Loa-loa Modelling

Our earlier work on Loa-loa modelling developed a predictive map to identify areas of high Loa loa prevalence by combining prevalence survey data with satellite-derived environmental covariates in a geostatistical model for the spatial variation in prevalence. More recently, we have been modelling the relationship between community-level prevalence and the distribution of intensity of Loa Loa infection amongst infected members of the community. This allows us to estimate the proportion of the community whose parasite counts exceed certain policy-relevant infection levels (for example 8,000 or 30,000 microfilariae/ml), using only prevalence data. The importance of this is that highly infected individuals are at risk of severe adverse events when treated with drugs used in mass administration programmes to control Onchocerciasis and Lymphatic Filariasis, but estimation of infection levels on a wide geographical scale is much more expensive and time-consuming than estimating prevalence.