I am working hard on updating all the evidence around “aerosol generating medical procedures” or AGMPs, but I think this is a failed and probably harmful paradigm. I will still complete my write-ups because the terminology is so widely used, but I think we would all be better off if we just abandoned the terminology of “aerosol generating medical procedure”. Patients produce aerosols constantly. COVID-19 is regularly spread through aerosols. Procedures aren’t needed to produce aerosols, and the production of aerosols has little to do with what makes a procedure high risk for disease transmission. The sooner we abandon this failed paradigm the better.
The “aerosol generating medical procedure” paradigm is based on the (often unstated) assumption that patients do not produce aerosols unless an “AGMP” is being performed. This is simply untrue. A single cough will produce hundreds to thousands of aerosols. (Nicas 2005; Morawska 2006; Fiegel 2006; Brown 2020) Sneezing can produce a few hundred thousand to a few million aerosols. (Nicas 2005; Morawska 2006; Fiegel 2006) Even normal breathing produces aerosols, and numbers increase with heavy breathing, talking, and shouting. (Tellier 2009; Morawska 2006; Fiegel 2006; Xie 2007; Asadi 2019; Stadnytskyi 2020)
In addition to just physically producing aerosols, there are a large number of studies that clearly demonstrate that viable pathogens (including tuberculosis, pseudomonas, and influenza) are frequently found in the aerosols produced by coughing patients. (Fennelly 2004; Fennelly 2012; Patterson 2018; Theron 2020; Wainwright 2009; Knibbs 2014; Lindsley 2010; Lindssley 2012; Bischoff 2013; Gralton 2013) The total number of viral particles produced in influenza varies drastically from patient to patient – explaining the super-spreader phenomenon we see in COVID-19 – but there is no doubt that coughing produces infectious aerosols. Viable pathogens have also been consistently found in aerosols produced by normal breathing. (Gralton 2013; Fabian 2008; Lowell 2013; Yan 2018; Leung 2020)
The phrase “aerosol generating medical procedure” sounds scientific without actually having a scientific basis. As I outlined in the original post on “aerosol generating procedures”, there is very little evidence that most of these procedures actually generate aerosols. For example, one study measuring aerosol levels in patient rooms found no change from baseline with any of non-invasive positive pressure ventilation, oxygen therapy, or chest physiotherapy (there was an increase with nebulization). However, the baseline aerosol concentration in patient rooms was well over 1,000,000 per cubic meter. (Simmonds 2010) In other words, patients are constantly producing aerosols, but “aerosol generating procedures” don’t seem to.
Similarly, intubation is widely recognized as one of the highest risk procedures for disease transmission. However, a standard uncomplicated intubation doesn’t actually create the physical conditions generally required to produce aerosols. In one operating room study, a single volitional cough produced aerosol concentrations that were 500 times higher than the concentrations produced while intubating patients. (Brown 2020)
Patients produce aerosols. “Aerosol generating medical procedures” don’t actually seem to.
The entire AGMP paradigm is a distraction from what we really care about. We want to know which scenarios put healthcare workers at the highest risk of getting sick, and what we can do to protect them.
It should have been very clear after SARS that the theory of “aerosol generating medical procedures” was irrational. Many of the activities that placed healthcare workers at the highest risk of contracting SARS had absolutely nothing to do with aerosols. Some of the highest risk procedures for transmission to healthcare workers included inserting a urinary catheter and placing ECG leads. (Loeb 2004; Raboud 2010) In a case control study, just having pulmonary congestion, requiring any oxygen therapy, and increasing disease severity were associated with increased transmission of the disease. (Yu 2007) None of this can be explained within the “AGMP” model.
“Aerosol generating procedure is a misleading term, and its use probably leads to overestimation of risk in stable patients while proved aerosol generating activities such as coughing and talking are neglected.” (Wilson 2020)
A more useful paradigm might be “prolonged close contact”. We know that all COVID-19 patients produce aerosols, but most encounters are actually relatively low risk. The activities that place healthcare workers (or anyone) at risk seem to be those that keep them in close contact with COVID positive patients for long periods of time. Although not universally true, the longer one stays in close contact with a patient the higher their risk will be of contracting COVID-19. This is why nurses contract the disease more often than doctors. This is one reason why large numbers of PSWs have become ill in nursing homes. The procedure is probably much less relevant than the total time spent exposed to the patient.
