A decision-making model for prediction of a stable disease course in chronic hepatitis B patients

Imri Ofri1,2, Noam Peleg2,3, Moshe Leshno4

  • 1Department of Medicine D and the Laboratory of Liver Research, Rabin Medical Center, Beilinson Hospital, 39 Jabotinsky Street, Petah-Tikva, Israel.

Scientific Reports
|December 28, 2023
PubMed

Insights

Identifying predictors for a stable chronic hepatitis B (CHB) course can optimize patient monitoring. Older age, lower baseline HBV DNA viral load (VL), and normal ALT levels predict a stable CHB disease course, potentially reducing unnecessary tests and healthcare costs.

Area of Science:

  • Hepatology
  • Virology
  • Clinical Medicine

Background:

  • Chronic hepatitis B (CHB) requires regular monitoring of HBV DNA and liver enzymes to guide antiviral therapy.
  • Identifying patients with stable disease can lead to less frequent monitoring, reduced testing, and cost savings.

Purpose of the Study:

  • To identify predictors of a stable disease course in patients with CHB.
  • To enable optimized monitoring strategies for CHB patients off antiviral treatment.

Main Methods:

  • Retrospective analysis of 220 CHB patients followed between 2004-2018.
  • Defined stable CHB as low, steady viral load (<2000 IU/ml) and normal ALT (<40 IU/ml) over 6 consecutive visits.
  • Utilized stepwise multivariate logistic regression and decision tree models to identify predictors.

Main Results:

  • Older age, higher percentage of women, and lower baseline AST, ALT, and viral load (VL) were associated with stable CHB.
  • Multivariate analysis showed age (OR 0.94), baseline ALT (OR 1.06), and VL (OR 1.05) significantly predicted stability.
  • A decision tree model identified patients aged 46-67 with baseline VL <149 IU/mL and ALT <40 IU/mL had a 91% probability of a stable course.

Conclusions:

  • Integrating patient age with baseline HBV DNA viral load and ALT levels can effectively predict a stable disease course in CHB patients not on treatment.
  • These findings support personalized monitoring schedules, potentially improving patient management and resource allocation.

Related Concept Videos