Topographic predictability of locations for targeting complementary interventions to eliminate residual malaria transmission in urban Dar es Salaam, Tanzania
Abstract
Background: Topographic indicators of local water accumulation potentials are widely used and considered to be accessible, affordable and scalable predictors of mosquito aquatic habitats and malaria risk. In Dar es Salaam,... [ view full abstract ]
Background: Topographic indicators of local water accumulation potentials are widely used and considered to be accessible, affordable and scalable predictors of mosquito aquatic habitats and malaria risk. In Dar es Salaam, local topography was used to identify high risk areas for directing complementary interventions to eliminate residual malaria transmission.
Methods: A comprehensive range of topographical indicators were derived from a freely available digital elevation model (DEM) and calibrated as predictors of malaria transmission risk using data from household cross-sectional surveys of malaria infection prevalence conducted from March 2010 to July 2012. Conditional Autoregressive Models were fitted to map areas that could be geographically and selectively targeted with interventions targeting malaria infection prevalence.
Results: As expected, malaria infection prevalence were intuitively associated with topographic predictors of aquatic habitat suitability such as low elevation (OR [95%CI] = 0.98 [0.97, 0.99], ρ=0.00791). Surprisingly, it was also associated with very well drained locations such as the edges of elevated areas with convex curvature (OR [95%CI] = 0.76 [0.58, 0.99], ρ=0.04675) and edges of valleys with steep slopes (β [95%CI] 0.75 [0.74, 0.75], ρ=0.00013). These counter-intuitive indicators provide topographic zones for human exposure as they define interface of human settlements (where residents are exposed to dispersing mosquitoes) and valleys (where mosquitoes emerge/oviposit). However, topographic indicators accounted for only a modest proportion of overall variance of malaria prevalence (R2=55%) and targeting efficiency (56% of cases in the 44% highest scoring locations), thus falling short of the 80/20 rule for effective targeting.
Conclusion: Variations in human resilience might presumably be limiting the power of topography in predicting malaria risk even in areas with high hazard. The interfaces between valleys and human habitation could be selectively targeted with complementary barrier interventions that increase human resilience by protecting households against new infections such as bednets, mosquito-proofed housing and repellents.
Authors
-
Victoria Mwakalinga
(Ardhi University)
-
Maureen Coetzee
(University of Witwatersrand)
-
Gerry Killeen
(Ifakara Health Institute)
-
Ben Sartorius
(University of KwaZulu-Natal)
-
Stefan Dongus
(University of Basel)
-
Daniel Msellemu
(Ifakara Health Institute)
-
Yeromin Mlacha
(Ifakara Health Institute)
-
Nicodem Govella
(Ifakara Health Institute)
Topic Area
V. Health indicators, spatial analysis and mapping: new tools, new methods 5.1 Spatial ana
Session
SPH-UH-01B » Spatializing Urban Health (08:00 - Friday, 1st April, TBA)
Paper
Mwakalinga_ISUH_2016_abstract.docx
Presentation Files
The presenter has not uploaded any presentation files.