Alaa Abd-Elsayed, Edward J Mascha, Dongsheng Yang, Daniel I Sessler, Andra Duncan
PURPOSE: Increased glucose variability may be associated with worse outcomes in critically ill patients. hyperinsulinemic normoglycemia provides intensive glucose control during surgery and may reduce glucose variability. Our objective was to compare glycemic variability between two methods of glucose control in cardiac surgical patients: hyperinsulinemic normoglycemia vs standard insulin infusion. We also assessed whether the effect differed between patients with and without diabetes mellitus.
METHODS: We compared measures of glycemic variability, including the primary outcome, average real variability (ARV), and secondary outcomes, within-patient standard deviation (SD) and glucose lability index (GLI), in 252 patients who received hyperinsulinemic normoglycemia and 266 patients who received standard therapy. Data was randomly sampled from each patient treated with hyperinsulinemic normoglycemia, so patients in each group had a similar number of glucose measurements. The significance level for each hypothesis was 0.05, and 0.025 within diabetic status.
RESULTS: For nondiabetic patients, hyperinsulinemic normoglycemia reduced mean glucose measure-to-measure variability for ARV by an estimated -0.23 (97.5% CI -0.30, -0.16) mg/dl/min (P < 0.001) versus standard care. There was no difference in glycemic variability between groups for diabetic patients, with difference in means (97.5% CI) of -0.10 (-0.20, 0.02) mg/dl/min, P = 0.07. Mean SD was lower for hyperinsulinemic normoglycemia patients overall, with difference in means (95% CI) of -19 (-22, -16), P < 0.001, with a stronger effect in nondiabetics (interaction P = 0.042). GLI was also lower with hyperinsulinemic normoglycemia.
CONCLUSION: hyperinsulinemic normoglycemia decreases glucose variability for cardiac surgical patients with a stronger effect in nondiabetic patients.
