Nov 13 - 17, 2016
EVALUATING STRATIFIED MALARIA CONTROL INTERVENTIONS IN BIOKO ISLAND: DIFFERENT APPROACHES TO FOCALIZED INTENSIFIED MALARIA CONTROL INTERVENTIONS THROUGH SPATIAL CLUSTERING AND RISK MAPS
Guillermo A. Garcia Contreras1 , Dianna Hergott1 , Christopher Schwabe1 , Wonder Phiri1 , Megan Perry1 , Immo Kleinschmidt2 , John Bradley2 , Jose Luis Segura1 , Jordan Smith1
1 Medical Care Development International, Silver Spring, MD, United States, 2 London School of Hygiene & Tropical Medicine, London, United Kingdom
The Bioko Island Malaria Control Project (BIMCP) created a geo-referenced mapping system in 2012 assigning a unique identifier to all households similar to an address. In 2014 this mapping system, based on ArcGIS software and satellite imagery, was linked to the Campaign Management Information System (CIMS), an Android-based tablet application, and is currently used to plan, implement, and monitor field malaria control activities on Bioko; ultimately allowing the BIMCP to accurately track all malaria control interventions at the household and community level over time. To account for budget constrains essential for long termsustainability of malaria control programs, a stratified control strategy can provide a possible sustainable and reproducible solution. In 2015, the BIMCP developed a framework for Indoor Residual Spraying (IRS) using a stratified methodology to target communities with higher risk of malaria prevalence. The model used: pre-existing Malaria Indicator Survey (MIS) data focusing on prevalence and risk of importation, housing characteristics for all households, spray coverage, and slope and altitude. Information is linked to the unique household identifier and a risk score is created at the community level identifying the most vulnerable communities that would be selected for IRS. Using this preexisting model for stratification, additional analysis will be carried out focusing on two different spatial clustering techniques: 1) Kulldorff's spatial scan statistics using SaTScan v9.4.2 and 2) Anselin’s Local Moran’s I statistics using ArcGIS v10.4; as well as regression analysis including data on socio-economic status, bed net coverage, and malaria incidence. Due to the unique characteristics of the methodologies that will be used for stratification, the parameters will not be analogous, but various thresholds will be taken into consideration in order to achieve a higher degree of comparability. Once all models have been completed and quantitatively verified, maps will be created for each methodology, overlaying the results into one map, in order to provide evidence of visual clustering of malaria risk areas.