The hyperglycemic response to major noncardiac surgery and the added effect of steroid administration in patients with and without diabetes

Abstract

Background: The pattern and magnitude of the hyperglycemic response to surgical stress, the added effect of low-dose steroids, and whether these differ in diabetics and nondiabetics remain unclear. We therefore tested 2 hypotheses: (1) that diabetics show a greater increase from preoperative to intraoperative glucose concentrations than nondiabetics; and (2) that steroid administration increases intraoperative hyperglycemia more so in diabetics compared with nondiabetics.

Methods: Patients scheduled for major noncardiac surgery under general anesthesia were enrolled and randomized to preoperative IV 8 mg dexamethasone or placebo, stratified by diagnosis of diabetes. Patients were part of a larger underlying trial (the Dexamethasone, Light Anesthesia and Tight Glucose Control [DeLiT] Trial). IV insulin was given when glucose concentration exceeded 215 mg/dL. The primary outcome measure was the change in glucose from the preoperative to maximal intraoperative glucose concentration. We also report the time-dependent pattern of intraoperative hyperglycemia.

Results: Ninety patients (23% with diabetes) were randomized to dexamethasone, and 95 (29% with diabetes) were given placebo. The mean ± SD change from preoperative to maximal intraoperative glucose concentration was 63 ± 69 mg/dL in diabetics and 72 ± 45 mg/dL in nondiabetics. The mean covariable-adjusted change (95% confidence interval) in nondiabetics was 29 (13, 46) mg/dL more than in diabetics (P < 0.001). For all patients combined, mean glucose increased slightly from preoperative to incision, substantially from incision to surgery midpoint, and then remained high and fairly stable through emergence, with nondiabetic patients showing a greater increase (P < 0.001). For nondiabetics, the mean increase in glucose concentration (97.5% CI) was 29 (9, 49) mg/dL more in patients given dexamethasone than placebo (P = 0.0012). However, there was no dexamethasone effect in diabetics (P = 0.99).

Conclusions: Treatment of intraoperative hyperglycemia should account for the hyperglycemic surgical stress response trend depending on the stage of surgery as well as the added effects of steroid administration. Denying steroid prophylaxis for postoperative nausea and vomiting for fear of hyperglycemic response should be reconsidered given the limited effect of steroids on intraoperative blood glucose concentrations.