ABSTRACT
<p><b>OBJECTIVE</b>To investigate the short-term prognostic value of the indocyanine green clearance test when used in combination with the model for end-stage liver disease (MELD) scoring system to assess patients with hepatitis B virus acute-on-chronic liver failure (HBV-ACLF).</p><p><b>METHODS</b>Clinical data of 105 patients diagnosed with HBV-ACLF were retrospectively analyzed. The indocyanine green retention rate at 15 minutes (ICGR15), clinical data within 24 h after diagnosis, Child-Turcotte-Pugh (CTP) classification, MELD score, MELD combined with sodium concentration (MELD-Na) score, and King's Hospital (KCH) criteria data were collected for analysis. Measurement data were assessed by t-test and count data by the chi-square test. Short-term predictive accuracy for patients with HBV-ACLF was compared between different models using the area under the receiver operating characteristic (ROC) curve (AUC).</p><p><b>RESULTS</b>The mortality rate for all patients was 45.71%. Comparison of the survivors versus the non-survivors showed that age, total bilirubin, albumin, cholinesterase, creatinine, international normalized ratio, and incidence positive rate of relative complications (hepatorenal syndrome, hepatic encephalopathy) were significantly different between the two groups (all, P less than 0.05). The ICGR15 was found to be positively correlated with MELD score (r = 0.205, P less than 0.05). The MELD-ICGR15 model constructed by logistic regression analysis was: Logit(P) = 0.193 * MELD + 0.130 * ICGR15 - 11.256. The AUC was 0.880 and the cut-off was -0.706, with 89.6% sensitivity and 75.4% specificity. The AUC of the MELD-ICGR15 model was significantly higher than that of the ICGR15 (0.820), MELD score (0.779), MELD-Na score (0.761), KCH criteria (0.680), and CTP classification (0.631) (all, P less than 0.05).</p><p><b>CONCLUSION</b>ICGR15, MELD score, and MELD-Na score had higher predictive values for HBV-ACLF than did CTP classification or KCH criteria. Furthermore, the MELD-ICGR15 model was better than any single parameter model for predicting the short-term prognosis of patients with HBV-ACLF.</p>