Differential impact of non-insulin-dependent diabetes mellitus and insulin-dependent diabetes mellitus on breast reconstruction outcomes

While the comparative safety of breast reconstruction in diabetic patients has been previously studied, we examine the differential effects of insulin and non-insulin-dependence on surgical/medical outcomes. Patients undergoing implant/expander or autologous breast reconstruction were extracted from the National Surgical Quality Improvement Program 2005-2012 database. Preoperative and postoperative variables were analyzed using chi-square and Student's t test as appropriate. Multivariate regression modeling was used to evaluate whether non-insulin-dependent diabetes mellitus (NIDDM) or insulin-dependent diabetes mellitus (IDDM) is independently associated with adverse 30-day events following breast reconstruction. Of 29,736 patients meeting inclusion criteria, 23,042 (77.5 %) underwent implant/expander reconstructions, of which 815 had NIDDM and 283 had IDDM. Of the 6,694 (22.5 %) patients who underwent autologous reconstructions, 286 had NIDDM and 94 had IDDM. Rates of overall and surgical complications significantly differed among non-diabetic, NIDDM and IDDM patients in both the implant/expander and autologous cohorts on univariate analysis. After multivariate analysis, NIDDM was significantly associated with surgical complications (OR 1.511); IDDM was significantly associated with medical (OR 1.815) and overall complications (OR 1.852); and any type of diabetes was significantly associated with surgical (OR 1.58) and overall (OR 1.361) complications after autologous reconstruction. Diabetes of any type was not associated with any type of complication after implant/expander reconstruction. In this large, multi-institutional study, diabetes mellitus was significantly associated with adverse outcomes after autologous, but not implant-based breast reconstruction. The multivariate analysis in this study adds granularity to the differential effects of NIDDM and IDDM on complication risk.