Font Size: a A A

Create A Predictive Model For Significant Liver Inflammation Or Fibrosis For The Patients With Chronic Viral Hepatitis B Based On Routine Serological Indexes

Posted on:2022-05-18Degree:MasterType:Thesis
Country:ChinaCandidate:K LiangFull Text:PDF
GTID:2544306602450544Subject:Internal medicine
Abstract/Summary:PDF Full Text Request
Objective To establish a non-invasive prediction model of significant liver inflammation or fibrosis(grade of liver histological inflammatory activity(G)≥2 or stage of liver fibrosis(S)≥2)for the chronic hepatitis B patients based on routine serological indexes,guiding their treatment.Methods A total of 149 patients diagnosed with chronic viral hepatitis B who had undergone liver biopsy in The First Affiliated Hospital of Guangxi Medical University from January 2015 to March 2021 were Included in the study,including 118 patients in the model group and 31 patients in the validation group.The general information,laboratory tests and pathological results of liver biopsy were collected.The patients were divided into groups according to the results of liver biopsy:those with G≥2 or S≥2 were included in the group of significant inflammation or liver fibrosis and those with G≤1 and S≤1 were included in the group of no significant inflammation or liver fibrosis.Univariate analysis and binary Logistic regression analysis were used to construct predictive models of significant liver inflammation or fibrosis.The ROC curve was calculated to evaluate the prediction performance of our prediction model,APRI and FIB-4.The prediction model was verified and evaluated with validation group Finally.Results In the modeling group,there were 64 patients(54.2%)in with significant inflammation or fibrosis,and 54 patients(45.8%)in the group without significant inflammation or fibrosis.Six indicators were included in the prediction model,including HBV DNA>5×10~2 IU/m L,PLT,ALB,AST,ALT and Cys C.The area under the ROC curve of the prediction model was 0.824,the optimal cut-off value was 0.482,the corresponding sensitivity was 0.813,and the specificity was 0.741,and Its area under ROC curve is larger than APRI and FIB-4.In the validation group,the positive predictive value was 0.840(21/25)and the negative predictive value was 0.833(5/6).Conclusion In this study,a non-invasive predictive model of significant liver inflammation or fibrosis was established for patients with chronic viral hepatitis B.The model has certain value in predicting significant liver inflammation or fibrosis,and can be used to guide the treatment of patients with chronic viral hepatitis B after evaluation and verification.So It is worth to promote.
Keywords/Search Tags:chronic hepatitis B, liver inflammation, liver fibrosis, Logitsic regression, non-invasive predictive model
PDF Full Text Request
Related items