Early and midterm clinical outcome in patients with severe left ventricular dysfunction undergoing coronary artery surgery

Background. Patients presenting with severe left ventricular (LV) dysfunction undergoing coronary artery surgery are at increased risk of perioperative morbidity and mortality. The present study investigated early and midterm outcomes in a consecutive series of patients with severe LV dysfunction undergoing coronary surgery at our institution. Methods. Data on 5,195 consecutive patients undergoing coronary artery bypass grafting (CABG) alone (in-hospital mortality 1.35%) from April 1996 to August 2002 were prospectively recorded in the Patient Analysis and Tracking System. Two hundred and fifty patients (median age 65 years [interquartile range, 57 to 70]) with preoperative left ventricular ejection fraction less than 30% (74 off pump; 29.6%) were identified and early and midterm clinical outcomes analyzed. Propensity scores were used to take account of the imbalance in the distribution of prognostic factors between the on-pump and off-pump groups. Results. Patients undergoing on-pump surgery were less likely to have current congestive heart failure, insulin-dependent diabetes, a history of hypertension, have had gastrointestinal tract surgery or an ulcer, or unstable angina. They had on average lower Parsonnet scores and New York Heart Association and Canadian Cardiovascular Score ratings. However they were more likely to have more extensive coronary artery heart disease and to require more grafts than those undergoing off-pump surgery. After adjustment for consultant team and propensity scores no differences between groups with regard to in-hospital mortality and morbidity were found. The only in-hospital outcome to show a significant difference after adjustment was the need for intraoperative inotropic support, which was higher in the on-pump group (odds ratio 5.1; 95% confidence interval 2.55 to 10.2; p < 0.001)). The median follow-up times for the on- and off-pump groups were 3.4 years and 1.4 years respectively. Three-year survival was higher with on-pump surgery (87% on-pump versus 73% off-pump) but this difference did not reach statistical significance after adjustment for prognostic variables (hazard ratio 0.54, 95%) confidence interval 0.22 to 1.26, p = 0.16). Conclusions. In-hospital mortality and morbidity in patients presenting with severe LV dysfunction is low with comparable results with both on- and off-pump coronary artery surgery. Midterm clinical outcome is encouraging and seems to justify surgical revascularization for this high-risk group of patients. (C) 2003 by The Society of Thoracic Surgeons.