Kate Battle

Semi-mechanistic modelling for serological survey data

Geospatial Modelling For Malaria Risk Stratification And Intervention Targeting

As low burden endemic countries move towards malaria elimination and the number of individuals carrying a microscopically-patent malaria infection at any given time declines, so too does the efficiency of parasite surveys for identifying pockets of residual transmission. The use of serological measurements targeting potentially long-lived products of the immune response to parasite invasion offers a means to devise new survey approaches probing exposure over historical windows of months and/or years, thereby increasing positive sample sizes and spatial risk mapping accuracy at the expense of temporal resolution.

Previous statistical models for relating serological markers to the strength of malaria transmission have focussed on single antigen observations and supposed a fixed exposure for the entire community. An ongoing strand of our research is to develop more sophisticated models based on a semi-mechanistic platform—a compromise between fully mechanistic models of the transmission cycle and a ‘black-box’ statistical approach—to allow the effective integration of serological data within our risk mapping framework.

Initial applications to serological data from Cambodia, Myanmar, and Haiti have demonstrated the potential of this approach and its importance to improved operational planning for malaria control.