Font Size: a A A

Comparative Study Of Seven Scoring Systems To Predict Mortality Of Patients With Hepatitis B Virus-related Acute-on-chronic Liver Failure

Posted on:2019-12-05Degree:DoctorType:Dissertation
Country:ChinaCandidate:D J ZhangFull Text:PDF
GTID:1364330548489890Subject:Eight-year clinical medicine
Abstract/Summary:PDF Full Text Request
BackgroundHepatitis B virus-related acute-on-chronic liver failure(HBV-ACLF)which is characterized by rapid deterioration of liver function and multisystem organ failure is the most common type of liver failure in China.There is still no ideal therapy available for ACLF untill now.Early diagnosis,accurate evaluation and timely intervention can greatly improve the outcome of patients with ACLF.However,due to the complex and multifactorial pathogenesis of ACLF,together with different clinical manifestations,intervention measures and the response to treatment in each patient,the prognosis of ACLF is quite difficult.Most of the commonly used scoring models are based on the research data of western populations,which may have low predictive value in assessing the prognosis of patients with HBV-ACLF.The prognostic significance of the newly established scoring model remains to be further verified.ObjectivesThis study aimed to analyze the indepatent risk factors of short-term prognosis in patients with HBV-ACLF,compare the predictive value of MELD,MELD-Na,iMELD,CLIF-SOFA,CLIF-C ACLF,LRM,COSSH-ACLF scoring models,and explore whether dynamic changes of these scoring models have higher prognostic value than that evaluated at the diagnosis of ACLF in an attempt to guide the therapeutic options and organ allocation.MethodsA total of 201 patients with HBV-ACLF diagnosed in Nanfang Hospital of Southern Medical University from September 2013 to July 2017 were included and divided into two groups according to the prognosis in 12 weeks in our study.The SPSS software version 22 was used to compare the clinical data between survival group and death group.Area under the receiver operating curve(AUC)was calculated to evaluate the prognostic value of the scoring models in predicting short-term prognosis of HBV-ACLF patients.Results1.A total of 201 patients with HBV-ACLF was enrolled in the study,including 181(90%)males and 20(10%)females.84 patients died in three months and the mortality rate was 41.8%?2.Logistic regression analysis showed that age,international normalized ratio(INR),Hepatorenal syndrome(HRS),hepatic encephalopathy(HE)and upper gastrointestinal bleeding were the indepatent risk factors of short-term prognosis in patients with HBV-ACLF.3.The mortality rates of HBV-ACLF patients at early,middle and late stage were 7.2%,30.4%,and 92.8%,respectively.The difference was statistically significant.Scoring models in early stage patients were obviously lower than that in the middle and late stage patients.And scores in middle stage patients were lower than that in the late stage patients.4.The seven scoring models of the survival group were significantly lower than that of the death group.The modality rate was positively associated with the value of these scoring models.The AUC values for MELD,MELD-Na,iMELD,CLIF-SOFA,CLIF-C ACLF,LRM and COSSH-ACLF scoring models were 0.822,0.823,0.831,0.787,0.813,0.762 and 0.846,respectively.The prognostic value of these scoring models at the third,seventh and fourteenth day of enrollment was higher than that at baseline.ConclusionThe prognostic of HBV-ACLF was very poor with a modality rate of 92.8%for patients in the late stage.Age,international normalized ratio(INR),Hepatorenal syndrome(HRS),hepatic encephalopathy(HE)and upper gastrointestinal bleeding were the indepatent risk factors of short-term prognosis in patients with HBV-ACLF.All of the seven scoring models can be applied to evaluate the prognosis of HBV-ACLF,and the predictive value of COSSH-ACLFs is higher than others.Sequential assessment of prognosis can significantly improve the predictive accuracy of the scoring models.
Keywords/Search Tags:acute-on-chronic liver failure, Hepatitis B, prognosis, scoring model
PDF Full Text Request
Related items