A Study Of Predictive Model For Prognosis Of Hepatitis B Related Cirrhosis

Objective: To analyze the relationship between routine laboratory results and other indicators of death risk in patients with hepatitis B related liver cirrhosis to find the major predictors to create an effective mathematical model for the prognosis of liver cirrhosis after hepatitis B.Methods: The medical documents of hepatitis B related liver cirrhosis patients were reviewed retrospectively and the data were collected.Univariate and multivariate Logistic regression analyses were performed, in which the discharged results (survival or death) as dependent variables and the results of liver function, kidney function, serum electrolytes blood cell analysis,coagulation function test,type-B ultrasonic,bleeding esophageal varices, hepatic encephalopathy and ascites as independent variables were included. The multivariate regression equation was induced to the model for the prediction of patient outcome and its predictive performance was evaluated. To compare the capability of the logistic model, model for end-stage liver disease (MELD) and Child-Turcotte-Pugh (CTP) in predicting the prognosis of patients using receiver operating characteristic (ROC) curve.Results: In univariate regression, the significant positive variables for death outcome were S/L,â–³PA,TBIL,DBIL,CRE,K,WBC,GRAN,INR,â–³APTT,APT,spleen thickness,upper gastrointestinal bleeding and hepatic encephalopathy, and the negative variables were oblique diameter of right hepatic. The variables entered the multivariate regression included S/L,â–³PA,DBIL,Cr,K,INR,â–³APTT,blique diameter of right hepatic,spleen thickness,upper gastrointestinal bleeding and hepatic encephalopathy. The logistic equation for hepatitis B related liver cirrhosis was that Logit(Y)= 0.808(S/L)+3.951(â–³PA)+ 0.175 (DBIL)+ 1.789(CRE)+ 0.981 (K)+ 2.760 (INR)+ 0.070 (â–³APTT)-0.047 (blique diameter of right hepatic)+ 0.069 (spleen thickness)+ 2.012 (upper gastrointestinal bleeding, 1 bleeding,0 without bleeding)+ 1.837 (1 hepatic encephalopathy, 0 wihtout hepatic encephalopathy) -5.254. The logistic model was (P)=1 /[1+ e^-Logit(Y)].The sensitivity, specificity and accuracy of the regression model for predicting death of hepatitis B related liver cirrhosis patients were 91.2%,93.3% and 93.2%,and the values of area under the curve (AUC) were 0.957,respectively, higher than those as CTP and MELD(0.730 and 0.844).Conclusions : The logistic equation for hepatitis B related liver cirrhosis is that Logit(Y)= 0.808(S/L)+3.951(â–³PA)+ 0.175 (DBIL)+ 1.789(CRE)+ 0.981 (K)+ 2.760 (INR)+ 0.070 (â–³APTT)-0.047 (blique diameter of right hepatic)+ 0.069 (spleen thickness)+ 2.012 (upper gastrointestinal bleeding, 1 bleeding,0 without bleeding)+ 1.837 (1 hepatic encephalopathy, 0 wihtout hepatic encephalopathy) -5.254. The logistic model is (P)=1 /[1+ e^-Logit(Y)].The sensitivity, specificity and accuracy of the regression model for predicting death of hepatitis B related liver cirrhosis patients are 91.2%,93.3% and 93.2%.