Effect of Acute Stress Glycemic Control and Long-Term Glycemic Control on the Incidence of Post-Operative Infection in Diabetics Undergoing Cardiac Surgery

Objective Post-operative infection after cardiac surgery causes prolonged hospital stay and increased mortality. In patients with diabetes, peri-operative and pre-operative glycemic control have been associated with increased risk of post-operative infection. Saudi Arabia is the 7th highest country in the world for the prevalence of diabetes. In our surgical population the incidence of diabetes is 77%. We were aware of a high incidence of post-operative infections in our institution. The aim of this work was to assess how peri-operative and pre-operative glycemic control was related to the six-week incidence of post-operative infection. Method We retrospectively collected data for 174 adult patients with diabetes undergoing cardiac surgery between January 2017 and June 2019. For group analysis of peri-operative glycemic control, a mean value of ≤10 mmol/l was categorized as optimal control and a mean value of >10 mmol/l as sub-optimal control. The admission glucose value, the maximum glucose value and glycosylated hemoglobin A1c (HbA1c) were separately recorded. Admission HbA1c was used for optimal long-term control group (HbA1c ≤ 7%) and sub-optimal long-term control group (HbA1c > 7%). Results Of the 174 patients 60 (34%) developed infection in the six-week post-operative period. No statistically significant difference in infections was seen in the optimal peri-operative control group (n = 24, 14%) compared to sub-optimal peri-operative control group (n = 36, 21%; p = 0.113). However, patients with infection had a significantly higher mean glucose (10.4 mmol/l versus 9.9 mmol/l, p = 0.0316) than no infection. Grouping according to their HbA1c: well controlled group (41, 24.0%) and poor control group (130, 76.0%) showed no difference in infections. However, patients with lower HbA1c had better glycemic control as measured by: initial glucose (r = 0.52, p=<0.001); mean peri-operative glucose (r = 0.45, p=<0.001); maximum recorded glucose (r = 0.41, p=<0.001). Conclusion The majority of our patients presented with sub-optimal long-term glycemic control which we linked to poor stress glycemic control perioperatively. Patients with post-operative infections had higher mean peri-operative blood glucose. With the high incidence of diabetes in Saudi Arabia we have demonstrated the importance of good pre-operative assessment which allows tighter peri-operative glycemic control to reduce post-operative morbidity.