Serological surveillance contributes valuable information about the circulation of infectious diseases and can often provide reasonable estimates of the relative size and timing of epidemics. Given that that antibody titers can decay or boost through time, however, and that seroprevalence typically indicates past exposure for only a sample of the population, the uses of serology data beyond simple surveillance have been relatively unexplored. Thus, quantitative methods are needed to integrate observation processes with antibody decay and boosting mechanisms to provide a comprehensive understanding of the link between serology data and underlying disease dynamics. Here, we develop a mechanistic model of quantitative serological dynamics for Leptospira in California sea lions (Zalophus californianus) at the population scale and show that surveillance data can be scaled to capture total incidence of infection.
Results/Conclusions
We analyze a 15-year time series of seroprevalence surveys in stranded California sea lions, combined with longitudinal measurements within individual animals, to show that the total leptospirosis incidence in the population ranges from <1% to 10% of individuals per month and that high antibody titers may be observed in individuals for multiple months. We anticipate that this modeling approach will prove beneficial for a multitude of infectious diseases and can ultimately be linked with additional data streams to estimate case fatality rates and population-wide transmission potentials.