EP508 - Undiagnosed Diabetes Prior to Colorectal Surgery is Associated with Postoperative Mortality: A Retrospective Cohort Study

Image or Table (Do Not add text here):

Purpose/Background: The prevalence of diabetes amongst patients undergoing surgical interventions is 10-20%. An even larger proportion, 23-60%, of patients may have undiagnosed diabetes. Only 85% of patients with diabetes have appropriate glycemic management postoperatively. Presumably the proportion of patients with undiagnosed diabetes receiving postoperative glycemic management would be lower. Postoperative hyperglycemia has been associated with increased risk of major postoperative morbidity and mortality. Thus, patients with undiagnosed diabetes may be at increased postoperative risk, yet this is a patient population that has been rarely evaluated. We aimed to identify patients undergoing colorectal surgery without a diagnosis of diabetes but with a preoperative hemoglobin A1C (HbA1C) value of ≥ 6.5 and compare their postoperative outcomes to patients without diabetes and normal HbA1C and patients with diabetes.

Methods/Interventions: Adult patients undergoing proctectomy or colectomy for colorectal neoplasia, diverticular disease, or inflammatory bowel disease from 2021-2022 were identified in the American College of Surgeons (ACS) National Quality Improvement Program (NSQIP) database utilizing Current Procedure Terminology (CPT) codes and International Classification of Diseases, Tenth Revision (ICD-10) codes. Patients without a preoperative HbA1C were excluded. The exposure of interest was a preoperative HbA1C ≥ 6.5 without a preoperative diagnosis of diabetes (Group 1). The comparators were patients with a preoperative diagnosis of diabetes (Group 2) and patients without a diagnosis of diabetes and an HbA1C < 6.5 (Group 3). The primary outcome was 30-day postoperative mortality. Secondary outcomes included 30-day major postoperative morbidity, septic complications, and prolonged postoperative length of stay (LOS). Multivariable regression models were fit for each outcome.

Results/Outcomes: Overall, 35,949 patients were included (Group 1: n=2,930 (8.2%); Group 2: n=12,539 (34.9%); Group 3: n=20,480 (57.0%)). Operative intervention was most commonly performed for colorectal neoplasia (n=25,207, 70.1%). Colectomy with primary anastomosis was the most common operation (n=26,487, 73.7%), and 11.7% (n=4,223) of patients underwent proctectomy. Postoperative mortality occurred infrequently (Group 1: 2.3%; Group 2: 2.1%; Group 3: 1.1%), but patients in Groups 1 and 2 were at increased risk as compared to Group 3 on multivariable analysis (Group 1 OR 1.53, 95%CI 1.14-2.07, p=0.005; Group 2 OR 1.32, 95%CI 1.08-1.61, p=0.006; Group 3 [reference]). On adjusted analyses, the odds of postoperative morbidity (Group 1 OR 1.10, 95%CI 0.98-1.23, p=0.114; Group 2 OR 1.01, 95%CI 0.94-1.08, p=0.86; Group 3 [reference]) and septic complications (Group 1 OR 0.94, 95%CI 0.79-1.11, p=0.46; Group 2 OR 0.98, 95%CI 0.88-1.09, p=0.67; Group 3 [reference]) were similar between groups. Patients in Group 1 were had increased odds of experiencing prolonged LOS (Group 1 OR 1.16, 95%CI 1.05-1.28, p=0.004; Group 2 OR 0.98, 95%CI 0.92-1.04, p=0.53; Group 3 [reference]).

Conclusion/Discussion: Patients without a formal diagnosis of diabetes but elevated preoperative HbA1C are at increased risk of postoperative mortality as compared to patients without a formal diagnosis of diabetes and normal preoperative HbA1C. These patients must be identified preoperatively, and postoperative glycemic management should be considered.