A systematic review of the data, methods and environmental covariates used to map Aedes-borne arbovirus transmission risk

Abstract

Background Aedes (Stegomyia)-borne diseases are an expanding global threat, but gaps in surveillance make comprehensive and comparable risk assessments challenging. Geostatistical models combine data from multiple locations and use links with environmental and socioeconomic factors to make predictive risk maps. Here we systematically review past approaches to map risk for different Aedes-borne arboviruses from local to global scales, identifying differences and similarities in the data types, covariates, and modelling approaches used. Methods We searched on-line databases for predictive risk mapping studies for dengue, Zika, chikungunya, and yellow fever with no geographical or date restrictions. We included studies that needed to parameterise or fit their model to real-world epidemiological data and make predictions to new spatial locations of some measure of population-level risk of viral transmission (e.g. incidence, occurrence, suitability, etc.). Results We found a growing number of arbovirus risk mapping studies across all endemic regions and arboviral diseases, with a total of 176 papers published 2002–2022 with the largest increases shortly following major epidemics. Three dominant use cases emerged: (i) global maps to identify limits of transmission, estimate burden and assess impacts of future global change, (ii) regional models used to predict the spread of major epidemics between countries and (iii) national and sub-national models that use local datasets to better understand transmission dynamics to improve outbreak detection and response. Temperature and rainfall were the most popular choice of covariates (included in 50% and 40% of studies respectively) but variables such as human mobility are increasingly being included. Surprisingly, few studies (22%, 31/144) robustly tested combinations of covariates from different domains (e.g. climatic, sociodemographic, ecological, etc.) and only 49% of studies assessed predictive performance via out-of-sample validation procedures. Conclusions Here we show that approaches to map risk for different arboviruses have diversified in response to changing use cases, epidemiology and data availability. We identify key differences in mapping approaches between different arboviral diseases, discuss future research needs and outline specific recommendations for future arbovirus mapping.

The overlapping global distribution of dengue, chikungunya, Zika and yellow fever

“…We included global raster layers associated with a range of factors hypothesised to be associated either with transmission of Aedes-borne arboviruses or with capacity to detect, diagnose and report cases. Our choice of covariates was based on a previous systematic review of arbovirus risk mapping studies 55 . Covariate data sources were prioritised by those that gave the highest spatial resolution and covered the 2010-2020 time period where most of our occurrence point data are concentrated.…”

“…We used a down sampled random forest approach which has been proven to outperform many other machine-learning approaches for the modelling of presence-only data across a range of examples 54 and has previously been used for global environmental suitability mapping applications 55 , including dengue 16,38 , chikungunya 17 , Zika 18 , and yellow fever 19 , as well as the global distribution map of Aedes vectors 20 . Specifically we selected a random forest (RF) with down-sampling approach that balances the presence and background points at a 1:1 ratio, as RF model has been shown to outperform other methods for presence-only species distribution modelling across a range of applications 54,71 .…”

“…We obtained previously published suitability maps from Messina et al 16 for dengue, Nsoesie et al 17 for chikungunya, Messina et al 18 for Zika, and Shearer et al 19 for yellow fever. These maps were chosen because they represent recent global predictions of disease risk and also utilise occurrence data and species distribution modelling approaches 55 , making a more useful assessment of the advances made in this work. The same threshold calculation above was performed for the previous suitability maps using the same set of presence and background points to convert them into binary maps.…”

Human movement and environmental barriers shape the emergence of dengue

“…Predictive models that characterise the relationships between dengue spread and its known drivers such as temperature, rainfall, and connectivity offer the best chance of inferring generalisable mechanisms from limited data and allow useful insights for future containment strategies. While dynamic models of single outbreaks have incorporated the dual effects of mobility and environmental drivers 18 , current frameworks for modelling long-term EID geographic spread have focussed on either connectivity 19 or environmental factors 20 , despite the spread of species and diseases relying on a close interaction of connectivity and environmental suitability 21 , 22 for dispersal. Environmental factors may play a greater role in directing early spread if the pathogen is already circulating among highly connected areas, while connectivity may become more important in the later stages of spread as marginally suitable areas require repeated introduction to trigger an outbreak 23 .…”