Bayesian nonparametric inference for heterogeneously mixing infectious disease models

Abstract

Significance Mathematical models of infectious disease transmission continue to play a vital role in understanding, mitigating, and preventing outbreaks. The vast majority of epidemic models in the literature are parametric, meaning that they contain inherent assumptions about how transmission occurs in a population. However, such assumptions can be lacking in appropriate biological or epidemiological justification and in consequence lead to erroneous scientific conclusions and misleading predictions. We propose a flexible Bayesian nonparametric framework that avoids the need to make strict model assumptions about the infection process and enables a far more data-driven modeling approach for inferring the mechanisms governing transmission. We use our methods to enhance our understanding of the transmission mechanisms of the 2001 UK foot and mouth disease outbreak.

Modeling foot-and-mouth disease dissemination in Brazil and evaluating the effectiveness of control measures

“…It is worth noting that FMD pathogenesis and transmission dynamics are different between species, including viral replication, persistence, incubation period, and potential to infect the same and/or different hosts with different transmission probabilities (Alexandersen et al, 2003; Arzt et al, 2011; Moreno-Torres et al, 2022; Stenfeldt et al, 2016), for example, pigs emit more viruses than cattle and sheep which increases the transmission rate when the pigs are the more infectious host when compared with other domestic species (Brown et al, 2022; Donaldson and Alexandersen, 2002; Moreno-Torres et al, 2022). Thus, it is pivotal that such heterogeneity in the transmission dynamics is considered while modeling either within or between-farm transmission dissemination (Seymour et al, 2022). Whilst, between farm transmission, primarily occurs through the direct contact among susceptible and infected animals (Alexandersen et al, 2003; van Andel et al, 2021), field observation and experimental trials demonstrated the spread of the virus via indirect contact with fomites and long-distance transport of aerosols, a process known as spatial transmission (Chakraborty et al, 2021; Guyver-Fletcher et al, 2021; Sellman et al, 2020; van Andel et al, 2021).…”

“…The most common simplification is to restrict the epidemiological models to a single species (Chanchaidechachai et al, 2021; Sseguya et al, 2022). Despite the different applications and efforts in modeling FMD, questions on measuring the epidemic trajectory and epidemic control strategies while considering heterogeneous transmission dynamics among the different susceptible species coexisting on the same farms remain outstanding (Seymour et al, 2022).…”

“…The official data from the World Organisation for Animal Health (WOAH) of more than 57 thousand outbreaks in 99 counties from 1996 until 2012 show that 68% of the cases were associated with cattle, 22% with swine, and 21% with small ruminants and other species (Santos et al, 2017). Even with strong evidence indicating that all susceptible species can drive large FMD epidemics (Bertram et al, 2020; Tildesley et al, 2012) for example, the 2001 UK epidemic was traced back to an infected Essex swine slaughterhouse (Davies, 2002) control measures and surveillance programs have mostly been developed to target dissemination within cattle populations, thus neglecting the role of other species in FMD transmission (Naranjo and Cosivi, 2013; Santos et al, 2017; Seymour et al, 2022; Singanallur et al, 2021).…”

“…For example, swine shed more viral particles than cattle and sheep, with widespread epidemics expected when swine are also infected 14 . Thus, it is pivotal to consider such heterogeneity in transmission dynamics when modeling within- or between-farm FMD dissemination 11 . While between-farm transmission primarily occurs through direct contact among susceptible and infected animals 12, 16 , field observations and experimental trials have demonstrated the spread of the virus via indirect contact with fomites and long-distance transport of aerosols, a process known as spatial transmission 16 .…”

“…A common simplification in modeling involves restricting the dynamics to a single species (4,19,20). Despite the varied applications and efforts in modeling FMD, outstanding questions persist regarding the measurement of epidemic trajectories and control strategies, particularly given the heterogeneous transmission dynamics among different susceptible species coexisting on the same premises (8,18).Here, we develop a multihost, single-pathogen, multiscale model designed to capture the dynamics of FMD transmission patterns across different host species. The model i) simulates the spread of FMD within the state of Rio Grande do Sul in Brazil; ii) describes the geodesic distances between the initial outbreak and secondary cases; and iii) compares the effects of control action strategies, including emergency vaccination, depopulation, various restrictions on between-farm movement, and surveillance activities within control zones, taking into account the initial number of infected farms at the onset of control measures.…”