(033) Analyzing Patient-Specific Predictors of Surgical Risk and Outcomes Following Gender-Affirming Surgery

Abstract Introduction Gender affirming surgery (GAS) has seen dramatic advances in phalloplasty and vaginoplasty for gender dysphoria. However, there is still need better characterization of patient demographics surgical outcomes. Recent interest emerged predictors post-surgical complications, such as frailty comorbidities preoperative risk assessment. Better understanding factors predicting outcomes can help with selection, counseling, quality improvement. Objective To evaluate complications utilizing the National Surgical Quality Improvement Program (NSQIP) database. Methods CPT codes were utilized to identify (59980) (59970, 57335) cases recorded NSQIP from 2006-2020. Descriptive statistics calculated baseline characteristics, operative parameters, 30-day complications. Frailty was using modified index (NSQIP-mFI-5), a previously validated 5-item score including points diabetes, impaired functional status, COPD, CHF, hypertension. Pre-operative morbidity probabilities derived “MORBPROB” variable, which probability, uses hierarchical regression analysis on patient-specific factors. Univariate logistic regression, ANOVA, t-test analyses performed appropriate time, hospital length stay (LOS). Results Out 9,637,305 patients database, 58 underwent 485 vaginoplasty. For vaginoplasty, mean age 38.4±11.8 36.6±12.6 years, respectively. The average time 331 minutes 263 respectively, LOS 5.1 days 4.3 days, overall complication rate 26%, 17% all experiencing minor urinary tract infections, superficial wound pneumonia, peri-operative bleeding requiring transfusion, 16% major occurrences sepsis, DVT, stroke, reintubation, renal failure, myocardial infarction, pulmonary embolism, septic shock, would dehiscence, deep cardiac arrest. overall, minor, rates vaginoplasties 14%, 7%, 9%. Unplanned readmissions reoperations each occurred 7% 5% patients. In either cohort, no deaths within 30 post-operation. Upon NSQIP-mFI-5 scores ≥0.2 versus 0 not predictive or procedure. Notably, vaginoplasties, higher numerically associated (OR 2.02, 95% CI 0.94-4.09, p=0.072). When comparing ≥10% <10%, this both 4.0, 1.1-19.6, p=0.038) 2.46, 1.4-4.3, p<0.001). probability also extended (6.3±1.3 vs. 2.9±0.8, p=0.03). Conclusions This study described database are largely aligned published studies. an effective predictor than index. Disclosure Any authors act consultant, employee shareholder industry for: Paul Chung - Consultant Boston Scientific Coloplast.