The Predicting Value Of PTAR,Child-pugh And MELD Score On The Occurrence Of Acute-on-chronic Liver Failure In The Patients With Cirrhosis

Objective:With the development of liver cirrhosis,acute-on-chronic liver failure(ACLF)can appear.Because of its characteristics,such as acute onset,poor prognosis and high mortality rate,it is significant to find a marking system that can predict the progression of ACLF in patients with cirrhosis.This study aims to find marking systems to predict the progression of ACLF more accurately by using PTAR score,Child-pugh score,MELD score and so on,which fit HBV-related cirrhosis and alcoholic cirrhosis.With the assistance of such marking systems,we can not only extract the high risk patients of ACLF,but also guide physicians adopt effective treatment programs to delay even avoid the occurrence of ACLF.Methods:A retrospective study is conducted on patients with HBV-related cirrhosis and alcoholic cirrhosis who were hospitalized in our Hospital from September 2016 to September 2019.According to the exclusion criteria,excluded the cases that do not meet the requirements,and finally included 261 patients with liver cirrhosis,including 148 patients with HBV-related hepatitis cirrhosis and 113 patients with alcoholic cirrhosis.According to the follow-up records 24 cases of ACLF are diagnosed,among which 13 cases are HBV-related and 11 cases are alcohol-related.HBV-related cirrhosis group is divided into A1 group and A2 group according to whether progressing to ACLF,if one progresses to ACLF,we classify him into A1 group.Alcoholic cirrhosis is divided into B1 and B2group in the same way.Indicators are collected including platelet,albumin,total bilirubin and so on in the first day of hospitalization,furthermore,ascites,hepatic encephalopathy,esophageal and gastric varices or not are collected.PTAR score,Chlid-pugh score,MELD score and so on are calculated respectively according to above indexes.The comparison of normal distribution data between groups use T test analysis,skewed distribution data using Mann Whitney U test analysis,count data using chi-square analysis.Using the single factor and multiple factors Logistic analysis and establish the receiver-operating characteristic curve to analyze the predictive value of ACLF-occurrence.Results:1.General data:In HBV-related cirrhosis,factors such as albumin,serum Na~+,Bilirubin,serum creatinine,urea nitrogen,PT,APTT,INR,and whether ascites have statistically significant differences between A1 group and A2 group(P<0.05).Scoring systems:Child-pugh,MELD,MELD-Na,ALBI and PTAR score are higher in A1 group than in A2 group,and the results are statistically different(P<0.05).In alcoholic cirrhosis,factors such as albumin,serum Na~+,Bilirubin,AST,serum creatinine,PT,APTT,INR,and whether hepatic encephalopathy have statistically significant differences between B1 group and B2 group(P<0.05).Scoring systems:Child-pugh,MELD,MELD-Na,ALBI and PTAR scored higher in B1group than in B2 group,and the results are statistically different(P<0.05).2.Univariate and multivariate regression analysis:Logistic univariate regression analysis of HBC-related group shows that Child-pugh,MELD,PTAR and whether ascites have statistical differences in the incidence of ACLF(P<0.05).Logistic multivariate regression analysis shows that only Child-pugh score and PTAR score have statistical differences in the incidence of ACLF(P<0.05).Child-pugh,MELD,PTAR and whether hepatic encephalopathy are the indicators with statistical difference(P<0.05)in affecting the occurrence of ACLF in the alcohol group.Logistic multivariate regression analysis shows that only PTAR and MELD scores are found to have statistical differences in the incidence of ACLF(P<0.05).3.ROC curve:In the HBV-related group,the Child-pugh score,PTAR score and PTAR-CP score are developed into ROC curve,and the AUC of the combined model is 0.805,larger than PTAR(0.755)and Child-pugh’s(0.691),and the optimal critical value of the combined model is-3.93.In the alcohol group,the AUC of the combined model is 0.773,larger than PTAR(0.744)and MELD’s(0.715),and the optimal critical value of the combined model is-2.15.Conclusion:1.HBV-related cirrhosis:The combined score of PTAR and Child-pugh(PTAR-CP)is more accurate in predicting the occurrence of ACLF than both single.When the PTAR-CP joint model scores>-3.93 is a strong indicator for the progressing of ACLF.2.Alcoholic cirrhosis:The combined model of PTAR and MELD(PTAR-MELD)has better predictive value for ACLF occurrence than both single.When the PTAR-MELD joint model scores>-2.15 is a strong indicator for the progressing of ACLF.3.ABIC and ALBI are of limited value in predicting the progression of ACLF in patients with cirrhosis.