Geospatial modeling of pre-intervention prevalence of Onchocerca volvulus infection in Ethiopia as an aid to onchocerciasis elimination

Abstract

Background Onchocerciasis is a neglected tropical and filarial disease transmitted by the bites of blackflies, causing blindness and severe skin lesions. The change in focus for onchocerciasis management from control to elimination requires thorough mapping of pre-control endemicity to identify areas requiring interventions and to monitor progress. Onchocerca volvulus infection prevalence in sub-Saharan Africa is spatially continuous and heterogeneous, and highly endemic areas may contribute to transmission in areas of low endemicity or vice-versa. Ethiopia is one such onchocerciasis-endemic country with heterogeneous O. volvulus infection prevalence, and many districts are still unmapped despite their potential for O. volvulus infection transmission. Methodology/Principle findings A Bayesian geostatistical model was fitted for retrospective pre-intervention nodule prevalence data collected from 916 unique sites and 35,077 people across Ethiopia. We used multiple environmental, socio-demographic, and climate variables to estimate the pre-intervention prevalence of O. volvulus infection across Ethiopia and to explore their relationship with prevalence. Prevalence was high in southern and northwestern Ethiopia and low in Ethiopia's central and eastern parts. Distance to the nearest river (-0.015, 95% BCI: -0.025 - -0.005), precipitation seasonality (-0.017, 95% BCI: -0.032 - -0.001), and flow accumulation (-0.042, 95% BCI: -0.07 - -0.019) were negatively associated with O. volvulus infection prevalence, while soil moisture (0.0216, 95% BCI: 0.014 - 0.03) was positively associated. Conclusions/Significance Infection distribution was correlated with habitat suitability for vector breeding and associated biting behavior. The modeled pre-intervention prevalence can be used as a guide for determining priority for intervention in regions of Ethiopia that are currently unmapped, most of which have comparatively low infection prevalence.

Synthesizing environmental, epidemiological and vector and parasite genetic data to assist decision making for disease elimination

“…We used variables corresponding to the year when the samples were collected for fitting the models: before 2001 for prevalence data, from 2010-2012 for O. volvulus, and 2013-2015 for S. damnosum. Continuous environmental rasters that might be ecologically relevant to the onchocerciasis distribution based on field experiments on blackflies (Cheke et al, 2017;Opoku, 2006) and previous geospatial modelling studies (Barro & Oyana, 2012;Cheke et al, 2015;Cromwell et al, 2021;O'Hanlon et al, 2016;Shrestha et al, 2022) included distance to the nearest river, soil moisture, elevation, slope, temperature, and precipitation (Barro & Oyana, 2012;Cromwell et al, 2021;Shrestha et al, 2022). Because the dispersal capacity of the Simulium vector is dependent on the vegetation type and time of the year (World Health Organization and Onchocerciasis Technical Advisory Subgroup, 2020), we included vegetation and seasonality-related variables.…”

“…The mean of the posterior prevalence was estimated based on the pre-MDAi microfilaria prevalence data using the Bayesian approach with Integrated Nested Laplace Approximation (INLA) (Moraga et al, 2015;Rue et al, 2009). To select the best model, we compared different model types (with and without spatial random effects) and different triangulation meshes using AIC and WAIC scores as described in (Shrestha et al, 2022). The detailed code for prevalence analysis is available on GitHub (https:// github.…”

“…We created a triangulation mesh with a 3 km cut-off; the maximum length of triangles inside and outside the boundary was set to 10 km and 100 km, respectively. Finally, we fitted the model and assessed the relationship of environmental variables with the prevalence data (as in (Shrestha et al, 2022)). The prediction of the posterior prevalence was made at a 2-km resolution considering the computational cost of prediction at a lower resolution.…”

“…Areas with high soil moisture tend to be arable land that is either inhabited by or worked by people who are then exposed to more vector bites (Adeleke et al, 2010;Opoku, 2006), and has been reported to be an important environmental feature influencing the occurrence of onchocerciasis in other studies (Cromwell et al, 2021;Shrestha et al, 2022).…”

“…The blackfly vectors have a narrow ecological niche for breeding and larval development: they require fast-flowing, highly oxygenated waters. This constraint drives spatial heterogeneity in the prevalence and transmission of onchocerciasis, with higher prevalence in communities near rivers (hence the name river blindness) (Ngoumou et al, 1994;O'Hanlon et al, 2016;Schmidt et al, 2022;Shrestha et al, 2022;Vieira et al, 2005;Zouré et al, 2014). The primary tool for onchocerciasis control is mass drug administration with ivermectin (MDAi), which initially focused on highly endemic communities (World Health Organization and African Programme for Onchocerciasis Control, 2010Control, , 2015; thus, there is also spatial heterogeneity in intervention history.…”