Mapping the endemicity and seasonality of clinical malaria for intervention targeting in Haiti using routine case data

Abstract

Towards the goal of malaria elimination on Hispaniola, the National Malaria Control Program of Haiti and its international partner organisations are conducting a campaign of interventions targeted to high-risk communities prioritised through evidence-based planning. Here we present a key piece of this planning: an up-to-date, fine-scale endemicity map and seasonality profile for Haiti informed by monthly case counts from 771 health facilities reporting from across the country throughout the 6-year period from January 2014 to December 2019. To this end, a novel hierarchical Bayesian modelling framework was developed in which a latent, pixel-level incidence surface with spatio-temporal innovations is linked to the observed case data via a flexible catchment sub-model designed to account for the absence of data on case household locations. These maps have focussed the delivery of indoor residual spraying and focal mass drug administration in the Grand’Anse Department in South-Western Haiti.

Spatial, environmental, and individual associations with Anopheles albimanus salivary antigen IgG in Haitian children

“…The regions of Haiti included as school sampling sites have high heterogeneity in malaria disease burden and vector density ( Frederick et al., 2016 ; Cameron et al., 2021 ). High levels of vector exposure can still occur even when P. falciparum is not prevalent in the human population.…”

“…Hispaniola, an island composed of the Dominican Republic and Haiti, is the only area in the Caribbean with endemic malaria, with Plasmodium falciparum as the primary species. Though recent malaria transmission in Haiti has been relatively low ( Lucchi et al., 2014 ; Jules et al., 2022 ), heterogeneity by spatial, individual, and environmental factors accentuates the need for enhanced surveillance methods to characterize higher-risk regions and population subgroups to further move towards malaria elimination ( Boncy et al., 2015 ; Cameron et al., 2021 ).…”

Improving access to care and community health in Haiti with optimized community health worker placement

“…In other situations, disease-specific risk could also be accounted for. For instance, in order to prioritise areas where significant malaria transmission is identified, the placement algorithm can be modified to include malaria incidence into the model in addition to capacity and walking time constraints, either by giving more weight or lower population numbers per CHW in malaria endemic areas, or by restricting the allocation to the areas where malaria risk is considered present with a predefined probability (exceedance probability maps) [34]. Additionally, detailed workload data can be used to calibrate the maximum capacity per CHW, which is very specific to the characteristics of a given CHW program [32].…”

“…The analysis focused exclusively on physical accessibility and did not consider other determinants of access to care, such as financial accessibility [4], the quality of services provided [36], and more generally demographic, economic and social factors impeding or facilitating healthcare access [37,38]. This analysis also focused exclusively on population coverage assuming that CHWs serve all individuals equally and therefore did not consider how the decisions regarding the geographical distribution of CHWs could depend on the presence of epidemiological risks, for example higher malaria risk in the Grande-Anse and Sud departments [34], or particular needs in specific population subgroups. Furthermore, the choice of thresholds for the maximum distance and maximum population in urban and rural areas could lead to border effects and gives a tendency for the model to place more CHWs in high-capacity areas, in order to minimise their total number.…”

A General Framework for Cutting Feedback within Modularized Bayesian Inference

“…This differs from standard Bayesian inference where PK and PD models are inferred simultaneously. More examples that adopt modularized Bayesian inference for this purpose include: Li et al (2013), which removes the influence from the suspect highest streamflow observations in hydrological modeling; Mikkelä et al (2019), which removes the influence from the less "valid" reported human salmonellosis cases data to the estimation of the distribution of salmonella subtypes in the food sources; Arambepola et al (2020), which removes the influence from the less reliable malaria incidence data to the malaria prevalence estimation; and Cameron et al (2021), which applies modularized Bayesian inference to focus on the estimation of endemic transmission intensity of malaria.…”