Developing and validating a preoperative risk score to predict 90-day mortality after liver resection

Area of Science:

• Hepatobiliary surgery
• Surgical oncology
• Patient outcomes research

Background:

• Liver resection is increasingly performed with combined surgical, interventional, and chemotherapy approaches.
• Improved perioperative care enables higher-risk patient selection for hepatectomy.
• Despite improvements, 90-day postoperative mortality after liver resection remains variable.

Purpose of the Study:

• To develop and validate a preoperative risk score for predicting 90-day mortality following liver resection (hepatectomy).

Main Methods:

• Retrospective analysis of hepatectomy patients from 2011-2016.
• Univariable and multivariable analyses identified mortality predictors.
• A risk score was derived and validated in separate patient cohorts.

Main Results:

• The 90-day mortality rate was 4.0% in the derivation cohort (n=1269).
• Predictors of mortality included older patient age, extensive resection, diabetes mellitus, and low preoperative sodium.
• The risk score demonstrated good predictive accuracy (AUROC 0.778 in derivation, 0.703 in validation).

Conclusions:

• A validated preoperative risk score can predict 90-day mortality after liver resection.
• This tool can assist in patient counseling, preoperative optimization, and risk-adjusted outcome evaluation.