Hypoglycemia with intensive insulin therapy in critically ill patients: Predisposing factors and association with mortality*

OBJECTIVES: : To examine the predisposing factors for hypoglycemia in medical-surgical intensive care unit patients treated with intensive insulin therapy and to assess its association with mortality. DESIGN: : Nested-cohort study within a randomized controlled trial. SETTING: : Tertiary care intensive care unit. PARTICIPANTS: : Medical-surgical intensive care unit patients with admission blood glucose of >6.1 mmol/L or 110 mg/dL who were enrolled in a randomized controlled trial comparing intensive insulin therapy with conventional insulin therapy. INTERVENTIONS: : None. EXPOSURE: : Hypoglycemia was defined as blood glucose < or =2.2 mmol/L or 40 mg/dL and intensive care unit mortality was the primary outcome. MEASUREMENTS AND MAIN RESULTS: : Among the 523 patients included in the study, hypoglycemia occurred in 84 (16%). Intensive insulin therapy was independently associated with increased risk of hypoglycemia (adjusted odds ratio, 50.65; 95% confidence interval, 17.36-147.78; p < .0001). Other variables associated with an increased risk of hypoglycemia included female gender, diabetes, Acute Physiology and Chronic Health Evaluation II, mechanical ventilation, continuous veno-venous hemodialysis, and intensive care unit length of stay. When adjusted to potential confounders, hypoglycemia was not significantly associated with increased mortality (adjusted hazard ratio, 1.31; 95% confidence interval, .70-2.46; p = .40). Patients with admission blood glucose of < or =10 mmol/L had an increased mortality with hypoglycemia (adjusted hazard ratio, 4.43; 95% confidence interval, 1.36-14.44; p = .01). Crude analysis showed significant association of mortality with blood glucose levels of < or =1.2 mmol/L (adjusted hazard ratio, 2.92; 95% confidence interval, 1.05-8.11; p = .04). When adjusted analysis was performed, similar trend was seen but was not statistically significant (adjusted hazard ratio, 2.56; 95% confidence interval, .85-7.70; p = .10). CONCLUSIONS: : Our study showed significant increase of hypoglycemia with intensive insulin therapy. Although hypoglycemia was not independently associated with increased risk of death, increased mortality could not be excluded with severe hypoglycemia and in patients admitted with blood glucose of < or =10 mmol/L.