Abstract
Seroepidemiology has significant potential for uncovering the unreported burden of infectious diseases. However, for diseases without well-defined serological correlates of protection, natural immune boosting-whereby pathogen exposure triggers a detectable immune response without causing a transmissible infection-can complicate the interpretation of serosurveys. This issue is relevant to pertussis, a vaccine-preventable disease that remains a significant public health concern worldwide. Here, we aimed to evaluate the reliability of pertussis serosurveys using a transmission model that tracked the dynamics of pertussis infection, natural immune boosting, and seroprevalence. By fitting this model to seroprevalence data from the late whole-cell pertussis vaccine era in six European countries, we estimated that protection against infection conferred by natural infection or vaccination was variable but lasted, on average, for several decades. We then predicted the positive predictive value (PPV) of seropositivity in serosurveys among adults across twelve countries that broadly captured transmission patterns worldwide. Overall, we predicted a low PPV across multiple scenarios, especially in adults aged 20-39 years, where it typically dropped below 50%. Thus, although serosurveys are unquestionably useful for quantifying pertussis exposure levels, the common interpretation of seroprevalence as a measure of recent infections may lead to overestimating pertussis circulation and underestimating the impact of pertussis vaccines.