Background: Different interventions targeting live poultry markets (LPMs) have been applied in China for controlling the avian influenza A (H7N9), including LPM closure and "1110" policy (i.e., daily cleaning, weekly disinfection, monthly rest day, zero poultry stock overnight). However, the effects of these interventions have not been comprehensively assessed.
Methods: Based on the available data (including reported cases, domestic poultry volume, and climate) collected in Guangdong Province between October 2013 and June 2017, we developed a new compartmental model that enabled us to infer H7N9 transmission dynamics. The proposed model incorporated the intrinsic interplay among humans and poultry as well as the effects of absolute humidity and LPM intervention, in which different intervention strategies were parameterized and estimated by Markov chain Monte Carlo (MCMC) method.
Results: There were 258 confirmed human H7N9 cases in Guangdong Province during the study period. If without interventions, the number would reach 646 (95%CI, 575-718) cases. The temporal, seasonal and permanent closures of LPMs can substantially reduce transmission risk, which might respectively reduce human infections by 67.2% (95%CI, 64.3%-70.1%), 75.6% (95%CI, 73.8%-77.5%), 86.6% (95%CI, 85.7-87.6%) in total four epidemic seasons, and 81.9% CI(95%, 78.7%-85.2%), 91.5% (95%CI, 89.9%-93.1%), 99.0% (95%CI, 98.7%-99.3%) in the last two epidemic seasons. Moreover, implementing the "1110" policy from 2014 to 2017 would reduce the cases by 34.1% (95%CI, 20.1%-48.0%), suggesting its limited role in preventing H7N9 transmission.
Conclusions: Our study quantified the effects of different interventions and execution time toward LPMs for controlling H7N9 transmission. The results highlighted the importance of closing LPMs during epidemic period, and supported permanent closure as a long-term plan.