Prognostic Value of Acute-On-Chronic Liver Failure (ACLF) Score in Critically Ill Patients with Cirrhosis and ACLF
Xinran Lin, Xielin Huang, Li Wang, Shuyi Feng, Xiaofu Chen, Weimin Cai, Zhiming Huang
Department of Gastroenterology and Hepatology, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, Zhejiang, China (mainland)
Med Sci Monit 2020; 26:e926574
Available online: 2020-08-05
In the intensive care unit (ICU), critically ill patients with cirrhosis and acute-on-chronic liver failure (ACLF) continue to have high mortality rates. The AARC ACLF score is a simple, newly-developed score based on Asian ACLF patients, which performs well in prognosis. The present study attempted to verify the prognostic ability of AARC ACLF in non-Asian critically ill patients with cirrhosis and ACLF.
MATERIAL AND METHODS: We enrolled 786 patients. Relevant clinical data were collected within 24 h after admission to compare the differences between survivors and non-survivors, and all the patients were followed up for at least 180 days.
RESULTS: The 28-day, 90-day, and 180-day mortality rates were 28.9% (227/786), 36.4% (286/786), and 40.3% (317/786), respectively. Multivariate Cox regression analysis showed that AARC ACLF score (HR: 1.375, 95% CI: 1.247-1.516, P<0.001) was an independent predictive factor of 28-day mortality, and the AUROC of the predictive ability in 28-day mortality of the AARC ACLF score was 0.754. In addition, the AARC ACLF score was regraded into 3 classes (low risk: AARC ACLF <9, intermediate risk: 9≤ AARC ACLF <12, and high risk: AARC ACLF ≥12). The AARC ACLF score can be used for dynamic assessment by retest at days 4-7.
CONCLUSIONS: The AARC ACLF score has a good predictive value for 28-day, 90-day, and 180-day mortality in non-Asian critically ill patients with cirrhosis and ACLF, which is not inferior to CLIF-C ACLFsLact and other models. It is easy to use at bedside, and it is dynamic and reliable.
Keywords: End stage liver disease, Liver Cirrhosis, Liver failure, Proportional Hazards Models