Bayesian estimation of the effective reproduction number for pandemic influenza A H1N1 in Guangdong Province, China

Abstract

PurposeDuring the course of a pandemic, it is necessary to understand its transmissibility, which is often summarized by the effective reproduction number. Accurate estimation of the effective reproduction number (R) is of vital significance in real-time decision making for coping with pandemic influenza. MethodsWe used daily case notification data in Guangdong Province, China, in conjunction with Bayesian inference of two different stochastic susceptible, infectious, recovered (SIR) models to estimate the effective reproduction number. The duration of infectiousness was taken from published literature, and the proportion of imported cases was obtained from individual-level data. ResultsAt the initial epidemic phase, 40% of the first 261 cases were not locally acquired. Explicitly accounting for imported cases and different infectious periods, the possible range of basic reproduction number was preliminarily estimated to be between 1.05 and 1.46. We showed how the daily case reports provided valuable information to estimate the effective reproduction number. We also found the potential delay in reporting had a relatively minor impact on estimating R. ConclusionsOur proposed models and findings provide a relevant contribution towards establishing a basis for monitoring the evolution of emerging infectious diseases in real time and understanding the characteristics of pandemic influenza A H1N1 in Guangdong Province.