Using Routinely Collected Health Records to Identify the Fine-Resolution Spatial Patterns of Soil-Transmitted Helminth Infections in Rwanda

Abstract

Background. Soil-transmitted helminths (STH) are parasitic diseases with significant public health impact. Analysis is generally based on cross-sectional prevalence surveys; outcomes are mostly aggregated to larger spatial units. However, recent research demonstrates that infection levels and spatial patterns differ between STH species and tend to be localized. Methods. Incidence data of STHs including roundworm (Ascaris lumbricoides), whipworm (Trichuris trichiura) and hookworms per primary health facility for 2008 were linked to spatially delineated primary health center service areas. Prevalence data per district for individual and combined STH infections from the 2008 nationwide survey in Rwanda were also obtained. Results. A comparison of reported prevalence and incidence data indicated significant positive correlations for roundworm (R2 = 0.63) and hookworm (R2 = 0.27). Weak positive correlations were observed for whipworm (R2 = 0.02) and the three STHs combined (R2 = 0.10). Incidence of roundworm and whipworm were found to be focalized with significant spatial autocorrelation (Moran’s I > 0: 0.05−0.38 and p ≤ 0.03), with (very) high incidence rates in some focal areas. In contrast, hookworm incidence is ubiquitous and randomly distributed (Moran’s I > 0: 0.006 and p = 0.74) with very low incidence rates. Furthermore, an exploratory regression analysis identified relationships between helminth infection cases and potential environmental and socio-economic risk factors. Conclusions. Findings show that the spatial distribution of STH incidence is significantly associated with soil properties (sand proportion and pH), rainfall, wetlands and their uses, population density and proportion of rural residents. Identified spatial patterns are important for guiding STH prevention and control programs.

Keywords: GIS; Rwanda; soil-transmitted helminth infection; spatial pattern.

Figure 1

District boundaries, primary health facility locations as provided by RBC–GIS unit and Health Facility Service Area (HFSA) boundaries delineated using spatial allocation tools and data from the community-based health insurance scheme in 2010.

Figure 2

STH incidence rates (per 10,000 persons) per district in Rwanda for 2008, as reported in RHIMS.

Figure 3

STH incidence and prevalence rates and their correlation at the district level. Starting from the left-upper corner, incidence maps for combined STHs (a), for T. trichiura (b), for A. lumbricoides, (c) and for hookworm (d). The prevalence maps and scatter plots are shown for combined STH cases (e), individual cases with T. trichiura (f), A. lumbricoides (g) and hookworm (h).

Figure 4

Incidence rates at HFSA level for combined STH infection (a), T. trichiura (b), A. lumbricoides (c) and hookworm (d).

Figure 5

Spatial distributions of Anselin local Moran’s I at the HFSA level, for (a) combined STH cases; (b) T. trichiura; (c) A. lumbricoides and (d) hookworm.