Comparison of coronary bypass surgery with and without cardiopulmonary bypass in patients with multivessel disease.

BACKGROUND Coronary artery bypass grafting can now be performed with or without cardiopulmonary bypass. Our objective was to determine whether off-pump coronary artery bypass grafting is associated with better early outcomes compared with conventional coronary artery bypass grafting. METHODS In 4 centers with off-pump coronary surgery experience, a retrospective analysis of all coronary artery bypass grafting in a 3-year period was performed. Groups were compared to determine selection criteria, mortality, and morbidity, then computer-matched by propensity score to control for selection bias. Multivariate logistic regression identified risk factors predictive of mortality. Specific subgroups most likely to benefit were identified. RESULTS In all, 17,401 isolated coronary artery bypass grafts were performed, 7283 (41.9%) off-pump coronary artery bypass grafts and 10,118 (58.1%) conventional coronary artery bypass with cardiopulmonary bypass. Factors determining selection of patients for off-pump coronary artery bypass grafting included female gender (55.5% vs 44.5%), preexisting renal failure (57.0% vs 43.0%), and reoperations (52.6% vs 47.4%). Operative mortality was 2.8%; off-pump coronary artery bypass grafting versus conventional coronary artery bypass with cardiopulmonary bypass (1.9% vs 3.5%, P <.001) had the same predicted risk. Of the patients with multivessel disease, 11,548 were matched by propensity scoring. Mortality was significantly less in the off-pump coronary artery bypass grafting group (2.8% vs 3.7%, P <.001). By multivariate logistic regression analysis of the matched sample, predictors for mortality were female gender (odds ratio 1.83, confidence interval 1.37-2.44), preexisting renal failure (odds ratio 2.85, confidence interval 2.64-4.95), history of stroke (odds ratio 1.74, confidence interval 1.08-2.80), previous coronary artery bypass grafting surgery (odds ratio 4.22, confidence interval 2.92-6.09), use of cardiopulmonary bypass (odds ratio 2.08, confidence interval 1.52-2.83), and recent myocardial infarction (odds ratio 2.31, confidence interval 1.68-3.22). Cardiopulmonary bypass was predictive of mortality in reoperations, female patients, and patients aged >or= 75 years. Off-pump coronary artery bypass grafting was associated with less morbidity, including reductions in blood transfusion (32.6% vs 40.6%, P <.001), stroke (1.4% vs 2.1%, P =.002), renal failure (2.6% vs 5.2%, P <.001), pulmonary complications (4.1% vs 9.5%, P <.001), reoperation (1.7% vs 3.2%, P <.001), atrial fibrillation (21.1% vs 24.99%, P <.001), and gastrointestinal complications (3.6% vs 4.8%, P =.02). CONCLUSION In 4 centers with beating-heart operation experience, there is an overall early benefit in off-pump surgery, especially in patients traditionally considered at high risk for coronary artery bypass grafting.