Comparative Analysis of Reintervention Rates in Mesh Versus No Mesh Hernia Repair Using Electronic Health Records

OBJECTIVES: Abdominal hernia repair exhibits varying rates of reintervention based on device and procedure characteristics. The granularity of this information, however, is often lacking in conventional data sources such as claims. This study delves into the impact of mesh usage and procedural and device characteristics on hernia repair reinterventions, leveraging unstructured clinical notes from electronic health records (EHR).

METHODS: This retrospective cohort study utilized data from the Arkansas Clinical Data Repository, comprising of patients undergoing hernia repair as the index procedure between October 2015 and September 2023. Reintervention, defined as repair for recurrent hernia more than 7 days post the index procedure, served as the primary outcome. Procedure and device-related characteristics were extracted from unstructured clinical notes. Two logistic regression models were employed: the first accounting for structured EHR data only and the second incorporating both structured and extracted unstructured data.

RESULTS: Among 1838 patients undergoing hernia repair, 1169 (63.40%) received mesh. Of the total, 71 (51.08%) with mesh and 68 (48.92%) without experienced reinterventions. Variables from unstructured data, included item name, anesthesia type, and surgical approach. Two logistic regressions were run. The first logistic regression controlled only for structured data (age, gender, comorbidity score) and showed 1.603 (95% Confidence Interval (CI):1.116-2.303) higher odds of reintervention with mesh compared to no mesh. The second model incorporated the extracted unstructured data with the structured data and indicated an odds ratio of 1.519 (95%CI:1.504-2.189), with the unstructured variables emerging as significantly influential factors.

CONCLUSIONS: Patients receiving mesh during their hernia repair exhibited significantly higher odds of reintervention in both models. The study highlights the utility of EHR data in providing detailed insights into device and procedure-related characteristics. More nuanced investigations and advanced methodologies with larger sample sizes are needed to further investigate the role of these characteristics on reintervention rates.