T. M. Barry1, H. Janjua1, E. Cousin-Peterson1, F. Agcaoili1, A. Mooney1, M. Ottinger1, P. C. Kuo1 1University Of South Florida College Of Medicine,Surgery,Tampa, FL, USA
Intro:Prior studies comparing patients cared for by male vs. female internists showed decreased 30-day mortality rates among female physicians; however there have been no studies evaluating the effect of surgeon gender on post-operative mortality in the United States.
Methods: A retrospective analysis was performed on hospital admissions using the Agency for Health Care Administration (AHCA) Florida database from 2010 to 2015 examining patients undergoing 1 of 25 surgical procedures. Surgeon gender, number of Medicare beneficiaries (surrogate for practice volume) and years of experience were imported from CMS Physician Compare/Provider Utilization & Payment Data set and linked on the basis of NPI. 25 procedures were selected using the modified Johns Hopkins Surgical Criteria, including common and complex procedures from all surgical subspecialties. The primary outcome was in-hospital death. For each procedure, admissions were propensity matched on the basis of surgeon gender via 1:1 optimal matching comprising admission priority, patient characteristics (age, sex, race, ethnicity, CCI), surgeon years of experience. Z test, Chi square and Students t-test were used to compare practice volume, total charges and length of stay(LOS). Backwards stepwise logistic regression was performed on matched data for 3 procedures of interest to predict mortality.
Results:There were 86,954 admissions for patients undergoing surgical procedures performed by 2737 surgeons(366 female, 2371 male). Fewer patients who had elective procedures performed by female surgeons died in the hospital(95/24846;0.38%) than those who had male surgeons(141/24624;0.57% P=0.002). Propensity matched populations ranged from 21 to 10261 patients per group. Those who underwent the following had significantly lower rates of in-hospital mortality if the surgeon was female: bypass/sleeve gastrectomy(2/873,0.23% vs 10/865,1.14%P=0.020), mastectomy(4/6616,0.06% vs 12/6608, 0.18%P=0.045), open cholecystectomy(43/1553,2.69% vs 63/1533, 3.95%P=0.048). Male surgeons did not have a mortality advantage for any procedure. AUC for the predictive model was 0.948 (95%CI 0.9197-0.9762). Odds ratio for surgeon gender was 1.795 demonstrating higher likelihood of inpatient mortality if the surgeon is male (β coef 0.58 P=0.0084). There was no significant difference in total charges, practice volume or LOS.
Conclusion:Patients who underwent bariatric procedures, mastectomy and open cholecystectomy had lower rates of in-hospital mortality if the surgeon was female while male surgeons did not have a significant mortality advantage for any of the 25 procedures. Further studies examining national data may provide additional insight regarding the effect of surgeon gender on patient outcomes.
