Short-term Prognostic Value of ALBI, NLR, FIB-4 and MELD Score in Patients With Acute-on-chronic Liver Failure Undergoing Plasma Exchange
10.3969/j.issn.1008-7125.2020.08.003
- Author:
Ying ZHANG
1
;
Fang MING
1
;
Lixin XU
1
;
Hong XUE
1
Author Information
1. Department of Intensive Infectious Diseases, Nantong Third People's Hospital
- Publication Type:Journal Article
- Keywords:
Acute-on-Chronic Liver Failure;
ALBI;
FIB-4;
Logistic Models;
MELD;
NLR;
Prognosis
- From:
Chinese Journal of Gastroenterology
2020;25(8):462-466
- CountryChina
- Language:Chinese
-
Abstract:
Background: Acute-on-chronic liver failure (ACLF) is a rapid-developing critical illness with a high mortality. Accurate prediction of the prognosis of patients with ACLF can guide the individualized therapy, and effectively and rationally utilize the scarce liver source. Aims: To investigate the short-term prognostic value of ALBI, NLR, FIB-4, and MELD score in ACLF patients undergoing plasma exchange (PE). Methods: A retrospective analysis was conducted on clinical data of consecutive patients with ACLF undergoing PE from Jan. 2015 to Jul. 2019 at the Nantong Third People's Hospital. According to the survival status at 3 months after admission, the eligible cases were allocated into improvement group and deterioration group (including clinical deterioration, liver transplantation and death cases). The laboratory parameters before PE were recorded to calculate the score of ALBI, NLR, FIB-4, and MELD. Multivariate logistic regression analysis was performed to identify the influencing factors for prognosis and a combined prognostic model was constructed. ROC curve was used to assess the performance of single and combined score for predicting the short-term prognosis. Results: A total of 147 ACLF patients were enrolled, 71 in improvement group and 76 in deterioration group. ALBI, NLR, FIB-4, and MELD score were all significantly increased in deterioration group than in improvement group (P<0.05). Multivariate logistic regression analysis revealed that all four single score were independent risk factor for poor short-term prognosis of ACLF patients undergoing PE. Area under the ROC curve (AUC) was 0.767, 0.884, 0.750 and 0.860 for ALBI, NLR, FIB-4, and MELD, respectively. When using triple (ALBI+NLR+FIB-4) or quadruple (ALBI+NLR+FIB-4+MELD) combined score established by logistic regression model, AUC could increase to 0.918 and 0.946, respectively. Conclusions: ALBI, NLR, FIB-4, MELD score has a good value for evaluating the short-term prognosis of ACLF patients undergoing PE. Combined model including these four single score has higher predictive value.
