Bilateral Internal Mammary Artery Use in Diabetic Patients: Friend or Foe?

Abstract

Bilateral internal mammary artery (BIMA) grafting in diabetic patients undergoing coronary artery bypass grafting remains controversial. Our study compared morbidity and mortality between (1) diabetic and nondiabetic BIMA patients and (2) diabetic BIMA versus diabetic patients who underwent left internal mammary artery (LIMA) grafting only. Patients who underwent isolated coronary artery bypass grafting from July 2011 to June 2016 at any of the 10 Maryland Cardiac Surgery Quality Initiative centers were propensity scored across 16 variables. Diabetic BIMA patients were matched 1:1 by nearest neighbor matching to nondiabetic BIMA patients and were separately matched 1:1 to diabetic LIMA patients. We calculated observed-to-expected (O/E) ratios for composite morbidity/mortality, operative mortality, unplanned reoperation, stroke, renal failure, prolonged ventilation, and deep sternal wound infection and compared ratios among matched populations. During the study period, 812 coronary artery bypass grafting patients received BIMA grafts, including 302 patients (37%) with diabetes. We matched 259 diabetic and nondiabetic BIMA patients. O/E ratios were higher in matched diabetic (versus nondiabetic) BIMA patients when comparing composite morbidity/mortality, reoperation, stroke, renal failure, and prolonged ventilation (all O/E ratios >1.0); however, the O/E ratio for operative mortality was higher in nondiabetic BIMA patients. We additionally matched 292 diabetic BIMA to diabetic LIMA patients. Diabetic BIMA patients had a higher O/E ratio for composite morbidity/mortality, operative mortality, stroke, renal failure, and prolonged ventilation. In this statewide analysis, diabetic patients who received BIMA grafts (compared with diabetic patients with LIMA grafts or nondiabetic patients with BIMA grafts) had higher O/E ratios for composite morbidity/mortality as a result of higher O/E ratios for major complications.