Glycemic Status and Elective Surgery Risk: Clinically Recognized Hyperglycemia vs Diabetes Diagnosis Alone in a National Cohort.

Abstract

Background: Diabetes affects 20% to 25% of surgical patients, but whether elective surgery should be deferred for patients with diabetes absent clinically recognized hyperglycemia remains unclear. We evaluated perioperative risk by glycemic status in adults undergoing elective surgery.

Study design: Retrospective cohort study of the National Inpatient Sample, 2016 to 2022. Adults undergoing elective joint arthroplasty, lumbar spine fusion, cholecystectomy, colectomy, cardiac surgery, or hysterectomy were classified as no diabetes, diabetes without hyperglycemia, or diabetes with hyperglycemia using ICD-10 Clinical Modification codes. The primary endpoint was composite major adverse events (mortality, acute kidney injury, sepsis, MI, ischemic stroke, or venous thromboembolism). Survey-weighted multivariable logistic regression adjusted for demographics, payer, procedure type, year, hospital characteristics, and Elixhauser comorbidities.

Results: Among 2,104,338 hospitalizations (~10.5 million weighted), 1,641,797 (78.0%) had no diabetes, 398,095 (18.9%) had diabetes without hyperglycemia, and 64,446 (3.1%) had diabetes with hyperglycemia. Adjusted major adverse event probability was 5.31% for no diabetes, 5.85% for diabetes without hyperglycemia, and 9.16% for diabetes with hyperglycemia. Diabetes without hyperglycemia was associated with modestly increased odds of major adverse events (adjusted odds ratio 1.13; 95% CI 1.11 to 1.15; absolute risk difference 0.55%; number needed to harm 183) and ranked 16th of 17 surgical risk factors examined. Diabetes with hyperglycemia was associated with substantially higher risk (adjusted odds ratio 2.00; 95% CI 1.94 to 2.05; absolute risk difference 3.85%; number needed to harm 26).

Conclusions: Diabetes without clinically recognized hyperglycemia was a comparatively weak risk factor, whereas clinically recognized hyperglycemia identified a substantially higher perioperative risk. These findings suggest glycemic status may better inform elective surgical risk stratification than diabetes diagnosis alone.