State transition detection in the spatio-temporal incidence of malaria.

Mosquito-borne disease spread might exhibit irregular epidemic fronts caused by ecological heterogeneity in the risk factors. To determine Plasmodium vivax infection spread in north-eastern Venezuela, we used the State Transition Index (STI) to detect the spatial locations of malaria incidence boundaries and their dynamics over time. Then, we evaluated the role of population size on disease persistence. Boundary locations of malaria were found to be highly spatially variable. Waves of infection were observed in the spatial mosaics of large and small nearby localities due to a strong asynchrony in the epidemic dynamics between both host populations. Our results suggest that the epidemic spatial diffusion follows a hierarchy from large, populated villages (with few or no seasonal parasite fadeouts) to smaller, less populated localities, where infection outbreak was irregular or disease dynamics showed frequent fadeouts. Our findings stress the importance of malaria surveillance and control in these larger communities.