Hypoglycemia with intensive insulin therapy after cardiac surgery: predisposing factors and association with mortality.

Background: Intensive insulin therapy has become a major therapeutic target in cardiac surgery patients. It has been associated, however, with an increased risk of hypoglycemia compared with conventional insulin therapy. Our study sought to identify the factors predisposing to hypoglycemia with intensive insulin therapy and investigate its effect on early clinical outcomes after cardiac surgery.

Methods: A concurrent cohort study of 2,538 consecutive patients undergoing cardiac surgery (coronary artery bypass grafting, valve, or bypass grafting and valve surgery) from January 2005 to March 2010 was carried out. Multivariable logistic regression analysis and propensity score matching were used (1) to identify the risk factors for developing hypoglycemia (blood glucose < 60 mg/dL) after cardiac surgery and (2) to compare major morbidity, operative mortality, and actuarial survival between patients in whom hypoglycemia developed (n = 77) and those in whom it did not (n = 2461). The propensity score-adjusted sample included 61 patients in whom hypoglycemia developed and 305 patients in whom it did not (1 to 5 matching).

Results: Risk factors for hypoglycemia included female gender (odds ratio [OR] = 2.3, 95% confidence intervals [CI] = 1.4-3.7; P < .001), diabetes (OR = 2.8, CI = 1.7-4.5; P < .001), hemodialysis (OR = 3.0, CI = 1.3-6.8; P = .009), intraoperative blood product transfusion (OR = 2.0, CI = 1.2-3.4; P = .010), and earlier date of surgery (years of surgery, 2005-2007; OR = 2.1, CI = 1.2-3.7; P = .007) . Hypoglycemia increased the risk for operative mortality in univariate (hypoglycemic 10% vs normoglycemic patients 2%; P < .001) but not in propensity score- adjusted analysis (OR= 2.5, 0.9-6.7; P = .11). The propensity score-adjusted analysis demonstrated a significant increase in hemorrhage-related reexploration (P = .048), pneumonia (P < .001), reintubation (P < .001), prolonged ventilatory support (P < .001), hospital length of stay (P < .001), and intensive care unit length of stay (P < .001) for the hypoglycemic compared with normoglycemic patients. Five-year actuarial survival was similar in the compared patient groups (hypoglycemic 75% vs normoglycemic 75%; P = .22).

Conclusions: Hypoglycemia with intensive insulin therapy is independently associated with increased risk for respiratory complications and prolonged hospital and intensive care unit lengths of stay after cardiac surgery. In our study, hypoglycemia was not independently associated with increased risk of death.