SS07. Poor Glycemic Control Is a Strong Predictor of Postoperative Morbidity and Mortality in Patients Undergoing Vascular Surgery

Abstract

Hyperglycemia is a common occurrence in patients undergoing cardiovascular surgery. In several surgical cohorts, especially those undergoing cardiac and colorectal surgical procedures, it has been established that improved perioperative glycemic control has reduced postoperative morbidity and in-hospital mortality. A significant portion of the population with peripheral vascular disease suffers from the sequelae of diabetes and/or metabolic syndrome. Although previous studies have included vascular procedures, little data exist regarding perioperative glycemic control and vascular surgery patient outcomes from a large, single-center study. Our objective was to better understand this relationship and determine which negative perioperative outcomes could be eventually be targeted for intervention. A retrospective review was performed using a prospectively maintained vascular patient database at a large academic center from 2009 to 2013. Patients who underwent carotid endarterectomy, carotid stenting, endovascular aortic repair, and all suprainguinal and infrainguinal lower extremity revascularization procedures were included. Data were collected on demographics, postoperative outcomes, and glucose levels in the perioperative period. Perioperative hyperglycemia was defined as at least one glucose reading >200 mg/dL. The primary outcome was 30-day mortality, with secondary outcomes of complications, return to the OR, and hospital readmission. A total of 1086 patients were identified, of whom 271 (25.0%) had at least one perioperative glucose reading >200 mg/dL within 72 hours of surgery, and these patients had a higher 30-day mortality rate (6.3% vs 1.1%; P < .0001), increased rates of acute renal failure (4.8% vs 1.5%; P = .0016), postoperative stroke (3.0% vs 1.0%; P = .0197), and surgical site infection (6.3% vs 2.7%; P = .0062). In addition, these patients were more likely to require mechanical ventilation >48 hours (8.5% vs 2.8%; P < .0001) and return to the OR (5.9% vs 2.5%; P = .0060). The preoperative diagnosis of diabetes was not associated with an increase in 30-day mortality (2.2% vs 2.5%; P = .8312) or return to the OR (3.7% vs 3.2%; P = .6806). On multivariate logistic regression, independent predictors of 30-day mortality included glucose ≥200 mg/dL within 72 hours of surgery (OR, 5.134; 95% CI, 2.212-11.916; P < .0001), case urgency (OR, 6.745; 95% CI, 2.584-17.606; P < .0001), and male gender (OR, 0.354; 95% CI, 0.152-0.823; P = .0158). This study demonstrates a correlation between perioperative glucose levels and postoperative outcomes in vascular patients. We show a strong association between poor postoperative glycemic control and 30-day mortality. Further investigation is needed to clearly characterize “true” uncontrolled perioperative hyperglycemia and define a level at which we must take action to reduce adverse outcomes.