Simulating the early mpox outbreak: Dynamic-spread assessment via vSEIR
model and kink detection in disease transmission
Abstract
This paper proposes a varying coefficient
Susceptible-Exposed-Infected-Removed (vSEIR) model to dynamically
simulate the early mpox epidemic that sparked panic in 2022, considering
the time-varying infection rate and the group protected by the smallpox
vaccination. We apply the recursive least squares algorithm with a
forgetting factor for real-time identification of time-varying infection
rates and the efficacy of non-pharmacological interventions. The sparse
Hodrick-Prescott (HP) filter, tuned with leave-one-out cross-validation,
captures mpox epidemic kinks via the effective reproduction number R t
obtained from the discrete vSEIR model. We experiment with this approach
in Brazil, Spain, UK and US, comparing COVID-19 and mpox outbreaks based
on those kinks and transmission cycles, identifying that except for
Spain, mpox epidemic reached its decline period earlier than COVID-19
without strong interventions. Additionally, the result regarding
sensitivity analyses shows that the total number of mpox outbreak
infections would have increased by 12% without smallpox vaccination and
the data uncertainty can bring great variations in R t .