• S Xiao, J W Tang, D S Hui, H Lei, H Yu, and Y Li.
• Department of Mechanical Engineering,The University of Hong Kong,Hong Kong SAR,China.
• Epidemiol. Infect. 2018 Jul 1; 146 (9): 1114-1122.

AbstractInfluenza is a long-standing public health concern, but its transmission remains poorly understood. To have a better knowledge of influenza transmission, we carried out a detailed modelling investigation in a nosocomial influenza outbreak in Hong Kong. We identified three hypothesised transmission modes between index patient and other inpatients based on the long-range airborne and fomite routes. We considered three kinds of healthcare workers' routine round pathways in 1140 scenarios with various values of important parameters. In each scenario, we used a multi-agent modelling framework to estimate the infection risk for each hypothesis and conducted least-squares fitting to evaluate the hypotheses by comparing the distribution of the infection risk with that of the attack rates. Amongst the hypotheses tested in the 1140 scenarios, the prediction of modes involving the long-range airborne route fit better with the attack rates, and that of the two-route transmission mode had the best fit, with the long-range airborne route contributing about 94% and the fomite route contributing 6% to the infections. Under the assumed conditions, the influenza virus was likely to have spread via a combined long-range airborne and fomite routes, with the former predominant and the latter negligible.

21 keywords...

hide…