Glycemic Control in ICU Patients

The BG target range during the study period was 90 to 120 mg/dl for all patients admitted to the ICU, a modest upward revision of the target range shown to improve mortality and morbidity of populations of critically ill patients in previously published interventional trials [5 (link),6 (link)] (Additional file 1). This target was chosen explicitly to maximize the percentage of values within a broader 70 to 140 mg/dl range, a range that the ICU nurses felt that they could achieve. We chose this range because (1) <70 mg/dl is a widely used definition of hypoglycemia and (2) ≥140 mg/dl is a widely accepted threshold for hyperglycemia. Nurses performed BG monitoring using ACCU-CHEK Inform II glucose meters (Roche Diagnostics, Indianapolis, IN, USA) to test capillary, venous or arterial blood. Monitoring guidelines precluded use of capillary blood in the setting of shock or marked peripheral edema. The measurement frequency was every 3 hours at a minimum for all patients. Sustained hyperglycemia—two consecutive BG readings ≥180 mg/dl—triggered the institution of continuous intravenous regular insulin infusion and hourly BG measurement. The nurses treated lesser degrees of hyperglycemia with subcutaneous insulin aspart at an interval of every 3 hours. It is the standard of care in the ICU to initiate nutritional support in the first 24 to 48 hours of admission. Patients requiring more than 10 U/day of insulin who were receiving a continuous source of calories were typically administered insulin glargine to supply a portion of their daily insulin requirement. The typical starting dose of insulin glargine was one-third to one-half of the previous 24-hour insulin requirement.