AAPS - Abdominal Wall Reconstruction In Ventral Hernia Repair: Do Current Models Predict Surgical Site Risk?

Back to 2025 Resident Posters

Abdominal Wall Reconstruction In Ventral Hernia Repair: Do Current Models Predict Surgical Site Risk?
Zachary Gala, MD, Medhi S. Lemdani, BS, Dustin T. Crystal, MD, Jane N. Ewing, BS, Robyn Broach, PhD, John P. Fischer, MD, Stephen J. Kovach, MD.
University of Pennsylvania, Philadelphia, PA, USA.

PURPOSE: Complications following ventral hernia repair (VHR) pose significant healthcare burdens. Risk stratification models are incentivized to improve outcomes and cost-effectiveness. The Ventral Hernia Risk Score (VHRS) and Ventral Hernia Work Group Classification (VHWG) attempt to predict surgical site infections (SSI) and surgical site occurrences (SSO) based on patient characteristics. Our study aims to evaluate these models and assess validity.
METHODS: A 2013-2022 retrospective review of VHRs was conducted. Non-SSI and non-SSO cohorts were compared to SSI and SSO cohorts. Outcomes were evaluated relative to VHRS and VHWG scores. The Youden index of the Receiver Operating Characteristic (ROC) curves defined optimal score cutoffs for each model. Area under curve (AUC) was reported.
RESULTS: 1,414 patients underwent VHR, of which 175 (12.4%) experienced SSI and 367 (26.0%) SSO. Mean follow-up was 1.72 years. Cohort demographics were similar. However, comorbidities including prior non-VHR abdominal surgery (SSI: p<0.001; SSO: p<0.001), prior-VHR (SSI: p=0.001; SSO: p-0.012), and prior mesh infection (p=0.004) were different between non-SSI/SSO and SSI/SSO cohorts. Youden index of VHRS suggested a cutoff score of 7 for increased SSI risk (AUC 0.609) and 6 for SSO risk (AUC 0.588). VHWG grade of 3 was the cutoff for both SSI (AUC 0.616) and SSO (AUC 0.614).
CONCLUSION: Our study presents the largest external validation assessing the VHRS model. The VHRS was not superior to the VHWG for SSI/SSO prediction. The VHRS is convenient for simplicity, however our results suggest that additional factors are relevant in prognosticating SSI/SSO, which limits model predictability and generalizability.

Back to 2025 Resident Posters