The geography of malaria elimination in Bangladesh: combining data layers to estimate the spatial spread of parasites

Abstract

Malaria control programs face difficult resource allocation decisions. Of particular concern for countries aiming for malaria elimination, the regular movement of individuals to and from endemic areas undermines local interventions by reintroducing infections and sustaining local transmission. Quantifying this movement of malaria parasites around a country has become a priority for national control programs, but remains methodologically challenging, particularly in areas with highly mobile populations. Here, we combined multiple data sources to measure the geographical spread of malaria parasites, including epidemiological surveillance data, travel surveys, parasite genetic data, and anonymized mobile phone data. We collected parasite genetic barcodes and travel surveys from 2,090 patients residing in 176 unions in southeast Bangladesh. We developed a genetic mixing index to quantify the likelihood of samples being local or imported. We then inferred the direction and intensity of parasite flow between locations using an epidemiological model, and estimated the proportion of imported cases assuming mobility patterns parameterized using the travel survey and mobile phone calling data. Our results show that each data source provided related but different information about the patterns of geographic spread of parasites. We identify a consistent north/south separation of the Chittagong Hill Tracts region in Bangladesh, and found that in addition to imported infections from forested regions, frequent mixing also occurs in low transmission but highly populated areas in the southwest. Thus, unlike risk maps generated from incidence alone, our maps provide evidence that elimination programs must address ongoing movement of parasites around the lower transmission areas in the southwest.

Managing the spread of disease with mobile phone data

“…Several papers produce a static analysis of the relationship between mobility and disease, identifying sources and sinks of the disease without factoring in seasonal changes in mobility and disease prevalence ( Wesolowski et al, 2012 ; Chang et al, 2018 ; Ihantamalala, 2018 ; Tatem et al, 2009 ; Le Menach, 2011 ; Ruktanonchai et al, 2016 ). Seasonal variation in biological factors and seasonal population movements are important for many infectious diseases, and failing to account for seasonality could lead to misallocation of resources ( Buckee et al, 2017 ).…”