Cost‐effectiveness of augmenting current perinatal hepatitis B prevention program with maternal antiviral therapy

Abstract

Supported, in part, by grants from Jiangsu Provincial Department of Science and Technology (no. BE2015655), China, and the National Natural Science Foundation of China (nos. 81370520 and 81373060). Potential conflict of interest: Nothing to report. TO THE EDITOR: We read with interest the well‐designed study by Fan et al.1 on cost‐effectiveness of the current perinatal hepatitis B (HepB) screening and immunization program versus augmenting this strategy with maternal antiviral therapy, with a Markov decision tree model involving perinatal hepatitis B virus (HBV) infection risks, long‐term health, and economic consequences. However, several aspects need clarification in this article. First, Fan et al. perform a thorough cost‐effectiveness analysis (CEA) on the lifelong costs and outcomes of three strategies for preventing perinatal HBV infection. A 78‐year cycle with a 1‐year interval is specified for the Markov model. Given that most infants and children with perinatal HBV infection maintain immune tolerance, the researchers start the Markov process at age 20 years. Considering the hypothetical 2010 birth cohort of 4 million infants and approximately 190 million quality‐adjusted life‐years (QALYs) for each strategy, there is an obvious discrepancy between the average 47.5 QALYs and the 78 years of life expectancy, probably attributed to exemption of the first 20 years' QALYs for the entire cohort. We suggest that these common QALYs should be included in the total rewards, while the final incremental cost‐effectiveness ratio (ICER) results remain the same. Second, the researchers summarize the parameter values and their ranges for the decision tree and Markov model in Table 1. Two baseline parameters, that is, cost for hepatitis B surface antigen screening test and HepB vaccination, are lacking. Actually, these costs contributed substantially to the total cost and should be listed in the article, as in other reports.2 Third, using one‐way sensitivity analyses, the researchers demonstrate that the antiviral prophylaxis remains cost‐effective over wide ranges of time‐varying and district‐specific factors. However, the results in Table 4 might be misleading because the ranges of output ICERs are not consistent with the interpretation within the text. For example, bigger reduction in perinatal HBV transmission (20%‐80%) from maternal antiviral therapy makes antiviral prophylaxis a more cost‐effective program (ICER $97,749/QALY—cost saving, rather than cost saving—$97,749/QALY), compared with the current strategy. Similarly, the ranges of $2,886‐$15,552/QALY, $3,961‐$11,131/QALY, cost saving—$269,796/QALY, $4,635‐$8,446/QALY, cost saving—$21,549/QALY and $6,876‐$8,926/QALY should be reversed or presented in a Tornado diagram as in other reports.3 Last, in CEA, the strategy of choice is usually determined by comparing the ICER to what decision makers are willing to pay (WTP) for an additional QALY.4 The discussion of WTP is lacking in the article.