Mapping the Geographic Distribution of Tungiasis in Sub-Saharan Africa

Abstract

The geographic distribution of tungiasis is poorly understood, despite the frequent occurrence of the disease in marginalized populations of low socioeconomic status. To date, little work is available to define the geography of this neglected tropical disease (NTD). This exploratory study incorporated geostatistical modeling to map the suitability for tungiasis transmission in sub-Saharan Africa (SSA). In SSA, environmental suitability is predicted in 44 countries, including Angola, Nigeria, Ghana, Cameroon, Cote de Ivoire, Mali, Ethiopia, the Democratic Republic of the Congo, Kenya, Gabon, Uganda, Rwanda, Tanzania, Zambia, Zimbabwe, Madagascar, and South Africa. In total, an estimated 668 million people live in suitable areas, 46% (304 million) of which reside in East Africa. These evidence-based maps provide vital evidence of the potential geographic extent of SSA. They will help to guide disease control programs, inform policymakers, and raise awareness at the global level. Likewise, these results will hopefully provide decisionmakers with the pertinent information necessary to lessen morbidity and mortality in communities located in environmentally suitable areas.

Keywords: GIS; disease mapping; medical geography; neglected tropical disease; precision public health; tungiasis.

Figure 1

Severe tungiasis in the feet of a 10-year-old girl. Adapted from Barbosa and Barbosa [9]. License Number: 4832790723180.

Figure 2

Model calibration region and study area defined by the WHO Global Burden of Disease Regions (GBD). Green dots correspond to occurrence locations (n = 86). The accepted historic arrival in Africa of Tunga penetrans in Ambriz, Angola (1872), and the probable presence at Oualata (1324) are highlighted.

Figure 3

Workflow for tungiasis environmental suitability mapping.

Figure 4

(A) Predicted distribution across sub-Saharan Africa (SSA). Probability of occurrence is provided on a scale from 0 (low probability) to 1 (high probability). Green dots correspond to occurrence locations (n = 86). (B) Status of occurrence reporting by country in SSA based on the weighted mean suitability model. Insets correspond to São Tomé and Príncipe and Comoros.

Figure 5

Variable contributions (top) and predictive algorithm performances (bottom). A. PC: principal component contribution. B. ROC: receiver operating characteristic, KAPPA: Cohen’s Kappa, and TSS: true skill statistic.

Figure 6

(A,C) The proportion of people living on $1.25 a day in Nigeria, Uganda, and Kenya. Red corresponds to a higher proportion of the population living on $1.25 a day; yellow represents a lower proportion of the population living on $1.25 a day. (B,D) The coefficient of variation (CV) or measure of model uncertainty. Lighter colors = less uncertainty; darker colors = increased uncertainty.

Figure 7

(A) The proportion of people living on $1.25 a day in Tanzania and Malawi. Red corresponds to a higher proportion of the population living on $1.25 a day; yellow represents a lower proportion of the population living on $1.25 a day. (B) The coefficient of variation (CV) or measure of model uncertainty. Lighter colors = less uncertainty; darker colors = increased uncertainty.