Title: Postoperative complications after coronary artery bypass graft surgery: a comparison between the off-pump and on-pump techniques
Abstract: EDITOR: Coronary artery bypass grafting (CABG) is one of the most frequently performed operations worldwide and any improvement in its safety and efficacy or any means to increase its applicability and/or reduce hospital costs would have a major impact. One potential method for achieving some of these objectives is off-pump CABG (OP-CABG). Although this technique is intellectually appealing, with many theoretical and practical advantages, there are still major concerns [1-3]. OP-CABG has been performed with increasing frequency in our institution since 1998. Data have been prospectively collected and entered in a computerized database and we have reviewed our experience from January 1998 through July 2002. We identified 2552 patients who underwent isolated, elective CABG. We excluded emergency and combined operations. One thousand and nine hundred and fourteen (75%) procedures were performed with cardiopulmonary bypass (CPB) and 638 (25%) were performed without CPB. The percentage of OP-CABG increased from 6% in 1998 to 47% in 2001. It decreased to 35% in 2002. Procedure selection was at the discretion of the operating surgeon and varied accordingly. All patients received a standard premedication (morphine 0.1 mg kg−1, scopolamine 0.25 mg intramuscularly, and oral diazepam) 1 h before surgery. Anaesthesia was induced with fentanyl and propofol. Orotracheal intubation was facilitated by pancuronium. Anaesthesia was maintained with propofol (2-4 mg kg−1 h−1), isoflurane (end-tidal concentration <1 minimum alveolar concentration (MAC)) and additional doses of fentanyl (maximum total dose OP-CABG: 20 μg kg−1, CPB-CABG: 25 μg kg−1). The initial heparin dose was 300 IU kg−1 for CPB-CABG and 150 IU kg−1 for OP-CABG. Activated clotting time was maintained >480 s for CPB-CABG and >300 s for OP-CABG. Heparin was reversed with protamine sulphate in a 1:1 ratio. Standard normothermic CPB (flow rate 2.4 L min−1 m−2) and myocardial protection with hypothermic blood cardioplegia were used in all CPB patients. In the OP-CABG group the latest generation of stabilizing retractors was used. Coronary shunts were used in most patients to avoid ischaemia during the anastomosis. Haemodynamic instability during cardiac manipulation was managed by placing the patient in the Trendelenburg position, atrial pacing, intravenous (i.v.) fluids and/or use of inotropes. Thermal homeostasis was maintained by keeping the operating room temperature at 20°C, covering the non-operated body of the patient with a warm airflow blanket and warming i.v. fluids. After surgery the patients remained intubated under sedation with propofol (2 mg kg−1 h−1). They were transferred to the intensive care unit (ICU). Extubation and discharge from intensive care followed clinical criteria. Patients were compared according to the surgical treatment (OP-CABG vs. CPB-CABG). Preoperative patient characteristics, individual risk factors, intraoperative course and operative outcomes were compared by univariate and multivariate analyses using SAS (release 8.00 by SAS Institute, Cary, NC, USA) statistical software. Univariate comparisons were done using the two tailed χ2-test with the Yates correction or with the t-test for variance when appropriate. Stepwise multivariate logistic regression analysis (entry values of P < 0.05, exit values P < 0.05) was used to determine the surgeon's propensity to use off-pump technique and the independent predictors of short-term outcome which are reported as odds ratios with 95% confidence intervals (CI). The groups were similar with respect to gender (16% female), history of stroke (6%), diabetes (21%), and previous CABG (5%). Compared to the CPB-CABG group, the OP-CABG group was older (64 ± 9.4 vs. 63 ± 9.9 yr, P < 0.001) and had more patients with previous myocardial infarction (MI) (55% vs. 47%, P < 0.001), renal failure (16% vs. 9%, P = 0.003), chronic obstructive pulmonary disease (15% vs. 12%, P = 0.02) and low left ventricular ejection fraction (21% vs. 14%, P < 0.001). A stepwise multiple logistic regression model was used with the same cut-offs to identify the surgeon's preference for off-pump technique in patients with chronic renal failure (odds ratio: 1.7; 95% CI: 1.3-2.2; P = 0.0001), ejection fraction <40% (odds ratio: 1.5; 95% CI: 1.2-1.8; P = 0.001), preoperative MI (odds ratio: 1.3; 95% CI: 1.1-1.6; P = 0.003) and older age (odds ratio: 1.02; 95% CI: 1.005-1.025 per year; P = 0.002). The mean number of grafts per patient was 2.0 ± 1.08 (range 1-6) in the OP-CABG group vs. 2.9 ± 0.79 (range 1-6) in the CPB-CABG group (P < 0.001). In the OP-CABG group a trend towards an increase in the number of grafts was observed during the study period (1.2 in 1998 and 2.3 in 2002) although it remained significantly lower than in the CPB group during the entire study period (P < 0.001). Postoperative data and complications in the overall population and in the two groups are given in Table 1.Table 1: Postoperative complications (n, %).The stepwise multiple logistic regression model used to identify the perioperative risk factors of short-term outcome confirmed the beneficial effects of OP-CABG in terms of a more frequent uneventful perioperative period with lower need for transfusion of blood products, lower incidence of postoperative MI and/or dysrhythmia, reduced time on mechanical ventilation and ICU stay. Hospital mortality (1.5%) was similar in the two groups (2.2% in OP-CABG vs. 1.3% in CPB-CABG, P = 0.1). Several known risk factors were confirmed to have a strong positive association with mortality in this population: previous CABG (odds ratio: 6.5; 95% CI: 3.0-14.0; P < 0.001), ejection fraction <40% (odds ratio: 3.2; 95% CI: 1.6-6.3; P = 0.0007), chronic renal failure (odds ratio: 2.9; 95% CI: 1.4-6.2; P = 0.005) and chronic obstructive pulmonary disease (odds ratio: 2.3; 95% CI: 1.1-4.9; P = 0.03). These high-risk categories had high hospital mortality rates: previous CABG (10/132 = 7.6%), ejection fraction <40% (17/414 = 4.1%), chronic renal failure (13/280 = 4.6%) and chronic obstructive pulmonary disease (11/326 = 3.4%). In our experience, OP-CABG patients, as compared to those receiving CPB, more frequently had an uncomplicated postoperative course with a shorter duration of postoperative mechanical ventilation and stay in the ICU. These results were obtained in spite of the surgeon's propensity to operate on high-risk patients off-pump and in the absence of an anaesthesiology fast-track protocol. Our observations confirm that off-pump coronary artery surgery is a safe and effective strategy to improve short-term outcome and to optimize hospital resources. A. Zangrillo A. Romano G. Landoni D. Sparicio C. Redaelli F. Pappalardo G. Marino Department of Anesthesiology; Vita-Salute University of Milan; Milan, Italy S. Gulletta Department of Cardiology; Vita-Salute University of Milan; Milan, Italy G. Aletti Department of Mathematics; University of Milan; Milan, Italy O. Alfieri Department of Cardiac Surgery; Vita-Salute University of Milan; Milan, Italy Acknowledgements We are indebted to Santisi Michele, RN, Mella Francesca, RN, Castelnuovo Lara, RN for their support.