A nomogram for predicting outcomes following pedicled flap reconstruction of the lower extremity

Introduction: Lower extremity reconstructions with soft tissue flaps account for a significant proportion of reconstructive surgery. Pedicled flaps are a procedure of choice, particularly in multimorbid patients and those with small to medium-sized defects. Complication rates are high and should not be underestimated, with accurate preoperative risk factor assessment being imperative. Methods: Using the American College of Surgeons National Surgical Quality Improvement Program Database (2008-2021), we analyzed all patients who underwent pedicled flap reconstruction of the lower extremity. Demographic data, comorbidities, and perioperative data were extracted. We identified risk factors of major surgical complications within 30 days of surgery through uni- and multivariate regression analyses. On this basis, a nomogram for predicting the risk of complications was developed to allow for feasible point-of-care risk assessment. Results: A total of 6475 adult patients were identified. The most common comorbidities were hypertension (n = 3363, 51.9%) and obesity (n = 2306, 35.6%) and most common American Society of Anesthesiologists (ASA) class was class 3 (n = 3703, 57.2%). The most common complication was bleeding/transfusion (n = 1293, 19.9% ). Multivariate regression revealed longer procedures, inpatient setting, higher ASA scores, hypertension, and procedures per- formed by orthopedic and vascular surgeons as predictors for higher risk of post-operative complications. Conclusion: Leveraging a risk- and case-mix-adjusted multi-institutional database, a nomogram for post-operative surgical complications within 30 days after pedicled flap reconstruction in the lower extremity was developed. This broadly applicable risk prediction tool can aid in decision-making when assessing patient eligibility. (c) 2024 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.