We also need to consider the spectrum of patients. Patients in respiratory distress produce more aerosols. Coughing patients produce more aerosols. The sicker a patient is, the more likely I am to require an N95, even for short clinical encounters. Unfortunately, this approach is clearly imperfect as asymptomatic spread plays a major role in COVID-19, and most super-spreader events occur with pre-symptomatic patients. (Although that fact is biased by the fact that symptomatic patients are generally isolated, and so have fewer opportunities to act as super-spreaders.) We clearly need more science to determine the highest risk patients.
How does that change our approach? I am obviously still wearing an N95 during intubation, but we need to acknowledge the many other high risk scenarios that don’t currently fall under the AGMP umbrella. For example, if you are going to spend 30 minutes at the bedside of a palliative COVID patient requiring a nonrebreather, you should probably be wearing an N95 even if you aren’t going to be performing any “aerosol generating procedures”. On the other hand, if I am doing a 3 minute assessment of a well patient with a fever and a cough, I am probably going to be OK with just a surgical mask and normal droplet contact precautions.
“These data suggest that health-care workers should be protected from these potentially infectious aerosols when working in close proximity to patients.” (Fennelly 2020)
In an ideal world, we would use airborne precautions for all COVID-19 suspected patients. That is how we managed SARS and MERS. It is probably the only way to completely eliminate risk. However, we don’t work in a perfect world, and PPE shortages are a reality. There is a risk of contracting COVID-19 when seeing an average patient with just a surgical mask, but that risk is very small. In our imperfect world, it makes sense to reserve higher level PPE for the encounters that place us at the highest risk. Those encounters aren’t well defined by the faulty “aerosol generating medical procedure” paradigm. All patients produce aerosols. We need better science to guide these decisions, but for now “prolonged close contact” is probably much safer, and much better supported by science than “aerosol generating medical procedure”.
Asadi S, Wexler AS, Cappa CD, Barreda S, Bouvier NM, Ristenpart WD. Aerosol emission and superemission during human speech increase with voice loudness. Sci Rep. 2019;9(1):2348. Published 2019 Feb 20. doi:10.1038/s41598-019-38808-z PMID: 30787335
Bischoff WE, Swett K, Leng I, Peters TR. Exposure to Influenza Virus Aerosols During Routine Patient Care . 2013; 207(7):1037-1046.
Brown J, Gregson FKA, Shrimpton A, Cook TM, Bzdek BR, Reid JP, Pickering AE. A quantitative evaluation of aerosol generation during tracheal intubation and extubation. Anaesthesia. 2020 Oct 6:10.1111/anae.15292. doi: 10.1111/anae.15292. Epub ahead of print. PMID: 33022093
Fabian P, McDevitt JJ, DeHaan WH, et al. Influenza Virus in Human Exhaled Breath: An Observational Study PLoS ONE. 2008; 3(7):e2691-.
Fennelly KP, Martyny JW, Fulton KE, Orme IM, Cave DM, Heifets LB. Cough-generated aerosols of Mycobacterium tuberculosis: a new method to study infectiousness. Am J Respir Crit Care Med. 2004 Mar 1;169(5):604-9. doi: 10.1164/rccm.200308-1101OC. Epub 2003 Dec 4. PMID: 14656754
Fennelly KP, Jones-López EC, Ayakaka I, et al. Variability of Infectious Aerosols Produced during Coughing by Patients with Pulmonary Tuberculosis Am J Respir Crit Care Med. 2012; 186(5):450-457.
Fennelly KP. Particle sizes of infectious aerosols: implications for infection control The Lancet Respiratory Medicine. 2020; 8(9):914-924. [free full text]
Fiegel J, Clarke R, Edwards DA. Airborne infectious disease and the suppression of pulmonary bioaerosols. Drug Discov Today. 2006;11(1-2):51–57. doi:10.1016/S1359-6446(05)03687-1 PMID: 16478691
Gralton J, Tovey ER, McLaws M, Rawlinson WD. Respiratory virus RNA is detectable in airborne and droplet particles J. Med. Virol.. 2013; 85(12):2151-2159.
Knibbs LD, Johnson GR, Kidd TJ, et al. Viability of in cough aerosols generated by persons with cystic fibrosis. Thorax. 2014; 69(8):740-745.
Leung NHL, Chu DKW, Shiu EYC, et al. Respiratory virus shedding in exhaled breath and efficacy of face masks Nat Med. 2020; 26(5):676-680.
Lindsley WG, Blachere FM, Thewlis RE, et al. Measurements of Airborne Influenza Virus in Aerosol Particles from Human Coughs PLoS ONE. 2010; 5(11):e15100-.
