Quantifying model evidence for yellow fever transmission routes in Africa

Abstract

Yellow fever is a vector-borne disease endemic in tropical regions of Africa, where 90% of the global burden occurs, and Latin America. It is notoriously under-reported with uncertainty arising from a complex transmission cycle including a sylvatic reservoir and non-specific symptom set. Resulting estimates of burden, particularly in Africa, are highly uncertain. We examine two established models of yellow fever transmission within a Bayesian model averaging framework in order to assess the relative evidence for each model’s assumptions and to highlight possible data gaps. Our models assume contrasting scenarios of the yellow fever transmission cycle in Africa. The first takes the force of infection in each province to be static across the observation period; this is synonymous with a constant infection pressure from the sylvatic reservoir. The second model assumes the majority of transmission results from the urban cycle; in this case, the force of infection is dynamic and defined through a fixed value of R0 in each province. Both models are coupled to a generalised linear model of yellow fever occurrence which uses environmental covariates to allow us to estimate transmission intensity in areas where data is sparse. We compare these contrasting descriptions of transmission through a Bayesian framework and trans-dimensional Markov chain Monte Carlo sampling in order to assess each model’s evidence given the range of uncertainty in parameter values. The resulting estimates allow us to produce Bayesian model averaged predictions of yellow fever burden across the African endemic region. We find strong support for the static force of infection model which suggests a higher proportion of yellow fever transmission occurs as a result of infection from an external source such as the sylvatic reservoir. However, the model comparison highlights key data gaps in serological surveys across the African endemic region. As such, conclusions concerning the most prevalent transmission routes for yellow fever will be limited by the sparsity of data which is particularly evident in the areas with highest predicted transmission intensity. Our model and estimation approach provides a robust framework for model comparison and predicting yellow fever burden in Africa. However, key data gaps increase uncertainty surrounding estimates of model parameters and evidence. As more mathematical models are developed to address new research questions, it is increasingly important to compare them with established modelling approaches to highlight uncertainty in structures and data.

The effect of climate change on Yellow Fever disease burden in Africa

“…This meant that we adapted our previous best-fit model (Garske et al 2014) to include temperature range, temperature suitability and precipitation rather than enhanced vegetation and landcover. This change slightly reduced fit quality, giving an AUC of 0.9004 as opposed to to 0.9157 (Gaythorpe et al 2019). Vegetation is a key factor determining habitat of non-human primates, an element that may not be captured by the temperature suitability index which focuses on the vector Ae.…”

“…We reformulate an established model of YF occurrence to accommodate nonlinear dependence on temperature and rainfall. We couple this with established results from a transmission model of serological status to estimate transmission intensity across the African endemic region at baseline/ current environmental conditions (Garske 4 et al 2014;Jean et al 2018;Gaythorpe et al 2019). Then, we project transmission intensity for four climate scenarios given projected changes in temperature and rainfall.…”

“…We utilise previously estimated models of seroprevalence informed by serological survey data, demography and vaccination coverage information (Garske et al 2014;Gaythorpe et al 2019). The transmission intensity 5 is assumed to be a static force of infection, akin to the assumption that most YF infections occur as a result of sylvatic spillover (Garske et al 2014;Gaythorpe et al 2019). The force of infection is assumed to be constant in each province over time and age.…”

“…Jean et al 2018;Gaythorpe et al 2019); and incorporate covariates indicative of climate change that also have projections available in years 2050 and 2070 for different emission scenarios. As such, we omit enhanced vegetation index and land cover from the best fitting model ofGarske et al (2014) in favour of the temperature suitability index which depends on the average temperature, the temperature range and average rainfall.…”

Eliminating yellow fever epidemics in Africa: Vaccine demand forecast and impact modelling

“…Combining estimates of model variants that rely on different assumptions and structures in a way that can reliably inform decision making remains a challenge. Bayesian model averaging methods have been suggested to better combine such estimates [24].…”