Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Aug;39(8):5035-5045.
doi: 10.1007/s00464-025-11927-7. Epub 2025 Jul 7.

Fairness of machine learning readmission predictions following open ventral hernia repair

Tyler Zander  1   2 , Melissa A Kendall  3   4 , Rachel L Wolansky  3   4 , Emily A Grimsley  3   4 , Rajavi Parikh  3   4 , Joseph Sujka  3 , Paul C Kuo  3   5
Affiliations

Fairness of machine learning readmission predictions following open ventral hernia repair

Tyler Zander et al. Surg Endosc. 2025 Aug.

Abstract

Introduction: Few models have predicted readmission following open ventral hernia repair (VHR), and none have assessed fairness. Fairness evaluation assesses whether predictive performance is similar across demographic groups, ensuring that biases are not propagated. Therefore, we generated an interpretable machine learning model to predict readmission following open VHR while assessing fairness.

Methods: NSQIP (2018-2021) was queried for open VHR. We developed an XGBoost model to predict unplanned readmissions within 30 days of surgery with fivefold cross-validation. Performance and fairness were assessed by demographic groups: gender (female vs. male), ethnicity (Hispanic vs. non-Hispanic), and race (non-White vs. White). We identified influential features within demographic groups using SHapley Additive exPlanations (SHAP).

Results: 59,482 patients were included with a readmission rate of 5.5%. The model had an AUC of 0.72 and a Brier score of 0.16. Fairness metrics revealed minimal performance differences between demographic groups. SHAP revealed that influential factors were similar across demographic groups and included days from operation to discharge, morbidity probability, and operative time.

Conclusion: Using interpretable machine learning, we identified unique predictors for unplanned readmission following open VHR. Fairness metrics revealed minimal differences in performance between demographic groups. SHAP showed similar influential factors across demographic groups. Future surgical machine learning models should similarly assess models using fairness metrics and interpretation of predictions.

Keywords: Algorithmic bias; Fairness; Interpretable machine learning; Risk prediction.

PubMed Disclaimer

Conflict of interest statement

Declarations. Disclosures: Dr. Joseph Sujka is a consultant for Intuitive, Medtronic, and Enterra Medical. Drs. Tyler Zander, Melissa Kendall, Rachel Wolansky, Emily Grimsley, Rajavi Parikh, and Paul Kuo have no conflicts of interest or financial ties to disclose.

Figures

Fig. 1
Fig. 1
Cohort selection. *Any procedures except initial or recurrent open ventral hernia repairs for reducible or incarcerated hernias with or without mesh, exploratory laparotomy, enterolysis, omentectomy, enterorrhaphy, closure enterostomy, enterectomy, panniculectomy, or component separation. NSQIP National Surgical Quality Improvement Program, VHR ventral hernia repair, AMA against medical advice, LOS length of stay
See this image and copyright information in PMC

References

    1. Mehtsun WT, Papanicolas I, Zheng J, Orav EJ, Lillemoe KD, Jha AK (2018) National trends in readmission following inpatient surgery in the hospital readmissions reduction program era. Ann Surg 267(4):599–605 - PubMed
    1. Feimster JW, Whitehurst BD, Reid AJ, Scaife S, Mellinger JD (2022) Association of socioeconomic status with 30- and 90-day readmission following open and laparoscopic hernia repair: a nationwide readmissions database analysis. Surg Endosc 36(7):5424–5430 - PubMed
    1. Nelson JA, Fischer J, Chung CC, Wink J, Wes A, Serletti JM, Kovach S (2015) Readmission following ventral hernia repair: a model derived from the ACS-NSQIP datasets. Hernia 19(1):125–133 - PubMed
    1. Baltodano PA, Webb-Vargas Y, Soares KC, Hicks CW, Cooney CM, Cornell P, Burce KK, Pawlik TM, Eckhauser FE (2016) A validated, risk assessment tool for predicting readmission after open ventral hernia repair. Hernia 20(1):119–129 - PubMed
    1. Nguyen MT, Li LT, Hicks SC, Davila JA, Suliburk JW, Leong M, Kao LS, Berger DH, Liang MK (2013) Readmission following open ventral hernia repair: incidence, indications, and predictors. Am J Surg 206(6):942–948 - PubMed

LinkOut - more resources