quality at low-volume hospitals.