Long-term prognosis varies widely among patients with hepatitis B virus (HBV)-related liver cirrhosis. Our study aimed to investigate the applicability of albumin-bilirubin (ALBI), Child-Pugh and model for end-stage liver disease (MELD) scores to the long-term prognosis prediction of HBV-related cirrhosis. Patients diagnosed with HBV-associated cirrhosis from the First Affiliated Hospital of Wenzhou Medical University between January 2010 and December 2015 were enrolled in this study. The patients were followed up every 3 months. The prognostic performance of ALBI in long-term outcome prediction for HBV-related cirrhosis was compared with Child-Pugh and MELD scores using time-dependent receiver operating characteristic curve (tdROC) and decision curve analysis. A total of 806 patients were included in our study with 275 (34.1%) deceased during the follow-up. Multivariate Cox regression analysis showed that ALBI grade was an independent predictor associated with mortality. The tdROC analysis showed that ALBI score (0.787, 0.830 and 0.833) was superior to MELD (0.693, P=.003; 0.717, P<.001; 0.744, P<.001) and Child-Pugh score (0.641, P<.001; 0.649, P<.001; 0.657, P<.001) for predicting 1-year, 2-year and 3-year mortality. Additionally, decision curves also got the similar results. In addition, patients with lower ALBI score had a longer life expectancy, even among patients within the same Child-Pugh class. Thus, ALBI score was effective in predicting the long-term prognosis for patients with HBV-related cirrhosis and more accurate than Child-Pugh and MELD scores.
Keywords: albumin-bilirubin; cirrhosis; decision curves analysis; hepatitis B; time-dependent receiver operating characteristics.
© 2016 John Wiley & Sons Ltd.