COMPARATIVE STUDY IN EMPIRICAL CALIBRATION OF A HEPATITIS C VIRUS (HCV) NATURAL HISTORY MODEL IN HCV-INFECTED CHINESE POPULATION WITH THE UNITED STATES POPULATION

Purpose:

   In China, hepatitis C virus (HCV) infection represents a major public health challenge, with 10 million chronically infected. Because of distinct epidemiological features of China’s HCV epidemiology (greater transmission due to contaminated blood and unsafe medical procedures relative to transmission from injection drug use), using models and parameters developed for other countries may be inappropriate. A further challenge to constructing a Chinese HCV model is the availability of fragmentary epidemiological data. Hence, we use model calibration to combine such data and explore the uncertainty in the drivers of China HCV epidemic.

Method:

   We construct an HCV microsimulation model to reproduce the observed prevalence of HCV infection and liver fibrosis stage distribution in China. The model includes risks of HCV infection, fibrosis progression, and transitions to decompensated cirrhosis and liver cancer. Risks of HCV infection and fibrosis progression depend on age, sex, and alcohol intake and are calibrated to a national HCV sero-epidemiological study along with Chinese epidemiological targets identified through systematic review. Prior distributions cover ranges those reported in U.S./European HCV models. We sampled 100,000 parameter sets from the priors, using each to simulate 10,000 individuals over their lifetimes. We computed a goodness-of-fit score for each set relative to the epidemiological targets. We formed the posterior distribution from sets with the top 1% of scores. We compared these posteriors to analogous parameters from the U.S. setting.

Result:

  Calibration achieved consistency with multiple Chinese epidemiological targets. Parameters from the calibrated model are consistent with underlying biological processes (fibrosis progression) while highlighting the extent to which epidemiological differences are driven by differences in patterns of infection. The calibrated Chinese monthly risk of fibrosis progression for lower risk individuals (age at infection <40, no alcohol) was 0.01, consistent with that of U.S. HCV-infected U.S. individuals (0.008). However, unlike the U.S. where HCV infection is concentrated among 20-50 year-olds, our infection rates increase with age. The infection rate in Chinese elderly is 10 times higher than the U.S. rate (27.4 vs. 0.5 per 100,000). 

Conclusion:

   Our calibrated HCV model reproduces key features of HCV epidemiology in China, reflecting rates that are distinct from those in the U.S. It represents a necessary first step to accurately assess the cost-effectiveness of HCV prevention and treatment interventions for the Chinese population.