Presenter of 1 Presentation
PRE-RECORDED: GEOSTATISTICAL ANALYSIS OF ACTIVE CYSTICERCOSIS: RESULTS OF A LARGE-SCALE STUDY IN 60 VILLAGES IN BURKINA FASO (ID 1316)
Abstract
Introduction
Cysticercosis (CC) is a neglected tropical disease caused by Taenia solium which has been observed to cluster spatially. Geostatistical models can predict outcome prevalence at unsampled locations, and thus guide prevention and control strategies. The goal of this study was to fit, for the first time, a geostatistical model to human CC data.
Methods
Baseline data from a randomized control trial (EFECAB) conducted in 60 villages in Burkina Faso were used. A generalized linear geostatistical model (GLGM) was run, with active human CC (B158/B60 Ag-ELISA) as outcome and a set of environmental variables linked with infection transmission and spread as explanatory variables for the spatial distribution.
Results
The final GLGM retained precipitation, distance to the nearest river and night land temperatures as predictors for active human CC. The range of spatial correlation was estimated at 45 meters for the participant-level data, while it was 28.2 km using the village-level data. The prediction maps unravelled large areas with human CC prevalence estimates of at least 4% in the south-east, extreme south, and north-west of the study area.
Conclusions
This study shows the merits of geostatistical models for identifying target areas for intervention. Future studies should improve sampling strategies to ensure appropriate characterisation of the spatial variability.