Model-Based Estimation of the Prevalence of Chronic Hepatitis B in Canada and Cost-Effectiveness Analysis of Implementing a Universal Birth Hepatitis B Vaccination Program in Ontario, Canada

Abstract

BACKGROUND Hepatitis B is a potentially life-threatening acute or chronic disease that is caused by hepatitis B virus (HBV). The development of chronic disease is age-dependent, with the highest risk in the infant population. As such, the World Health Organization recommends universal hepatitis B vaccination within the first 24 hours of birth. Yet, hepatitis B vaccination is provided at age 12 years in Ontario, Canada. Chronic hepatitis B (CHB) is also characterized by the lack of symptoms until progression to end-stage liver disease, making it difficult identify infected patients early in the disease. This has resulted in uncertainties regarding the true prevalence and the impact of CHB in Canada.

OBJECTIVES The objectives of this thesis are to: 1) assess the cost-effectiveness of implementing a universal hepatitis B vaccination program in newborns versus adolescents in Ontario, and 2) estimate the true prevalence of CHB and the proportion of undiagnosed cases in Canada.

METHODOLOGY Two models were developed to achieve the study objectives. First, a state-transition model representing the natural history of acute and chronic hepatitis B was developed in TreeAge Pro to assess the cost-effectiveness of two hepatitis B vaccine schedules. Analyses were performed from a public payer perspective with a lifetime time horizon and a 1.5% annual discount rate. Second, a modified version of the natural history model was adopted to create a prevalence model in MATLAB to estimate the prevalence of CHB using Markov Chain Monte Carlo. Model input data were obtained from peer-reviewed literature and publicly available databases from Statistics Canada and Public Health Agency of Canada.

RESULTS Birth vaccination was found to be cost-saving compared to the current adolescent vaccination strategy in Ontario. Probabilistic analysis resulted in a mean cost of $317,261 and 43.36 QALYs for birth vaccination versus $317,735 and 43.18 QALYs for adolescent vaccination. A microsimulation showed that the birth vaccination strategy leads to decreases in liver-related cases by 15.96% in acute hepatitis B, 44.27% in CHB, 47.45% in compensated cirrhosis, 47.54% in hepatocellular carcinoma, 56.44% in decompensated cirrhosis, 50.00% in liver transplant, and 51.16% in liver death. In Canada, the model estimated both the prevalence of CHB and proportion of undiagnosed cases to have trended downwards in the total population from 2011 to 2017. Overall, when all age cohorts were combined, CHB prevalence was estimated to be 0.85% and the undiagnosed proportion was estimated to be 32.77%. The model-generated estimate for CHB prevalence of 0.85% was approximately doubled the previously estimated seroprevalence of 0.4% from a national seroprevalence study.

CONCLUSION The results of the study indicate that by switching to a birth hepatitis B vaccination program, the Ontario government can save healthcare spendings while increasing clinical benefits. The results of the study provide policy makers with actionable recommendations on re-assessing the current hepatitis B vaccination schedule in Ontario. The second model also showed that the prevalence of CHB may be much higher than previously estimated and that a significant proportion of patients remain undiagnosed. The prevalence model demonstrates a feasible framework for future analyses using administrative databases to more accurately identify the true burden of CHB in Canada.