Detection of influenza and other respiratory viruses in air sampled from university campus : a longitudinal study

Abstract

Influenza virus-laden particles have been reportedly detected in exhaled breath of symptomatic influenza patients or in the air at health care settings, demonstrating the feasibility of airborne transmission of influenza viruses. However, little knowledge is available on the temporal frequency of detecting airborne particles at public locations, which may serve as key influential nodes in transmission networks during epidemics. To fill this research gap, a longitudinal study was performed in Hong Kong, a city with subtropical climate, to monitor the temporal trend of influenza and other common human respiratory viruses in air sampled from a university campus where young adults converge daily.

From September 2016 to April 2017 and from September 2017 to April 2018, air was sampled weekly from lecture halls, shuttle buses, canteens, and the University Health Service on campus. Parameters including indoor temperature, relative humidity (RH), ventilation rate, and the number of occupants were recorded at each sampling event. Airborne viral particles were collected by the NIOSH bioaerosol sampler (BC251) that separates particles into three size fractions at > 4 μm, 1-4 μm, and < 1 μm and were tested by real-time quantitative RT-PCR to detect influenza A and B viruses, human rhinovirus (HRV), respiratory syncytial virus (RSV), and human coronaviruses (HCoV) OC43 and 229E. Influenza A (17.03%, 109/640) and B (4.84%, 31/640) viruses were the most frequently detected respiratory viruses when compared to HRV (2.22%, 6/270), RSV (0.37%, 1/270), and HCoV (0%, 0/152). In addition, infectious influenza A (0.94%, 6/640) and B (0.16%, 1/640) viruses were detected at low frequency in airborne particles.

A temporal correlation was noted when analyzing influenza activity in the community (eg. weekly influenza positive rate among respiratory specimens) and the weekly detection frequency of airborne influenza-laden particles at campus. Influenza epidemic strains can be detected in air sampled from campus prior to the peak of the epidemics but not afterwards. Human density and sampling performed on teaching days were the most important factors associated with the detection of airborne particles. Increased indoor temperature and increased indoor-outdoor RH differences were also found as significant risk factors by univariate and multivariate logistic regression models.

Compared to other public locations such as the transport systems, our study is biased towards the young-adult age group and the detection of respiratory viruses this population is susceptible of. We showed that influenza and other respiratory viruses were detected at different frequencies in air, which may have implications on their airborne transmission potential. Influenza A and B viruses were detected at a moderate frequency from airborne particles. In particular, epidemic strains can be detected in air prior to the peak of influenza activity observed in the community, which highlights the potential of applying air sampling as an easy-implemented surveillance tool at public locations. Factors associated with the detection of airborne particles provide insights for improving or implementing environmental interventions to reduce the airborne transmission risk of influenza.