Debunking the myth that N95s are super protective for 40 hours of wear (let's stop recommending it!)

Two studies came out this year looking at how extended wear time affects the filtration efficiency and fit of N95s.

Based on these studies, it is really unsafe to recommend that people wear the same respirator for 40 hours.

It seems like this recommendation comes from a combination of: the CDC authorizing healthcare workers to wear respirators for 5 shifts if (and only if) there was a PPE shortage, 3M statements and individual people performing experiments using fit tests and/or filtration efficiency measurements on their own respirators.

Onto the results from the studies:

• In both studies, respirators were retired if they failed fit tests, were heavily soiled or deformed

After ~40 hours of wear:

• On average, N95s worn for 40 hours do not reach N95 standards of filtration efficiency

• After 40 hours of wear, 40 % of N95s do not reach N95 standards of filtration efficiency

• After 40 hours of wear in the filtration efficiency study, I think they state that 100 % of the respirators had been retired but it’s not super clear

• After 32-48 hours of wear in the fit test study, 92.8 % of the total respirators had been retired

After 32 or 32-48 hours:

• On average, N95s worn for 32 hours do reach N95 standards of filtration efficiency

• After 32 hours, 34.5 % of N95s do not reach N95 standards

• After 32 hours, 93.2 % of respirators were retired in the filtration efficiency study

• After 32-48 hours, 92.8 % of the total respirators were retired in the fit test study

After 24 or 24-36 hours:

• On average, N95s worn for 24 hours do reach N95 standards of filtration efficiency

• After 24 hours, 28.8 % of N95s do not reach N95 standards

• After 24 hours, 85.2 % of respirators were retired in the filtration efficiency study

• After 24-36 hours, ~82 % of the total respirators were retired in the fit test study

After 16 or 16-24 hours:

• On average, N95s worn for 16 hours do reach N95 standards of filtration efficiency

• After 16 hours, 10.1 % of N95s do not reach N95 standards

• After 16 hours, 71.0 % of respirators were retired in the filtration efficiency study

• After 16-24 hours, ~65 % of the total respirators were retired in the fit test study

After 8 or 8-12 hours:

• On average, N95s worn for 8 hours do reach N95 standards of filtration efficiency

• After 8 hours, 1.8 % of N95s do not reach N95 standards

• After 8 hours, 46.6 % of respirators were retired in the filtration efficiency study

• After 8-12 hours, 38.7 % of the respirators were retired in the fit test study

What about if we view the data differently (by study, and so we can see how many respirators were retired):

Filtration efficiency study:

The intention of the filtration efficiency study was to look at the filtration efficiencies of N95s at multiple time points up to 40 hours of wear time. Unfortunately, many of the N95s had to be retired at the end of each shift due to failing fit tests, deformations and/or heavy soiling.

• 365 (# N95s worn for 8 hr)

• ↓ -170 (46.6 % of total N95s retired)

• 195 (# N95s worn for 16 hr)

• ↓ -89 (71.0 % of total N95s retired)

• 106 (# N95s worn for 24 hr)

• ↓ -52 (85.2 % of total N95s retired)

• 54 (# N95s worn for 32 hr)

• ↓ -29 (93.2 % of total N95s retired)

• 25 (# N95s worn for 40 hr)

Fit test study:

• 803 (# N95s worn for 8-12 hr)

• ↓ -316 (39.4 %* of total N95s retired)

• 487 (# N95s worn for 16-24 hr)

• ↓ -223 (67.1 %* of total N95s retired)

• 264 (# N95s worn for 24-36 hr)

• ↓ -133 (83.7 %* of total N95s retired)

• 131 (# N95s worn for 32-48 hr)

• ↓ -65 (91.8 %* of total N95s retired)

• 66 (# N95s worn for 40-60 hr)

*These percentages are slightly different than the ones I reported previously. Unlike in their analysis, this is the raw data with the two rounds of testing combined and without statistical processing.

A look at the average filtration efficiencies (FE) of N95s after different wear times:

• 99.5 % (average FE after 0 hr)

• 99.2 % (average FE after 8 hr)

• 97.8 % (average FE after 16 hr)

• 96.0 % (average FE after 24 hr)

• 95.2 % (average FE after 32 hr)

• 94.2 % (average FE after 40 hr)

Sample calculations using some filtration efficiencies from the study:

The average FE was 94.2 % after 40 hours of wear.

And for 40 % of these N95s, the FE was <95 %.

By looking at a graph from the study, one of these N95s had a FE ~69 % and one had one of ~82 %.

Let’s get into what 99.5 %, 94.2 %, 82 % and 69 % filtration efficiencies could mean!

There was a modelling study where the authors estimated that under certain conditions, it would take 30 seconds to inhale an infectious dose of SARS-CoV-2 (the virus that causes COVID-19) from aerosols <12 um.

