Statistical Inference of Lassa Fever Transmission Dynamics from Routine Surveillance Data

Lassa fever, a zoonotic viral disease endemic in Nigeria, periodically erupts, causing considerable morbidity and mortality. To estimate the dynamics of its transmission and aid in its control, we require an understanding of the underlying dynamics. Yet, in most cases, the missing information pertains to the hidden reservoir dynamics. Accordingly, we applied the stochastic renewal equation model parameterized for the number of confirmed cases and deaths in Nigeria during weeks 1 through 48 of the year 2025 to investigate the dynamics of its transmission over time. Our model also incorporates seasonality through a periodically forced reproduction number and overdispersion in observation through an overdispersed count process. Through the use of a biologically motivated delay linking incidence and mortality rates, we can reconstruct both from a single surveillance effort. Our fit does exhibit strong seasonality in transmission, with a baseline reproduction number of 1.06 and frequent crossings of the epidemic threshold. In our stochastic simulations, there is a great deal of short-term variability even in the baseline scenario. In our scenario analyses, a reduction in transmission of 20-40% can lead to a significantly decreased total number of cases. This study shows that incidence-based renewal models have proved to be a very valuable method for the investigation of endemic zoonotic diseases such as Lassa fever.