Explaining rapid reinfections in multiple-wave influenza outbreaks: Tristan da Cunha 1971 epidemic as a case study

Camacho, Anton; Ballesteros, Sébastien; Graham, Andrea L.; Carrat, Fabrice; Ratmann, Oliver; Cazelles, Bernard

In this study, we assessed and compared six potential antigenic and immunological drivers of multiple-wave influenza A outbreaks on a two-wave influenza A/H3N2 epidemic that occurred on the island of TdC in 1971. We translated these hypotheses unambiguously into six mechanistic stochastic models, and employed a rigorous statistical framework based on ML [20] for parameter inference and model selection. In addition, we performed complementary statistical analyses, based on extensive simulations, to evaluate and compare the goodness of fit of the predictions of our six models. Our findings emphasize that a stochastic formulation is essential to capture demographic stochasticity induced by small populations [24] and/or low-prevalence inter-wave periods. We show that two mechanisms-both invoking host heterogeneity rather than viral heterogeneity-are significantly better supported by the data. Both mechanisms challenge the efficiency of the human immune response following primary influenza infection, indicating that, after a first attack, some individuals with delayed (Win) or deficient (AoN) humoral immune response could be reinfected by the same strain.