Intended for healthcare professionals

Rapid response to:


Face masks for the public during the covid-19 crisis

BMJ 2020; 369 doi: (Published 09 April 2020) Cite this as: BMJ 2020;369:m1435

Read our latest coverage of the coronavirus pandemic

Rapid Response:

Re: Face masks for the public during the covid-19 crisis

Dear Editor

Discussion on whether face masks are effective generally fails to differentiate between who is being protected, in what physical manner and what the spatial relationship between host and recipient is. This can be assessed better by splitting up the transmission process into sections, as is done routinely in industrial gas-liquid separation technology.

At the start of the process, “ordinary” masks such as surgical or home-made will reduce outward velocities of ejected particles/droplets from a host, whether by cough, sneezing or ordinary out-breathing. The mask fabric will also coalesce small droplets into larger ones, shifting the droplet size distribution upwards, which in turn makes it more likely they will drop out of the air before reaching a recipient. If the host’s mask is moist or even wet, the velocity reduction and coalescence properties will be largely retained, which is apparently against the accepted wisdom on this. It is, however, true, is based on simple physics, and is an established principle used in industrial mesh pad mist eliminators. An N95 or similar mask will do the job somewhat better, but with insufficient supplies, simple masks can be quite effective for protecting others from the host.

The spatial relationship between host and recipient of the potential viral load is also vital. Simple droplet-in-gas calculations show that 2 m distancing is, for the smallest droplets, completely inadequate, even in still air. In a wind, small and even some large droplets will be blown considerable distances from the host. Whilst 2 m distance may be adequate to prevent a host from infecting a recipient if the host is downwind of the recipient, even 10 m distance may be insufficient if the host is upwind. Both parties are “safe” in a crosswind. If a host coughs or sneezes near an air conditioning system intake, the AC will distribute the small droplets around the building.

To protect the recipient from droplets finally reaching their own mask does require an N95 or similar spec mask (or full air breather/filter), in agreement with the established wisdom, as its primary function is to absorb or filter these out.

These factors must be more deeply considered. Existing studies on mask effectiveness that I am aware of rarely make these differentiations. Such qualitative factors can be modelled and quantified using CFD modelling, a technique ever more being used in medical research.

Competing interests: No competing interests

10 April 2020
Cliff Grover
Chartered Chemical Engineer (ret'd)
Specialist in gas/liquid separation processes.