The Relevance of Optimal Preoperative Glycemic Control for Outcomes of Patients With Diabetes Undergoing Surgery

To investigate the association between glycemic control, measured by glycated hemoglobin (HbA 1c ) levels, and postoperative complications across various procedures, identify the prevalence of patients with undiagnosed prediabetes or diabetes undergoing surgery, and explore whether better glycemic management is associated with reduced short-term postoperative complications.

The presence of diabetes and the level of long-term glycemic control, as measured using HbA 1c , can influence perioperative morbidity and mortality. Current evidence is largely based on small, single-center studies, with heterogeneous analysis strategies.

Data from the 2021-2022 multi-institutional American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) were utilized. Adult patients with recorded HbA 1c were selected and categorized into 8 groups based on the presence of a diabetes diagnosis and their HbA 1c values. Various preoperative, surgical, and postoperative variables were extracted and analyzed using multivariate regression models.

The study included 502,478 patients with recorded HbA 1c . Among those diagnosed with diabetes (n=179,697; 35.8%), HbA 1c levels were distributed as follows: 27,516 (5.5%) had HbA 1c <6% (<42 mmol/mol), 63,644 (12.7%) had HbA 1c 6-6.9% (42-52 mmol/mol), 45,534 (9.1%) had HbA 1c 7% to 7.9% (53-63 mmol/mol), 19,985 (4.0%) had HbA 1c 8% to 8.9% (64-74 mmol/mol), and 23,018 (4.6%) had HbA 1c ≥9% (≥75 mmol/mol). Of the patients without a diabetes diagnosis (n=322,781; 64.2%), 91,097 (18.1% of total) had HbA 1c in the prediabetes range (5.7-6.4%, 39-46 mmol/mol) and 25,818 (5.1% of total) in the diabetes range (>6.4%, ≥47 mmol/mol). Multivariate logistic analysis revealed that, compared with patients without prediabetes or diabetes, patients with undiagnosed prediabetes had a lower risk of surgical complications, while patients with undiagnosed diabetes had a higher risk of medical complications. Patients with diagnosed diabetes with near-normal glycemia and those with very poor glycemic control had higher odds of any complication. Patients with undiagnosed prediabetes and undiagnosed diabetes had significantly lower chances of transfusion. Patients with undiagnosed diabetes and patients with diagnosed diabetes with near-normal glycemia, suboptimal glycemic control, or very poor glycemic control had significantly longer hospital stays compared with patients without prediabetes or diabetes.

Glycemic control significantly impacts morbidity and mortality in surgical patients. A total of 23% of patients were patients with undiagnosed prediabetes or diabetes, underscoring the importance of preoperative HbA 1c screening for all patients. Both very low and very high HbA 1c levels should be preoperatively addressed, with moderate control (HbA 1c 7%-8%, 53-64 mmol/mol) identified as optimal. Overall, these findings emphasize the need for personalized diabetes management plans tailored to each patient's needs and should inform clinical guidelines.