Assessing surgical risk calculators on hernia repair candidates with cirrhosis

S Y Xu1, J Jackson2, M I Goldblatt3

  • 1Division of Minimally Invasive and Gastrointestinal Surgery, Department of Surgery, Medical College of Wisconsin, 8701 W Watertown Plank Rd, Wauwatosa, WI, 53226, USA.

Insights

The NSQIP Surgical Risk Calculator best predicts mortality for hernia repair in cirrhosis patients. Mayo Clinic

Area of Science:

  • Hepatology
  • Surgical Outcomes Research

Background:

  • Risk calculators aid surgeons in estimating mortality for hernia repair in patients with severe liver disease.
  • Accurate prediction of surgical risk is crucial for patient management and decision-making.

Purpose of the Study:

  • To evaluate the accuracy of existing risk calculators for predicting mortality in patients with cirrhosis undergoing hernia repair.
  • To identify the optimal patient population for utilizing these risk assessment tools.

Main Methods:

  • Utilized American College of Surgeons National Surgery Quality Improvement Program (NSQIP) datasets (2013-2021).
  • Assessed four risk calculators: Mayo Clinic's 'Post-operative Mortality Risk in Patients with Cirrhosis,' Model for End-Stage Liver Disease (MELD) score, NSQIP Surgical Risk Calculator, and a modified 5-item frailty index.
  • Employed Receiver Operating Characteristic (ROC) curve analysis to determine predictive accuracy.

Main Results:

  • The NSQIP Surgical Risk Calculator demonstrated the highest accuracy (AUC = 0.803, p < 0.001).
  • Mayo Clinic's calculator showed good performance for specific etiologies (AUC = 0.722, p < 0.001), as did the MELD score (AUC = 0.709, p < 0.001).
  • The modified frailty index had lower predictive accuracy (AUC = 0.583, p = 0.04).

Conclusions:

  • The NSQIP Surgical Risk Calculator is superior for predicting 30-day mortality in patients with ascites undergoing hernia repair.
  • In cases where NSQIP calculator data is incomplete, Mayo Clinic's calculator is recommended over the MELD score for risk assessment.
Abstract