Aerosol generating procedures: are they of relevance for transmission of SARS-CoV-2?

It is now generally accepted that SARS-CoV-2 can be spread by aerosols as well as larger droplets from the upper respiratory tract, although the relative importance of aerosol transmission remains incompletely answered.¹ Despite this, current UK infection control guidance for hospitals is centred on the premise that aerosols are only generated by specific medical interventions designated as aerosol generating procedures (AGPs).² This draws from epidemiological observations during the 2003 outbreak of severe acute respiratory syndrome, during which certain procedures appeared to be associated with an increased risk of staff infection (particularly tracheal intubation), and these procedures had a theoretical risk of viral aerosolisation.³ However, the evidence supporting aerosolisation during these procedures was, before the pandemic, remarkably slim, with aerosolisation being assumed on the basis of the precautionary principle and low quality mechanistic studies.⁴

This view of aerosol generation subsequently led to a dichotomisation—later codified in international guidance²—that categorised all medical activities into either AGPs, where potentially infectious aerosols are generated, versus everything else, where the risk of potentially infectious aerosol is presumed to be negligible. The logical extension of this dichotomy has resulted in health-care workers in many countries undertaking interventions classified as AGPs wearing higher levels of personal protective equipment (PPE), such as FFP3 or N95 masks, whereas those health-care workers providing other medical care have not been afforded the same protection, as infectious aerosol is not considered a risk outside of AGPs.⁵

Although this dichotomy was reasonable at the start of the pandemic, recent aerosol sampling studies from multiple groups investigating several currently defined AGPs have revealed more information on the potential risk of aerosols from these procedures. In fact, the aerosol emissions from tracheal intubation, high-flow nasal oxygen, and non-invasive ventilation are low, with similar sampled aerosol concentrations to tidal breathing and speaking.6, 7, 8, 9

Critically, these aerosol studies also confirm that coughing by both healthy volunteers, and patients (with or without COVID-19), generates orders of magnitude more aerosol than many AGPs.6, 7, 8 Despite this, the generation of a cough during a procedure is not considered sufficient for that procedure to be designated an AGP. Consequently, current UK infection control guidance advises the highest aerosol precaution PPE for procedures that are demonstrably not high risk (compared with coughing) and advises lower grade droplet precaution PPE when infectious aerosol risk is high (eg, caring for coughing patients with confirmed COVID-19 for a prolonged period of time in poorly ventilated health-care settings).

Although we and others have sought to quantify aerosol generation, it should also be clear that a simple observation of increased aerosol emission does not confirm the potential for pathogen transmission; the much more exacting task of quantifying airborne infectious SARS-CoV-2 concentrations carried in aerosols has so far proven intractable.

More research is ongoing to quantify other AGPs across a range of clinical settings. However, based on the research to date, a coughing patient with acute COVID-19 is likely to generate more infectious aerosol than many AGPs. This appears to be supported by the epidemiological evidence, which points to an increased risk of infection for ward medical staff (who care for acutely dyspnoeic, coughing patients with COVID-19) compared with intensive care staff—although it should be noted that interpretation of that data is confounded by patient mix, among other factors.¹⁰

We propose an end to the term aerosol generating procedure, as it is neither accurate (aerosol is not generated above a cough for many of these procedures), implies aerosol emission is only from specific procedures (rather than being generated during normal respiratory events), potentially misidentifies the source of infection risk, and applies a binary definition to a situation that is more complex. Instead, we propose that clinicians follow an evidence-based framework that accounts for the major drivers of risk, with a focus on physical exposure to patients with suspected or confirmed COVID-19 as the critical component (panel ).

Panel. Proposed factors to be included in risk matrix for respiratory transmission of SARS-CoV-2.

Patient risk (by far the largest risk factor)

The probability of the patient having the infection, and time since acquisition. Risk based on symptoms, PCR positivity, and vaccination status. Note symptoms such as coughing and increased respiratory exertion are likely to be major factors in aerosol generation.

Duration of exposure

The duration that the risk is in place. The length of time required to be in close proximity naturally increases risk of both aerosol and droplet transmission.

Health-care practitioner risk from COVID-19

Age, sex, body mass index, comorbidities, vaccination status.

Proximity risk

Exposure to any care intervention requiring close patient contact increases risk. This includes personal care (such as mouthcare) and physical examination (especially relating to upper respiratory tract such as nasal or throat swab, nasendoscopy, or intubation)

Environmental risk

Ventilation, humidity, temperature

Subsequently, additional factors known to be relevant in viral transmission, such as ventilation, proximity, and duration of exposure to patients, should be included in assessing risk, while recognising the changing epidemiology of infection with setting.

In summary, it is increasingly clear that transmission of SARS-CoV-2 via aerosol is possible and might represent a significant transmission route. However, emerging evidence indicates that many currently defined AGPs are unlikely to play any significant role in generation of infectious aerosol that poses a risk to staff. In view of this, the term AGP has neither face validity nor construct validity. Instead, we should focus on the risk in plain sight: close, physical exposure to people suspected, or known to have, COVID-19 for prolonged time or where ventilation remains poor.

© 2021 Usaid Design Cells/SPL

We declare no competing interests.