Lindsley WG, Pearce TA, Hudnall JB, et al. Quantity and Size Distribution of Cough-Generated Aerosol Particles Produced by Influenza Patients During and After Illness Journal of Occupational and Environmental Hygiene. 2012; 9(7):443-449.
Loeb M, McGeer A, Henry B, et al. SARS among critical care nurses, Toronto. Emerg Infect Dis. 2004;10(2):251–255. doi:10.3201/eid1002.030838 PMID: 15030692
Milton DK, Fabian MP, Cowling BJ, Grantham ML, McDevitt JJ. Influenza Virus Aerosols in Human Exhaled Breath: Particle Size, Culturability, and Effect of Surgical Masks PLoS Pathog. 2013; 9(3):e1003205-.
Morawska L. Droplet fate in indoor environments, or can we prevent the spread of infection?. Indoor Air. 2006;16(5):335–347. doi:10.1111/j.1600-0668.2006.00432.x PMID: 16948710
Nicas M, Nazaroff WW, Hubbard A. Toward understanding the risk of secondary airborne infection: emission of respirable pathogens. J Occup Environ Hyg. 2005;2(3):143–154. doi:10.1080/15459620590918466 PMID: 15764538
Patterson B, Morrow C, Singh V, Moosa A, Gqada M, Woodward J, Mizrahi V, Bryden W, Call C, Patel S, Warner D, Wood R. Detection of Mycobacterium tuberculosis bacilli in bio-aerosols from untreated TB patients. Gates Open Res. 2018 Jun 8;1:11. doi: 10.12688/gatesopenres.12758.2. PMID: 29355225
Raboud J, Shigayeva A, McGeer A, et al. Risk factors for SARS transmission from patients requiring intubation: a multicentre investigation in Toronto, Canada. PLoS One. 2010;5(5):e10717. Published 2010 May 19. doi:10.1371/journal.pone.0010717 PMID: 20502660
Tellier R. Aerosol transmission of influenza A virus: a review of new studies. J R Soc Interface. 2009;6 Suppl 6(Suppl 6):S783–S790. doi:10.1098/rsif.2009.0302.focus PMID: 19773292
Theron G, Limberis J, Venter R, et al. Bacterial and host determinants of cough aerosol culture positivity in patients with drug-resistant versus drug-susceptible tuberculosis Nat Med. 2020; 26(9):1435-1443.
Wainwright CE, France MW, O’Rourke P, et al. Cough-generated aerosols of Pseudomonas aeruginosa and other Gram-negative bacteria from patients with cystic fibrosis Thorax. 2009; 64(11):926-931.
Wilson N, Corbett S, Tovey E. Airborne transmission of covid-19 BMJ. 2020;
Yan J, Grantham M, Pantelic J, et al. Infectious virus in exhaled breath of symptomatic seasonal influenza cases from a college community Proc Natl Acad Sci USA. 2018; 115(5):1081-1086.
Morgenstern, J. “Aerosol Generating Medical Procedure” is a faulty paradigm, First10EM, December 5, 2020. Available at:
4 thoughts on ““Aerosol Generating Medical Procedure” is a faulty paradigm”
What about “AGP” in the dental field?
I don’t think aerosol generating procedures are a thing in any field. I have not looked at the literature for dentistry specifically, but the same principles apply. The big difference is that in medicine we are generally talking about symptomatic patients (febrile, coughing, short of breath), which wouldn’t apply in dentistry. Patients without COVID symptoms (like a patient with dental pain), we would be seeing in a normal surgical mask.
The existence of presymptomatic spread makes this complicated for everyone. Ultimately, the best way to keep all of us safe is to get COVID under control in the community.
“Prolonged close contact” is a better paradigm, but I feel like the wording could be better. An ECG or a UCath takes maybe 10minutes. That’s about the time I would spend with a patient with mild fever and coryza. It’s not the same level of closeness though (although looking into people’s throats is just asking for them to cough on you). I feel as though there are 2 variables – time, and closeness – that we’re trying to put into 1 idea and we need to be wary of how it will be interpreted. I feel the paradigm should be – Are you touching the patient? If yes, wear an N95. Are you going to be there for more than 5 minutes? If yes, wear an N95. If no, then don’t. This will probably lead to a sharp rise in PPE utilisation, which we will have to balance.
For me personally, I will be putting on my KN95 and keeping it on for my 8 hour shift until I have access to a vaccine.