Development of an Integer-based Risk Score to Predict 90-Day Mortality After Hepatectomy in Patients With Hepatocellular Carcinoma.

BACKGROUND

The incidence of liver cancer has more than tripled since 1980. Hepatectomy represents the major curative treatment for liver cancer. The risk factors associated with 90-day mortality after hepatectomy are not well understood and there are currently no good prediction models for this outcome. The objectives of the current study were to identify risk factors of 90-day mortality after hepatectomy in patients with hepatocellular carcinoma and to develop an integer-based risk score using the National Cancer Database.

METHODS

Hepatectomies recorded in the National Cancer Database during 2004-2012 were reviewed for 90-day mortality. Risk factors were identified by multivariate logistic regression models. An integer-based risk score was developed using the β coefficients derived from the logistic regression model and tested for discriminatory ability. According to the total risk score, patients were grouped into 4 risk groups.

RESULTS

The overall 90-day mortality was 10.2%. Ten risk factors were identified, which included sex, age, race/ethnicity, insurance status, education, annual hospital volume, stage, tumor grade, Charlson-Deyo Score, and surgical procedure. The risk of 90-day mortality was stratified into 4 groups. The calculated 90-day mortality rates were 2.47%, 5.88%, 12.58%, and 24.67% for low-risk, medium-risk, high-risk, and excessive-risk groups, respectively. An area under the receiver operating characteristic curve of 0.69 was obtained for model discrimination.

CONCLUSIONS

The integer-based risk score we developed could easily quantify each patient's risk level and predict 90-day mortality after hepatectomy. The stratified risk score could be a useful addition to perioperative risk management and a tool to improve 90-day mortality after hepatectomy.