Sep 3 2015
By Shreeya Nanda, Senior medwireNews Reporter
A risk score incorporating routinely collected clinical information can identify patients with chronic hepatitis B virus (HBV) infection at high risk of developing acute-on-chronic liver failure (ACLF) following severe acute exacerbation, research suggests.
Among 382 chronic HBV patients with severe acute exacerbation, 52 (13.6%) developed ACLF an average 9 days after hospitalisation. And 15 died of ACLF-related causes within 56 days, report lead investigator Xian-Bo Wang, from Capital Medical University in Beijing, China, and team.
Comparison of patients with and without ACLF showed that four factors were independently associated with ACLF risk: age of 40 years or more, total bilirubin levels of at least 171 μmol/L, prothrombin activity of 40% to 60% and HBV DNA levels higher than 107 copies/mL.
To develop the risk prediction model, patients were randomly assigned to the derivation or validation cohorts in a 2:1 ratio. The researchers allocated an integer value based on the regression coefficients to each variable and summed up the risk points to give a risk score between 0 and 7, where a score of 0 to 3 signified low risk and that of 4 and above indicated high risk.
Across all study participants, the cumulative risk of ACLF and ACLF-related mortality over 56 days was significantly higher for patients with a high-risk than those with a low-risk score, with respective incidence rates of 33.8% versus 2.0% (p<0.0001) and 9.4% versus 0.8% (p<0.0001).
The model distinguished patients who developed ACLF from those who did not with a concordance index of 0.857 in the derivation cohort and 0.889 in the validation cohort. In comparison, the Model for End-Stage Liver Disease (MELD) score had a concordance index of 0.601 and 0.632 in the derivation and validation cohorts, respectively, while the corresponding values for the Child-Turcotte-Pugh (CTP) score were 0.593 and 0.646.
The model had “good discrimination” and was “superior” to the MELD and CTP scores, say the authors in the World Journal of Gastroenterology.
Highlighting that the findings need to be confirmed in larger, prospective, randomised studies, they nonetheless conclude: “This model might be useful as a tool in identifying those patients at the highest risk of ACLF and in guiding monitoring and treatment decisions.”
Licensed from medwireNews with permission from Springer Healthcare Ltd. ©Springer Healthcare Ltd. All rights reserved. Neither of these parties endorse or recommend any commercial products, services, or equipment.