Elimination of cardiopulmonary bypass improves early survival for multivessel coronary artery bypass patients

• Published in: The Annals of thoracic surgery Vol. 73; no. 4; pp. 1196 - 1203
• Main Authors: Magee, Mitchell J, Jablonski, Kathleen A, Stamou, Sotiris C, Pfister, Albert J, Dewey, Todd M, Dullum, Mercedes K C, Edgerton, James R, Prince, Syma L, Acuff, Tea E, Corso, Paul J, Mack, Michael J
• Format: Journal Article
• Language: English
• Published: Netherlands 01.04.2002
• Subjects: Aged; Aged, 80 and over; Cardiopulmonary Bypass - adverse effects; Coronary Artery Bypass - adverse effects; Coronary Artery Bypass - mortality; Female; Humans; Logistic Models; Male; Postoperative Complications; Retrospective Studies; Risk Factors; Survival Rate
• ISSN: 0003-4975
• DOI: 10.1016/s0003-4975(01)03587-1

Coronary artery bypass graft (CABG) surgery performed without cardiopulmonary bypass (CPB) is currently increasing in clinical practice. Decreased morbidity associated with off-pump (OP) CABG in selected risk groups examined in relatively small, single institution groups has been the focus of most recent studies. The purpose of this study was to determine the independent impact of CPB on early survival in all isolated multivessel CABG patients undergoing surgery in two large institutions with established experience in OPCABG techniques. A review of two large databases employed by multiple surgeons in the hospitals of two institutions identified 8,758 multivessel CABG procedures performed from January 1998 through July 2000. In all, 8,449 procedures were included in a multivariate logistic regression analysis to determine the relative impact of CPB on mortality independent of known risk factors for mortality. Procedures were also divided into two treatment groups based on the use of CPB: 6,466 had CABG with CPB (CABG-CPB), 1,983 had CABG without CPB (OPCABG). Disparities between groups were identified by univariate analysis of 17 preoperative risk factors and treatment groups were compared by Parsonnet's risk stratification model. Finally, computer-matched groups based on propensity score for institution selection for OPCABG were combined and analyzed by a logistic regression model predicting risk for mortality. CABG-CPB was associated with increased mortality compared with OPCABG by univariate analysis, 3.5% versus 1.8%, despite a lower predicted risk in the CABG-CPB group. CPB was associated with increased mortality by multiple logistic regression analysis with an odds ratio of 1.79 (95% confidence interval = 1.24 to 2.67). An increased risk of mortality associated with CPB was also determined by logistic regression analysis of the combined computer-matched groups based on OPCABG-selection propensity scores with an odds ratio of 1.9 (95% confidence interval = 1.2 to 3.1). Elimination of CPB improves early survival in multivessel CABG patients. Rigorous attempts to statistically account for selection bias maintained a clear association between CPB and increased mortality. Larger multiinstitutional studies are needed to confirm these findings and determine the most appropriate application of OPCABG.