Authors
L A Chapman4; R Chowdhury1; C Bern3; G F Medley2; T D Hollingsworth4; 1 KalaCORE Programme, Bangladesh, Bangladesh; 2 London School of Hygiene & Tropical Medicine; 3 University of California, San Francisco, United States; 4 University of Warwick Discussion
Visceral leishmaniasis (VL), the world's second most deadly vector-borne parasitic disease, has been targeted for elimination in the Indian sub-continent by 2020. However, better understanding of spatial transmission of the disease and the contribution of asymptomatically infected individuals to transmission is required in order to achieve and sustain this goal. Existing models of VL fail to account for spatial heterogeneity in transmission, despite evidence that infection risk increases with proximity to a VL case. To address these issues we have developed an individual-based spatiotemporal model of VL transmission for a detailed epidemiological dataset from Bangladesh, and used it to estimate the rate at which infection risk decreases with distance from a case and the contribution of asymptomatics to transmission. The dataset includes information on clinical VL incidence and cross-sectional diagnostic survey results for 2494 individuals in 506 geo-located households in 3 villages in Fulbaria, Mymensingh, Bangladesh from 1999-2004. The parameter estimates obtained suggest that infection risk decreases relatively quickly with distance from a case (halving within 100m) and that clinical VL cases contribute four times more to transmission than asymptomatic individuals. This suggests that reactive insecticide spraying strategies, in which houses within a certain radius of a new VL case are sprayed to kill the sandfly vectors, could be effective in limiting transmission. 
