Impact of steroid use and glycemic control on postoperative complications in diabetic gynecologic oncology patients undergoing laparotomy

We aimed to assess the impact of preoperative steroid administration and perioperative glycemic control on postoperative complications in diabetic gynecologic oncology patients undergoing laparotomy.

This retrospective cohort study included gynecologic oncology patients with Type I and Type II diabetes (DM) undergoing laparotomy for any gynecologic indication at a single academic center from 10/2017 to 09/2020. The primary outcome was the rate of postoperative complications. Preoperative steroid administration and 24-hour postoperative average serum blood glucose (BG) ≥ 180 mg/dL were the studied exposures. Data was analyzed with SPSS Statistics v.28.

225 patients met inclusion criteria; 47.6 % had postoperative complications. Patient demographics were similar between patients with and without postoperative complications. Patients with complications had higher BMIs (36.8 vs. 34.0; p = 0.03), bowel surgery (33.0 % vs. 17.1 %; p = 0.008), operative time ≥ 240 min (14.2 % vs. 5.1 %; p = 0.02) and average BG ≥ 180 (63.6 % vs. 40.2 %; p < 0.01). On multivariate analysis, bowel surgery (OR 2.4 (1.2-4.8); p = 0.01) and average BG ≥ 180 (OR 2.8 (1.6-4.9); p < 0.01) remained significant predictors of postoperative complications. There were no differences in complication rates (42.3 % vs. 42.6 %; p = 1.0) between patients who received preoperative steroids and those who did not. When stratified by average postoperative BG < 180 mg/dL vs. BG ≥ 180 mg/dL, there was no difference in Clavien-Dindo classification, 30-day readmission rate (28.2 % vs. 22.1 %; p = 0.49) or 30-day mortality rate (2.9 % vs. 0.0 %; p = 0.53).

The administration of preoperative steroids did not increase complication rates. Perioperative hyperglycemia was associated with an increased risk of postoperative complications. Optimizing perioperative glycemic control is imperative to decrease postoperative complications.