Airborne transmission via droplet nuclei has rightfully been called the “elusive pathway” of infection (1). With the exception of tuberculosis, airborne transmission was traditionally viewed as an uncommon and ephemeral phenomenon for many bacterial and viral respiratory pathogens. It remains controversial, too, with widely differing conclusions about its importance being drawn from the same studies (2, 3). The evidence to support either side of the airborne debate has been thin on the ground compared with the relative wealth of knowledge on hand hygiene and large-droplet precautions. This reflects the difficulty in sampling and growing microorganisms from the air, because they are present in very low concentrations compared with nonbiological particles, and it requires specialized equipment and a nuanced approach. For example, viable airborne Pseudomonas aeruginosa have been found in cough aerosols from patients with cystic fibrosis, with the levels related to the infectious burden. Such potentially infectious cough aerosols remain viable up to 4 m away and for up to 45 minutes (4). However, it is difficult to place these results in a meaningful context, as the infectious inoculum is unknown. This is a recurring theme in many studies of airborne transmission.
The prescient work of Dr. William F. Wells and his mentee Dr. Richard L. Riley on tuberculosis transmission in the post–World War II period (5) still underpins much of what we know about the mechanisms and modeling of airborne transmission. Despite the justifiably long shadow cast by their studies, progress has been sporadic since then. Regardless of one’s interpretation of existing data on the relative importance of airborne transmission, it is hard to argue against the need for more good-quality experimental data.
In this issue of the Journal, Kulkarni and colleagues (pp. 308–316) present results that suggest hospitalized infants with respiratory syncytial virus (RSV)-positive bronchiolitis produce appreciable quantities of airborne RSV that can be detected in room air (6). They also demonstrate that the recovered virus is capable of infecting human ciliated respiratory epithelial cells from both healthy volunteers and those with chronic obstructive pulmonary disease. Over the course of two bronchiolitis seasons, they sampled air proximate to 24 infants on an open pediatric ward, of whom nine were nursed in an open “bronchiolitis bay” for suspected RSV cases, and the remainder in ward cubicles. Sampling was also done near an additional 10 additional infants in intensive care, of whom seven were nursed in the open (six/seven ventilated) and three were in isolation (three/three ventilated). Eight infants admitted for nonrespiratory illness were used as controls. Kulkarni and colleagues took a distance-based approach and sampled 1 m from each infant, with additional samples collected at 5 and 10 m from the bronchiolitis bay (6). The effect of time on airborne RSV was also assessed for eight infants, with samples collected 2 hours after discharge (representing the typical time of arrival for the next inpatient).
Kulkarni and colleagues detected a diverse range of room air RSV levels (measured as plaque-forming units) spanning three orders of magnitude in the vicinity of RSV-positive infants, but not from the eight controls (6). In those infants that were sampled at both distances, the authors found a reduction from 1 to 5 m. No RSV was detected at 10 m. Levels were also lower 2 hours postdischarge for seven infants. The majority of plaque-forming RSV was deposited on the lower stages (<4.7 μm) of the impactor the researchers used to sample air. For that size range, up to about 10% of inhaled particles (via nose breathing) would be expected to reach the tracheobronchial airways, and about 15% would reach the alveolar region, according to standard respiratory deposition models (7).
The room air sampling results presented by Kulkarni and colleagues are certainly interesting, and a useful contribution to the scant literature (6). However, although their results are suggestive, it is impossible to conclusively link RSV in room air to the index infants. This would require sampling under conditions in which all other aerosol sources could be eliminated, which was not feasible in the operational context of their study (i.e., inpatient infants). Also, the room air ventilation rate and air movement patterns were not measured, both of which are important determinants of indoor airborne pathogen transmission (8). The fluid dynamics of air in a typical hospital environment are complex to say the least, and bear consideration in all airborne studies. The air sampling results need to be interpreted with that in mind.
The headline finding from Kulkarni and colleagues’ work is that RSV recovered from room air appears readily capable of infecting healthy and chronic obstructive pulmonary disease ciliated respiratory epithelial cells (6). This is useful new evidence suggesting that airborne transmission could be a viable transmission pathway. The authors cite data to support the dose of RSV required to induce infection as being small, although it is not possible to convert their in vitro inoculation to that occurring during natural in vivo infection. The results also do not tell us to what extent airborne transmission of RSV may have been underestimated compared with contact transmission and large droplets. However, they do provide some much-needed grist to the airborne transmission mill and suggest that focusing on contact and droplet prevention strategies for RSV may not be the whole story. The implications for reviewing patient management and infection control practices to prevent nosocomial RSV are thus important, even in the absence of a smoking gun.
Some of the key scientific questions that still need to be resolved include: How long does airborne RSV remain infectious? What are the determinants of RSV production from an individual? How much RSV is produced by breathing, talking, coughing, and sneezing by adults? Which control strategies are most effective at preventing airborne RSV transmission? These are all issues that are relevant beyond just the hospital setting.
As always, there are more questions than answers. We hope Kulkarni and colleagues’ findings motivate further studies to address these and other important questions. We hope Wells, Riley, and all who followed in their footsteps would be gratified to see that airborne transmission continues to attract the attention of researchers more than half a century later.