Let’s assume that:

• 20 % of the aerosols in sizes <12 um are about 300 nm* (this is a very rough estimate from looking at graphs from multiple studies, example here, and I’m including a range of ~100-500 nm because when N95s drop in FE they seem to drop in FE in at least this range)

• And that 25 % of particles around that size are deposited in the body (specifically lungs) when breathed in (consistent with the many studies on the subject, example here)

• *Particles ~300 nm make it through N95s the best out of all particle sizes

With those assumptions in mind and using an example from a modelling study where they estimated that the time to inhale an infectious dose of SARS-CoV-2 unmasked was 30 seconds (0.5 minutes) for aerosols <12 um, the time I estimate it would take to deposit an infectious dose in the lungs while wearing an N95 would be:

• 33.3 hours (2000 minutes) with a constant FE of 99.5 %* and no leaks

• 2.9 hours (172 minutes) with a constant FE of 94.2 % and no leaks

• 0.9 hours (56 minutes) with a constant FE of 82 % and no leaks

• 0.5 hours (32 minutes) with a constant FE of 69 % and no leaks

• 11.1 hours (667 minutes)# with a constant FE of 99.5 %* and a 1 % inward leak

• 2.5 hours (147 minutes)# with a constant FE of 94.2 % and a 1 % inward leak

• 0.9 hours (53 minutes)# with a constant FE of 82 % and a 1 % inward leak

• 0.5 hours (31 minutes)# with a constant FE of 69 % and a 1 % inward leak

• 1.6 hours (95 minutes)# with a constant FE of 99.5 %* and a 10 % inward leak

• 1.1 hours (63 minutes)# with a constant FE of 94.2 % and a 10 % inward leak

• 0.6 hours (36 minutes)# with a constant FE of 82 % and a 10 % inward leak

• 0.4 hours (24 minutes)# with a constant FE of 69 % and a 10 % inward leak

*N95s don’t have constant FEs. for example, the evidence suggests that it would be impossible for an N95’s FE to remain at 99.5 % for 33.3 hours

#This is based only on small aerosols ~100-500 nm and thus this is an underestimation, because way more particle sizes get through in the case of a leak

To summarize this with simpler numbers, it would take one tenth (or 10 %, or 10 times less) of an amount of time to inhale an infectious dose of SARS-CoV-2 in an N95 with a FE of 95 % compared to one with an FE of 99.5 % (assuming no leaks).

Another important thing to note:

As mentioned previously, if aerosols of these sizes are deposited in the body, it is in the lungs. This means that nasal sprays, nasal rinses and mouthwashes cannot help. In general these products have overstated effectiveness, major issues with the associated clinical trials and no robust evidence suggesting that they help prevent COVID-19 (see my posts on how there’s no convincing evidence that nasal sprays prevent, and prevent nor treat, COVID-19).

With time, more and more N95s were deemed unsuitable for further wear in both of these studies. Most of this was due to fit test failure, which means the N95s had leaks.

My interpretation of these studies is that the fit test failures are more of a concern than the drop in FE. Though the drops in FE are enough of a concern on their own to discourage extended use of a respirator if possible, the FE is essentially meaningless if you have a leak and are therefore breathing in fully unfiltered air.

More N95 FE info and context about the filtration efficiency study:

The N95 FE standard is greater than or equal to 95 % for certain particle sizes (~185-300 nm).

However, most new N95s have a FE of ~99.5 % for those particle sizes.

In the filtration efficiency study, the drop in FE to an average of 94.2 % after 40 hours of wear time means that the average respirator did not meet N95 standards anymore. In other words, the respirators were no longer N95s.

Some considerations:

• The healthcare workers in these studies donned and doffed the N95s a median of 4 times during each shift

• It is unknown how much the participants checked the fit and adjusted their respirators for a good fit

• Respirators were reworn the very next day if they weren’t retired, for a total of 5 consecutive days in some cases

• Those of us still masking may do things differently

• In the fit test study, they state that their fit testing method might be overly sensitive, where an N95 is deemed to have failed a fit test when it shouldn't have. However, after reviewing the study they cite about that, I disagree. Basically, in the study they cite, the threshold for considering a fit test passed was too low.

My takeaways:

• Based on these studies, I would not suggest rewearing the same respirator for anywhere near 40 hours of wear time

• If you do wear respirators for over 8 hours of wear time, it is super important to check the seal and fit and get the respirator to fit as well as possible

• If you do wear respirators for over 8 hours of wear time, it would probably be safer to wear newer masks in higher risk environments and masks with more wear time in lower risk environments (both for fit and filtration efficiency reasons)

• Before we worry about drops in filtration efficiency, we should worry about leaks in the mask or fit test failure

• If we do manage to get a good seal and avoid leaks, a filtration efficiency of 94.2 % (or generally a drop in filtration efficiency over time) is concerning

• If you need respirators, check out covidactionmap.org, maskbloc.org and if you’re in so-called canada, go to donatemask.ca