Open Access Research article

Elevation and cholera: an epidemiological spatial analysis of the cholera epidemic in Harare, Zimbabwe, 2008-2009

Miguel A Luque Fernandez1*, Michael Schomaker1, Peter R Mason2, Jean F Fesselet3, Yves Baudot4, Andrew Boulle1 and Peter Maes5

Author Affiliations

1 Centre of Infectious Disease Epidemiology and Research (CIDER), University of Cape Town, Cape Town, South Africa

2 , Biomedical Research & Training Institute and the University of Zimbabwe College of Health Sciences, Harare, Zimbabwe

3 Medecins sans Frontieres / Public health department: Water, Hygiene and Sanitation Unit, Amsterdam Operational Centre, Amsterdam, Netherlands

4 NADAR sprl, Geographic Information Systems, Marchin, Belgium

5 Medecins sans Frontieres / Medical department: Water, Hygiene and Sanitation Unit, Brussels Operational Centre, Brussel, Belgium

For all author emails, please log on.

BMC Public Health 2012, 12:442  doi:10.1186/1471-2458-12-442

Published: 18 June 2012

Abstract

Background

In highly populated African urban areas where access to clean water is a challenge, water source contamination is one of the most cited risk factors in a cholera epidemic. During the rainy season, where there is either no sewage disposal or working sewer system, runoff of rains follows the slopes and gets into the lower parts of towns where shallow wells could easily become contaminated by excretes. In cholera endemic areas, spatial information about topographical elevation could help to guide preventive interventions. This study aims to analyze the association between topographic elevation and the distribution of cholera cases in Harare during the cholera epidemic in 2008 and 2009.

Methods

We developed an ecological study using secondary data. First, we described attack rates by suburb and then calculated rate ratios using whole Harare as reference. We illustrated the average elevation and cholera cases by suburbs using geographical information. Finally, we estimated a generalized linear mixed model (under the assumption of a Poisson distribution) with an Empirical Bayesian approach to model the relation between the risk of cholera and the elevation in meters in Harare. We used a random intercept to allow for spatial correlation of neighboring suburbs.

Results

This study identifies a spatial pattern of the distribution of cholera cases in the Harare epidemic, characterized by a lower cholera risk in the highest elevation suburbs of Harare. The generalized linear mixed model showed that for each 100 meters of increase in the topographical elevation, the cholera risk was 30% lower with a rate ratio of 0.70 (95% confidence interval=0.66-0.76). Sensitivity analysis confirmed the risk reduction with an overall estimate of the rate ratio between 20% and 40%.

Conclusion

This study highlights the importance of considering topographical elevation as a geographical and environmental risk factor in order to plan cholera preventive activities linked with water and sanitation in endemic areas. Furthermore, elevation information, among other risk factors, could help to spatially orientate cholera control interventions during an epidemic.