Hospital coronary artery bypass graft surgery volume and patient mortality, 1998-2000

Objective: To evaluate the association between annual hospital coronary artery bypass graft (CABG) surgery volume and in-hospital mortality. Summary Background Data: The Leapfrog Group recommends health care purchasers contract for CABG services only with hospitals that perform greater than or equal to500 CABGs annually to reduce mortality; it is unclear whether this standard applies to current practice. Methods: We conducted a retrospective analysis of the National Inpatient Sample database for patients who underwent CABG in 1998-2000 (n = 228,738) at low (12-249 cases/year), medium (250-499 cases/year), and high (greater than or equal to500 cases/year) CABG volume hospitals. Crude in-hospital mortality rates were 4.21% in low-volume hospitals, 3.74% in medium-volume hospitals, and 3.54% in high-volume hospitals (trend P < 0.001). Compared with patients at high-volume hospitals (odds ratio 1.00, referent), patients at low-volume hospitals remained at increased risk of mortality after multivariable adjustment (odds ratio 1.26, 95% confidence interval = 1.15-1.39). The mortality risk for patients at medium-volume hospitals was of borderline significance (odds ratio 1.11, 95% confidence interval = 1.01-1.21). However, 207 of 243 (85%) of low-volume and 151 of 169 (89%) of medium-volume hospital-years had risk-standardized mortality rates that were statistically lower or comparable to those expected. In contrast, only 11 of 169 (6%) of high-volume hospital-years had outcomes that were statistically better than expected. Conclusions: Patients at high-volume CABG hospitals were, on average, at a lower mortality risk than patients at lower-volume hospitals. However, the small size of the volume-associated mortality difference and the heterogeneity in outcomes within all CABG volume groups suggest individual hospital CABG volume is not a reliable marker of hospital CABG quality.