Abstract
BACKGROUND: Bangladesh first reported Japanese encephalitis virus (JEV) in 1977 and has seen regular cases since, however, no JEV vaccination program currently exists. A barrier to the use of JEV vaccines has been a limited understanding of the underlying burden. METHODS: We conducted a nationally representative serological community study in 70 communities in individuals of all ages (N=2,938, October 2015-January 2016). Serum samples were tested for IgG antibodies against JEV. We developed spatially explicit serocatalytic models to estimate the underlying force of infection across the country. We then used mathematical models to estimate the annual JE disease burden currently and under different vaccination strategies. FINDINGS: The overall JEV seroprevalence in Bangladesh was 3.4% (95%CI: 2.8-4.1, range 0-28% across communities). The annual probability of infection was 0.005 (95%CI: 0.003-0.009), with risk greatest near border regions. We estimated that annually there are 157 clinical cases (95%CI: 89-253) and 31 deaths (95%CI: 18-52). A vaccination strategy in the 10 most affected districts in 60% of 1-15 year olds would require 5 million doses and avert 1 case per 100,000 doses over five years compared to 35 million doses and 0.5 cases averted for a nationwide campaign. No vaccination scenario was cost-effective under a willingness-to-pay of three-times gross domestic product. INTERPRETATION: A spatially targeted vaccine campaign would be most effective in reducing JEV burden, however, would still not meet standard cost effectiveness targets. FUNDING: CDC.