What risk do aerosol‐generating procedures pose to health‐care workers?

This is the authors' abstract. We don't add key points for this paper.

In the COVID-19 pandemic, understanding SARS-CoV-2 transmission risk in the health-care setting is critical, as is ensuring appropriate and safe use of personal protective equipment (PPE). Given the emerging nature of this virus, understanding of SARS-CoV-2 is evolving in real time, causing uncertainty amongst health-care workers (HCWs) with particular concern around risks of workplace transmission. Much debate has centred around aerosol-generating procedures (AGPs) which may enhance transmission of the virus from an infected patient to surrounding patients and staff. AGPs are health-care procedures resulting in the production of aerosols which may have the capacity to transmit infection.1 Table 1 provides a comprehensive list. There are other procedures or behaviours that may generate aerosols but for which increased risk of transmission has not been demonstrated. While the degree of risk associated with different AGPs is likely to vary, it is generally accepted that intubation poses the highest-risk of these procedures.2 The exact mechanism is unclear, however one might postulate that transmission of virus during intubation is more likely to be due to close contact with the respiratory tract and exposure to droplets containing a high viral load, rather than exposure to aerosols per se. Some experts believe better terminology for AGPs would be ‘increased transmission risk’ procedures. This more accurately reflects the reason driving current interest amongst HCWs in these procedures. Attributing risk to any particular procedure is difficult because procedures are not performed in isolation. We can therefore only estimate risk of transmission of infection from AGPs by studying aerosol and virus characteristics, but this has limited meaning in the clinical setting. A more rational approach would be to review virus transmission dynamics. Definitive transmission data on SARS-CoV-2 is still accumulating, but its behaviour under experimental conditions has similarities to SARS-CoV-1.3 Community transmission parameters are also comparable to other droplet-based respiratory viruses. Thus, data on transmission dynamics of SARS-CoV-1 may be reasonably extrapolated to SARS-CoV-2, whilst awaiting robust studies on the latter. A systematic review suggested that certain procedures capable of generating aerosols were associated with an increased risk of SARS-CoV-1 transmission from infected patients to HCWs.4 The review did not, however, enable authors to conclude with certainty that HCWs caring for patients undergoing AGPs were at higher risk of contracting infection compared to those caring for patients not undergoing AGPs.4 While there remains a gap in evidence on the infectivity of SARS-CoV-2 associated with AGP's, there are case reports that support limited airborne transmission and experimental evidence that supports similar aerosol characteristics between SARS-CoV-1 and SARS-CoV-2.3, 5, 6 During the current pandemic there are some reassuring early observations. The single study from China that suggested risk to HCWs was highest in areas where AGPs are performed is confounded by self-reported poor hand hygiene amongst these same HCWs.7 Case reports of HCWs exposed to patients with COVID-19 undergoing AGPs without recommended PPE documented minimal, if any, nosocomial infection.8-10 When PPE is used, data are encouraging. Small studies from Italy, Singapore and the UK show either no transmission of disease to HCWs wearing PPE (despite performing AGPs on infected patients) or no increased risk of transmission of infection to HCWs performing AGPs compared to HCWs with no patient contact.9, 11, 12 Data from the 2003 SARS epidemic demonstrated that – with appropriate use of PPE, patient risk stratification and re-organisation of workflow processes where AGPs were performed – risk of transmission of infection to HCWs was low.13 Available information, including a systematic review and meta-analysis by Chu et al., suggests that the current systems, processes, PPE and training implemented in our workplace settings are effective at protecting HCWs from SARS-CoV-2.1, 14, 15 The notion that viral shedding of SARS-CoV-2 may peak at or just prior to symptom onset and the observation that hospital care tends to be needed a week or so into symptomatic disease suggests that hospitalised patients undergoing AGP's may not be the cohort of patients that pose the greatest risk in the health-care setting.16 AGPs aside, based on current published information, it appears that the rate of HCW workplace acquisition of SARS-CoV-2 is no greater than the HCW community acquisition. This is true in settings where the prevalence of disease is low as well as in high-disease burden settings where there is increased pressure on the health-care system.17, 18 Early Australian data demonstrate that SARS-CoV-2 is uncommon in HCWs and that most Australian HCWs who have contracted infection did so outside of the workplace setting.19 There is limited evidence that AGPs can transmit respiratory coronaviruses to HCWs who do not wear appropriate PPE. Existing PPE protocols provide an additional level of protection supported by studies from viruses with comparable transmission dynamics.20 Meanwhile, there is reasonable evidence that HCWs who adhere to protocols in the workplace, including meticulous hand hygiene and recommended PPE use, are no more likely to acquire infection at work than in the community. Infection control efforts should emphasise comprehensive training and careful adherence to PPE recommendations. Furthermore, reframing PPE discussions around ‘increased transmission risk’ procedures rather than aerosol-generation per se as the key risk metric will enable more productive and cooperative efforts to minimise transmission in the health-care setting. Further evidence in this area will be welcomed to support guideline development that ensures both staff safety and quality patient care during this pandemic.