Exposure to cough aerosols and development of pulmonary COVID-19

Six months after the emergence of COVID-19 the scientific community does not agree on the relative importance of the different possible transmission routes of SARS-CoV-2. Respiratory viruses like SARS-CoV-2 are normally transmitted by contact (direct and/or indirect) or when the virus travels on tiny droplets released into the air. That's why, in the beginning of the COVID-19 pandemic it was anticipated that SARS-CoV-2 transmission started with a sneeze or cough, similarly to SARS-CoV-1. Cough aerosols contain a mix of droplets and airborne particles. Droplets remain larger than 5–10 micron while quickly falling to the ground. Airborne particles, formally called 'droplet nuclei', originate from droplets that become so small while evaporating that they remain suspended in the air. The World Health Organization believes that SARS-CoV-2 normally does not transmit through droplet nuclei, only through droplets and contact. Whether on droplets or droplet nuclei or a combination of both, a recent hospital study is suggestive for important SARS-CoV-2 transport through the air, demonstrated by high ratios of positive air and floor samples [1].

We previously reported that surgical masks worn by cystic fibrosis patients while coughing, not only stop droplets, but also block the generation of droplet nuclei containing viable Pseudomonas [2]. While a lot of air, if not most, escapes through the gaps between face and mask when coughing, the infectious particles did not [2, 3]. Our findings were confirmed by another group and very recently a study with seasonal coronaviruses reported similar pilot results [4, 5]. The behavior of virus-laden droplets and droplet nuclei in aerosols depends on the aerodynamic size of these carrier particles, not on the size of the virus itself.

We hypothesized that the transmission route of SARS-CoV-2 is a disease severity determinant and that most (more than 70%) patients with severe pulmonary COVID-19 were exposed to cough aerosols. Based on this, one could expect a significant effect of promoting cough etiquette and mass masking on COVID-19 morbidity and mortality.

We asked medical colleagues in our hospital to identify adult COVID-19 patients who were almost 100% certain that they knew from which person the virus transmitted to them, after explaining that on average people get infected with the novel coronavirus 5 d before they become ill, with a range between 2–14 d. Next, patients were asked only one question: 'Was coughing a symptom of the person from which the virus transmitted to you?'. Additionally, only information about disease severity was collected, but no patient identifiers. Confidence intervals (CI) for proportions were calculated using MedCalc^® (MedCalc Software, Ostend, Belgium). The study received a review exemption from the Ethics Committee of University Hospital Antwerp in Belgium.

Between the 18th March and the 17th April, we received 38 filled 'ONE-questionnaires', representing transmission events where an index patient could be identified with a high degree of certainty. Only 9 out of 25 hospitalized patients requiring supplemental oxygen (36%; 95% CI 18–57%) got infected by somebody who coughed (figure 1). This result does not support the hypothesis that exposure to cough aerosols is required (meaning observed in over 70% of the patients) to develop severe pulmonary COVID-19. Overall, only 14 out of 38 COVID-19 patients (37%; 95% CI 22–54%) were infected by someone who they reported had coughing as a symptom.

Zoom In Zoom Out Reset image size

Participating patients were only asked to answer one question, which is both a limitation and a strength of our study. As with any case-control study, there is a risk of recall bias and misclassification: patients may not remember coughing as a symptom of the person who infected them. Also, coughing patients may have been placed in isolation more easily, or at least have less contact time.

Many SARS-CoV-2 transmission events could result from pre- and asymptomatic index patients [6, 7]. Even though it is normal for healthy people to cough, it seems unlikely that coughing would play an important role in transmission events from infected individuals without COVID-19 symptoms. Based on this and the results from our study, we believe we may conclude that most SARS-CoV-2 transmission does not involve coughing. Alternatively, other respiratory maneuvers like sneezing, talking or breathing could have generated SARS-CoV-2 aerosols. When talking, vibration of the vocal cords generates 2.5 micron droplet nuclei [8]. Ten seconds saying 'aah' produces a comparable number of droplet nuclei as coughing for 10 s [9]. The vocal cords are being lubricated by saliva, which contains large amounts of SARS-CoV-2 in most COVID-19 patients [10]. During singing or talking with a loud voice, up to 50 times more droplet nuclei are being generated compared to talking with a quiet voice [11, 12]. Researchers also noticed that 1 in 5 people released 10 times more particles during talking than their peers for reasons not yet understood. Bringing all this information together, one could expect COVID-19 superspreader events in places where people talk loudly or sing, like bars and clubs with loud music, choirs, and crowded indoor markets. Lately these are increasingly being reported, in both non-scientific and scientific press. One could even argue that talking with a loud voice to hearing impaired elderly, could have contributed to the high COVID-19 incidence in nursing homes.

Droplets have enough mass and speed to get impacted in a mask when coughing, even the small ones that would otherwise evaporate to droplet nuclei [2]. During speech, large droplets also impact in a mask, but the effect of face masks on preventing small droplets from evaporating to droplet nuclei when just talking or breathing, could be smaller. According to Newton's second law of motion (F = m.a), less force is required to deflect the slowly moving particles during speech or breathing, which could make it easier for them to follow airflows through gaps between skin and mask. While face masks have the potential to block the generation of all kinds of respiratory aerosols, simple cough etiquette, i.e. coughing in a tissue/elbow/arm/sleeve, could only play a role in preventing transmission through coughing or sneezing.

After droplet or contact infection of the upper respiratory tract, SARS-CoV-2 could spread to the lower airways through a process of auto-inoculation, which is an alternative hypothesis for the pathogenesis of lung lesions in COVID-19 that would not involve direct inhalation of virus-laden particles. Another possible mechanism could involve hematogenous spread to the lungs.

In conclusion, our study could not confirm that exposure to cough aerosols is required for developing severe pulmonary COVID-19. Further research is needed to determine how SARS-CoV-2 spreads and to confirm the preventive role of face masks. Avoid talking when you are not wearing your mask (during lunch for example) and not talking with a loud voice, 'voice etiquette', could be other public health interventions worthwhile exploring.

None of the authors have any financial associations or other possible conflicts of interest to declare.

KVD and ED conceived the study, analyzed the data, and drafted the manuscript. JG and JN collected the data and reviewed the manuscript. All authors approved the final version of the article. KVD was the guarantor of this paper.

Not applicable.

None of the authors have any financial associations or other possible conflicts of interest to declare.