Deep sternal infections after in situ bilateral internal thoracic artery grafting for left ventricular myocardial revascularization: predictors and influence on 20-year outcomes

Background: The incidence and potential factors influencing deep sternal wound infection (DSWI) in a cohort of patients undergoing coronary artery bypass grafting (CABG) using skeletonized bilateral internal thoracic artery (BITA) was explored. Furthermore, we studied influence of DSWI on long-term survival, major adverse cardiac events (MACEs) and repeat coronary revascularization (RCR).Methods: The study cohort consisted of 1,325 consecutive patients who were divided in two groups: patients experiencing DSWI (n=33, group 1) and those who did not have sternal infection (n=1,292, group 2). A logistic regression model was employed to find predictors of DSWI whereas Cox regression and a competing risk models were carried out to test predictors of late death, MACE and RCR, respectively. Follow up was 100% complete and ranged from 1 to 245 months. Median follow-up was 103 months (IQR, 61 to 189 months). Cumulative follow-up was 16,430 patient years.Results: The incidence of DSWI was 2.4%. Multivariable logistic regression analysis found any single independent predictor of DSWI. However, the association of peripheral vascular disease (PVD) and diabetes increased the risk by 1.4 and 1.6 times. When DM was associated with obesity the risk increased by 2.1 and 2.6 times compared to the single factors, respectively. Obese female patients were at a 1.6-fold higher risk when compared to the association of DM with obesity. DSWI was not an independent predictor of long-term survival (HR, 2.31; 95% CI: 0.59–9.12), RCR (SHR, 2.89; 95% CI: 0.65–10.12), or MACE (SHR, 1.98; 95% CI: 0.44–8.56).Conclusions: With an accurate patient selection (i.e., exclusion of obese diabetic females) and strict DM control BITA represents a first choice for most of CABG patients, even at high risk for DSWI. The occurrence of DSWI does not influence long-term survival and late outcomes. Our findings should be confirmed by further larger research.