Despite the substantial impact of RSV globally, it is still unclear via which routes RSV is primarily transmitted and if and how long infectious virus is shed by infected individuals. Here, we determined the amount of RSV in air around hospitalized infants, in correlation with the viral load in their upper respiratory tract over time. We demonstrated that despite the presence of infectious RSV in nasopharyngeal samples of infants, only low amounts of RSV RNA, but no infectious virus, was detected in the air around three out of six patients. RSV RNA was only detected in large (> 7 µm) droplets.
For two of these patients, one or both parents also tested positive for RSV, so they may have contributed to the RSV RNA quantities collected from the air. For the third patient in whose room RSV was collected from the air, both parents were RSV negative, and therefore the RSV RNA must have been expelled by the patient.
For one patient who was co-infected with RV, low amounts of RV RNA were collected from the air on multiple days. Except for one positive sample that was recovered from stage 1 (droplets > 7 µm), the remaining three RV positive air samples were consistently recovered from stage 3 of the cascade impactor. This stage collects aerosols in the size range of 3.3 µm to 4.7 µm, indicating that at the time of air sampling RV RNA was contained in smaller particles than RSV RNA. Remarkably, on a day that the airway samples collected from the patient and both parents were negative for RV RNA, an air sample turned out to be positive. It is unclear if virus was still shed by the patient or the parents, but from an anatomical site of the respiratory tract that was not sampled, or that the air was contaminated by hospital personnel that were present that day.
Air sampling was only started a few days to one week after symptom onset during the late phase of infection, which may explain the low quantities of RSV and RV RNA collected from the air (Table 2). In several other studies, in which various air samplers were used, also low numbers of RSV RNA positive air samples were reported, with a detection rate ranging from 2.3 to 31.8 % (12–14, 16). Moreover, in a recent study by Chamseddine et al., none of the collected air samples around RSV infected patients were positive for RSV RNA, while half of the air samples collected around influenza A virus infected patients were positive for influenza virus RNA. However, attempts to isolate infectious influenza virus from these samples were not successful(15).
Contrasting results were previously reported by Kulkarni et al., where air sampling around infants with RSV-confirmed bronchiolitis in general wards of a pediatric hospital resulted in the collection of high amounts of infectious RSV from the air(17). Although viral quantities in the air decreased with increasing distance to the patient’s head, up to 105.6 plaque forming units (PFU) of infectious RSV were still collected 5 meters away from the patient’s head. As in the present study, Kulkarni and colleagues used a six-stage Andersen cascade impactor, however, liquid medium was used as a collection medium. We have recently shown in an in-vitro set-up that the collection of infectious virus using liquid medium is less efficient than when semi-solid gelatin is used (as in the current study), so this does not explain the differences in collected amounts of RSV between the studies(19). It should be noted that the Anderson cascade impactor was designed and validated with solid impaction media rather than liquids(18).
The small amounts of RSV RNA detected in large droplets and the total absence of infectious RSV in air around infected infants as presented here, and the low detection rates of RSV RNA in air in most other studies using various air samplers, indicate that transmission via the air is unlikely to be an route by which RSV spreads in the population. This observation is also supported by the fact that room sharing of RSV infected and non-infected patients did not seem to influence the risk of nosocomial infections(21). In addition, wearing gowns and gloves, and adhering to strict hygiene has been shown to reduce the risk of nosocomial RSV transmission considerably, further indicating that aerosol transmission is not efficient and possibly negligible in this context(22, 23).
To investigate the possibility of RSV transmission through fomites, we also took surface swabs of the bedrails and dataloggers on the last day of air sampling. In none of the surface swabs, RSV RNA was detected by qRT-PCR, which is in contrast to the study of Wan et al., where RSV RNA was detected on various objects(12). A reason for the conflicting results might be the timepoint when the surface swabs were taken. Wan and colleagues took surface swabs shortly after admission of patients. In our study, surface swabs were only taken on the last day of air sampling, during the late stage of RSV infection when RSV RNA levels in the patients had already decreased.