Application of the Balance of Risk Score to Liver Transplantation

Abstract

BackgroundTo evaluate the capacity of BARS to predict 90-day, 1-year, 3-year, and 5-year survival after liver transplantation (LTx). The Balance of Risk Score (BARS), proposed by the Swiss Hepato-pancreato-biliary and Transplant Center, University Hospital Zurich, Switzerland, was conceptualized to determine survival after LTx preemptively. Sample and MethodsThis was a retrospective observational study on 866 cases of LTx among adults (>18 years) performed within the transplantation program at Hospital Israelita Albert Einstein between January 1, 2010, and December 31, 2015. Cases of loss from follow-up, split-liver transplantation, transplantation from a live donor, and combinations of LTx with transplantation of other organs were excluded. BARS was calculated for each transplantation performed. The survival estimates were accompanied by 95% CIs, and the associations between the variables of interest and the patients’ overall survival were evaluated using Cox proportional risk models. Receiver operating characteristic curves and the estimated area under the curve were used with 95% CIs and P values for the equality test on the area at .5. ResultsThe area under the curve for the 90-day period was 0.725, with a 95% CI from 0.670 to 0.81 and a P value < .001 for the equality test at 0.5. In the stratified analysis, the score of 18 presented the highest sensitivity, taking a minimum specificity of 90%. The BARS >18 gave rise to a significant decline in survival, from 89.7% to 60.4% over the first 90 days, from 83.3% to 56.2% over a 1-year period, from 75.7% to 49.5% over 3 years, and from 72.2% to 43.4% over 5 years. ConclusionThe BARS was shown to be reproducible and can be used as a tool for estimating survival among LTx patients. LTx performed on patients with BARS >18 significantly predicts lower survival for such patients.