Comparing thirty-day outcomes in prosthetic and autologous breast reconstruction: A multivariate analysis of 13,082 patients?

Background There is a paucity of multi-institutional data that directly compares short term outcomes of autologous and prosthetic breast reconstruction. The National Surgical Quality Improvement Program provides a unique data platform for evaluating peri-operative outcomes of these two main categories of breast reconstruction. It has detailed data from nearly 250 hospitals and over 13,000 patients. We performed risk-adjusted analysis of prosthetic and autologous breast reconstruction to compare 30-day morbidity outcomes. Methods Patients who underwent prosthetic breast reconstruction or autologous tissue reconstruction from 2006 to 2010 were identified using operation descriptions. Over 240 tracked variables were extracted for patients undergoing breast reconstruction. Thirty-day postoperative outcomes were compared, and subgroup analysis was performed on the autologous population to describe outcomes of specific flap procedures. Reconstruction was analyzed as an independent risk factor for specific complications, with propensity scores used to help standardize compared patient populations. Regression analyses were performed using SPSS (version 20.0, Chicago, IL). Results A total of 13,082 patients underwent breast reconstruction; 9786 patients received prosthetic reconstruction and 3296 received autologous reconstruction. Within the autologous cohort, 1608 (48.8%) patients underwent a pedicle TRAM flap, 1079 (32.7%) had a LD flap, and 609 (18.5%) received a free flap. Autologous reconstruction patients had higher rates of overall complications (12.47% vs 5.38%, p <.001), wound infection (5.46% vs 3.45%, p <.001), prosthesis/flap failure (3.13% vs 0.85%, p <.001), and reoperation (9.59% vs 6.76%, p <.001). Risk-adjusted multivariate analysis also showed autologous reconstruction to be a significant independent predictor of specific short term outcomes. Conclusions Using risk-adjusted models of a large multi-institutional database, we found that - relative to prosthetic reconstruction - autologous reconstruction had higher rates of 30-day overall complications, wound infection, prosthesis/flap failure, and reoperation. This may be due, in part, to a concomitant increase in operative time and higher case complexity. Taken with other reports such as NMBRA, this study helps to educate patients and surgeons alike on potential, comparative complications during the perioperative period.