Hitting hotspots: spatial targeting of malaria for control and elimination

Abstract

Teun Bousema and colleagues argue that targeting malaria “hotspots” is a highly efficient way to reduce malaria transmission at all levels of transmission intensity.

Figure 1. Hotspots of malaria transmission in the dry and wet season.

Mosquito exposure and parasite carriage are highly focal in the dry season (A). People living in hotspots are exposed to higher mosquito densities and, because individuals in households belonging to hotspots are more likely to be infected and infectious, mosquitoes are more likely to acquire a malaria infection in these households. In the wet season, as mosquito density and geographic distribution increase, infectious mosquitoes drive infection out into the rest of the village (B).

Figure 2. Targeted and untargeted interventions with long-lasting LLINs and IRS in a malaria elimination scenario.

The simulations for the low endemic setting with a baseline parasite prevalence of ∼15% in the general population (A) are based on parasite prevalence and mosquito exposure data from Korogwe, northern Tanzania (2008) . Effective coverage with LLINs is scaled up over 6 years to 60% prior to the intervention, creating a starting point for interventions aiming towards malaria elimination . Subsequently, the impact of four intervention strategies is simulated using an individual-based simulation model : (i) increasing LLIN coverage to 80% in a untargeted manner (blue solid line); (ii) increasing LLIN coverage with the same number of LLINs but preferentially targeting hotspots where 90% coverage is reached (dashed blue line); (iii) increasing LLIN coverage to 80% and yearly introducing IRS at 20% coverage in a untargeted manner (red solid line); (iv) a targeted approach using the same resources as the third scenario, reaching 90% effective coverage with LLINs and 90% effective coverage with yearly IRS in hotspots (dashed red line). LLINs were replaced every 4 years. Simulations were repeated for an area of high endemicity with a parasite prevalence of ∼40% in the general population (B).