Mechanistic modeling of African swine fever: A systematic review

TL;DR

This systematic review analyzes mechanistic models of African swine fever, highlighting their objectives, limitations, and the need for integrated, multi-route models to improve control strategies and decision-making.

Contribution

The paper provides a comprehensive review of existing ASF mechanistic models, identifying gaps such as host specificity and lack of combined control strategy assessments.

Findings

Models vary widely in transmission parameter estimates

Most models focus on either domestic pigs or wild boar, not both

Current models assess control strategies in isolation rather than synergistically

Abstract

The spread of African swine fever (ASF) poses a grave threat to the global swine industry. Understanding transmission dynamics, such as through mechanistic modeling, is critical for designing effective control strategies. Articles were examined across multiple epidemiological and model characteristics. Model filiation was determined through creation of a neighbor-joined tree using phylogenetic software. Of 34 four articles qualifying for inclusion, four main modelling objectives were identified: estimating transmission parameters (11 studies), assessing determinants of transmission (7), examining consequences of hypothetical outbreaks (5), and assessing alternative control strategies (11). Estimated transmission parameters varied widely as did parameter assumptions between models. Uncertainties on epidemiological and ecological parameters were usually accounted for to assess the impact on the modeled infection trajectory. Almost all models are host specific, being developed for either domestic pigs or wild boar despite the fact that spillover events between domestic pigs and wild boar are evidenced to play an important role in ASF outbreaks. The development of models incorporating such transmission routes is crucial. All compared control strategies were defined a priori, and future models should be built to identify optimal contributions across many control methods. Further, control strategies were examined in competition, opposed to how they would be synergistically implemented. While comparing strategies is beneficial for identifying a rank-order efficacy of control methods, this structure does not necessarily determine the most effective combination of all available strategies. In order for ASFV models to effectively support decision-making in controlling ASFV globally, these modelling limitations need to be addressed.