Introduction: Transfusion of red blood cells and other blood products is common during cardiac surgery. Diabetes mellitus (DM), present in up to one in four cardiac surgery patients, is associated with chronic hyperglycemia that may enhance coagulation. This prothrombotic state could reduce bleeding and transfusion needs. While prior studies link poor glycemic control with altered hemostasis, the role of intraoperative glucose levels is less clear. This quality improvement initiative evaluated how diabetes and intraoperative hyperglycemia affect transfusion practices during and after cardiac surgery to optimize risk stratification and resource use.
Methods: We retrospectively analyzed 24,005 adult patients who underwent index cardiac surgery with cardiopulmonary bypass at a single center from January 2018 to June 2024. Exclusions included repeat surgeries and transplants. The primary outcome was the total number of blood product units transfused intraoperatively and through postoperative day one. Multivariable negative binomial regression was used, adjusting for clinical and surgical variables. Diabetes was defined based on preoperative diagnosis. Intraoperative glucose was analyzed as a continuous covariate.
Results: DM was present in 6,071 patients (25%). Median transfusion volume was 2 units [IQR 0–6] for patients with DM and 1 unit [IQR 0–4] for those without. After adjustment, DM was associated with fewer transfusions (adjusted geometric mean ratio [GMR] 0.94; 95% CI 0.90–0.98; p< 0.01). In contrast, each 10 mg/dL increase in intraoperative glucose correlated with higher transfusion needs (GMR 1.02; 95% CI 1.02–1.03; p< 0.01).
Conclusions: Our project evaluated the relationship between glycemic profile and transfusion practices in a large cohort undergoing cardiac surgery. We found that chronic hyperglycemia, reflected by a diagnosis of diabetes, was associated with lower transfusion needs, while acute intraoperative hyperglycemia correlated with increased blood product use. These findings highlight that chronic and acute hyperglycemia have opposing effects on transfusion requirements. Incorporating both baseline diabetic status and real-time glucose measurements may improve transfusion planning and perioperative resource use.
