Anyin Feng

1 Introduction In long-term follow-up studies, panel count data arise when the observation times for the recurrent events of interest are discrete time points. Unlike the recurrent event data, which record the exact occurrence times of the recurrent events, the available information in panel count data is the number of recurrent events occurring between two adjacent observation times. Many statistical methods have been developed for the panel count data over the past several decades.

Counterfactual model with booster “switched off” We generated simulations from a counterfactual model, in which the third booster was removed or “switched to zero”. This was achieved by running the baseline model with the same parameters (including the same ) apart from setting the vaccine efficacy of the third dose, α, to zero. The model thus simulated the number of infections expected in the absence of the booster.