The African swine fever modelling challenge: Model comparison and lessons learnt

Area of Science:

• Veterinary Epidemiology
• Mathematical Modelling
• Disease Ecology

Background:

• Effective epidemic management requires robust epidemiological knowledge and predictive modeling tools.
• Multiple modeling approaches can support timely decision-making during disease outbreaks.
• Understanding the strengths and weaknesses of different models is crucial for advancing the field.

Purpose of the Study:

• To present the results of the first animal health modeling challenge, the ASF Challenge.
• To compare the predictive performance of five independent modeling teams for African swine fever (ASF).
• To assess models' ability to predict epidemic spread and prioritize interventions at the domestic pig-wild boar interface.

Main Methods:

• A synthetic African swine fever epidemic on an island was simulated.
• Five independent international teams developed and submitted their modeling approaches.
• Model predictions of temporal and spatial epidemic expansion were compared against simulated data.
• The prioritization of control interventions by the models was also assessed.

Main Results:

• Model performance varied by challenge phase, host species, and whether spatial or temporal dynamics were predicted.
• Ensemble models, combining predictions from all teams, outperformed individual models in at least one phase.
• Accounting for the livestock-wildlife interface was identified as critical for effective disease control.

Conclusions:

• The ASF Challenge highlighted the importance of integrating domestic animal and wildlife dynamics in disease modeling.
• Model comparison improved the scientific community's readiness for future African swine fever epidemics.
• Lessons learned from the challenge provide guidance for decision-making in animal disease control.