Enhancing prediction of incisional hernia in liver transplant patients: performance assessment and stepwise revision of a prognostic model.

Abstract

<h4>Purpose</h4>Incisional hernia (IH) is a common complication following abdominal surgery, particularly after liver transplantation (LT). Given the impact on quality of life and associated health and financial burdens, accurate risk prediction is important for guiding preventive strategies. This study aimed to externally validate the Transplant‑Specific Penn Hernia Risk Calculator and to develop a revised prediction model tailored to LT patients.<h4>Methods</h4>A retrospective cohort study was conducted of patients ≥ 18 years who underwent LT between 2010 and 2019 at a tertiary academic centre. External validation was conducted using the Transplant-Specific Penn Hernia Risk Calculator variables and their corresponding coefficients. A revised model was developed using a multi-step methodology to optimize predictive performance. The updated model included five additional predictor variables (age ≥ 60, NYHA functional class ≥ 2, history of previous hernia, retransplantation, and polycystic liver disease) added to three original predictors (arterial hypertension, history of GI surgery, and obesity WHO class 3). Model performance was assessed for discrimination, calibration, and overall accuracy.<h4>Results</h4>Of 308 LT recipients included, 55 (17.9%) developed an IH requiring surgical repair. Analysis was restricted to 280 patients (90.9%) with ≥ 1 year follow-up. The Transplant-Specific Penn Hernia Risk Calculator demonstrated suboptimal discrimination (area under the receiver-operating-characteristic curve [AUC] = 0.649) and calibration (observed-to-expected [O/E] ratio = 5.514). The revised LT-specific model achieved superior discrimination (AUC = 0.807) and excellent calibration (O/E ratio = 1.007).<h4>Conclusion</h4>By incorporating LT-specific variables, the revised model improves predictive performance, enabling improved risk stratification. This might increase interest of using preventive mesh placement in this patient category.