Outcomes following coronary artery bypass surgery in diabetic treatment sub-groups. A propensity matched analysis of > 7000 patients over 18 years

Background European System for Cardiac Operative Risk Evaluation II incorporates insulin-controlled diabetes whilst omitting tablet-controlled diabetes. Differences in adverse clinical outcomes following coronary artery bypass graft between these groups are poorly established. Therefore, a propensity matched comparison of short and longer term mortality and morbidity in insulin-controlled diabetes, tablet-controlled diabetes and non-diabetic patients was undertaken. Methods Isolated first-time coronary artery bypass graft surgeries between April 1999 and April 2017 were propensity score matched by pre- and intra-operative variables. Results 8241 patients; 23.5% diabetics and 76.5% non-diabetics. The groups' demographical and clinical characteristics were comparable after matching. Insulin-controlled diabetes patients had significantly higher in-hospital mortality (3.8% vs. 1.7%, p < 0.05), multisystem failure (2.6% vs. 1.8%, p < 0.05), sternal wound infections requiring debridement (3.6% vs. 1.3%, p < 0.05), respiratory complications (25.6% vs. 21.9%, p < 0.05), new dialysis (4.7% vs. 0.9%, p < 0.05) and longer hospital stays (13.5 +/- 13.3 vs. 10.6 +/- 8.0, p < 0.05) compared to non-diabetic patients. Tablet-controlled diabetes patients had significantly higher strokes (2.9% vs. 1.2, p < 0.05), superficial sternal wound infections (6.7% vs. 5.4%, p < 0.05), respiratory complications (25.7% vs. 22.7%, p < 0.05), new dialysis (1.7% vs. 0.6%, p < 0.05), post-operative atrial fibrillation (37.1% vs. 33.9%, p < 0.05) and readmission with myocardial infarction (22.4% vs. 19.6%, p < 0.05) compared to non-diabetic patients. Conclusion Diabetic treatment sub-groups are an independent risk factor for sternal wound infection, new dialysis requirement, multisystem failure and readmission with myocardial infarction after isolated first coronary artery bypass graft surgery. The findings suggest the need for better risk stratification of diabetic groups prior to cardiac surgery and for improved cardiovascular risk management post-surgery in tablet-controlled diabetes patients.