Can we safely reuse N95 masks?

Can we reuse N95s

N95 masks are designed to be single use. In normal times, they shouldn’t be reworn. In normal times, they shouldn’t be decontaminated and reused. But these are not normal times. COVID-19 has resulted in drastic PPE shortages around the world, and has left many hospital systems considering options for reusing their normally disposable equipment. This is a controlled study looking at 4 different possible techniques for decontaminating and reusing N95 masks. 

The paper

Fischer RJ, Morris DH, van Doremalen N, et al. Assessment of N95 respirator decontamination and re-use for SARS-CoV-2. 2020. Prepublished online

The Methods

This is a laboratory based study, in which 4 different N95 decontamination methods were compared. The 4 methods were UV radiation, 70 degree Celsius heat, 70% ethanol, and vaporized hydrogen peroxide. (If you are actually using these methods, it is probably worth reading the paper to delve into the details.) They looked at each method’s ability to reduce contamination of SARS-CoV-2, and they also assessed the effect of each method on respirator function.

Decontamination was measured by performing titrated viral cultures at various time frames after each decontamination method and on a control.

Respirator function was determined by quantitative testing, using the “Portacount Respirator fit tester”, which measures aerosols both inside and outside the mask. They did dynamic testing, in that they had participants move and talk while wearing the mask. The results are reported as the “fit factor”, which is the ratio of particles outside the mask to those inside. (A fit factor of 100 is considered a requirement for the mask to pass this test.)

The Results

The only technique that completely eliminated the virus was the vaporized hydrogen peroxide, but they only performed cultures up to 100 minutes. Given enough time, UV light and heat both look like they might also have eliminated all the virus, but that wasn’t actually demonstrated, and longer times might have significant impacts on the integrity of the masks.  The ethanol technique is not time based, but rather based on the number of applications. 5 applications of 70% ethanol DOES NOT completely eliminate the virus, which the authors say is consistent with prior research. 

The median time until only 1 in 1 million viral particles remained alive was:

  • Control: 1550 minutes
  • Ethanol: 13 minutes
  • Heat: 93 minutes
  • UV: 125 minutes
  • Vaporized hydrogen peroxide: 16 minutes

The integrity of all masks decreased with each decontamination. Looking at just the point estimates, integrity was fine for all methods after the first decontamination, but fell below acceptable levels after the second decontamination with ethanol, and the third decontamination with UV light and heat. However, the 95% confidence intervals cross the accepted integrity target after 2 decontamination cycles for every method, so until we have better data, it appears that 1 decontamination cycle should be the max. (The average here is much less important than the worst numbers seen, as we can’t afford to have ineffective masks being reused.) 

My thoughts

This is a really valuable study, and reassures me that N95s can be decontaminated and reused, at least a limited number of times. Ideally though, we would want to see a lot more data before adopting such a high risk, novel intervention. 

In terms of decontamination, the authors seem really focused on the speed that the virus was killed. I am not sure that speed is the most important outcome. If 100% of the virus is killed, I don’t really care if it takes 10 minutes or 10 hours. Obviously, this matters in terms of the efficiency of the process, but by itself, speed should not be the deciding factor. Assuming a technique can be shown to be 100% effective, we should ignore the speed of the technique, and focus on the longevity and integrity of the masks.

It is hard to tell from their discussion section, but I think they only tested the decontamination on discs cut out of an N95 mask. (They did decontaminate full masks to test their integrity, but I believe the viral cultures were done on small pieces of N95 material rather than full masks.) This is an issue, as discs are likely easier to decontaminate. Masks have complex geometry and are likely to fold. Decontamination methods might not work as well in crevices as they do on flat pieces of fabric.

My big concern is the difference between a proof of concept study that sterilizes a few masks at a time and the industrial processes required to replicate this in bulk. Will the decontamination be less effective with thousands of masks in a machine? Will a large scale operation be more likely to damage masks? How are the masks being transported? When you sterilize 1 mask at a time, they are unlikely to be bent or folded. However, when applying this process to thousands of masks, bending and folding seems much more likely, which could damage the mask, change the fit, or simply block part of the mask from being fully decontaminated. Furthermore, they only tested the integrity of 1 mask (the 3M 9211+), and I could imagine different shaped masks being impacted differently by these techniques. 

We also need to be very careful about the extrapolation of these results. They demonstrated that the integrity of the masks might be maintained for 2 or 3 decontaminations, depending on the technique, but there is a clear degradation from all methods. Unfortunately, they didn’t do a fourth decontamination, but the slopes of the curves make it look like the masks would be below the accepted standard after a 4th decontamination in every technique. I am not sure how hospitals are going to track the number of times that masks are decontaminated, but based on this data, there should be a strict limit on that number.

More importantly, we need to interpret this data like we would a non-inferiority or equivalence trial. The point estimate is less important than the 95% confidence interval, because we can’t afford to provide healthcare workers with ineffective masks. The 95% confidence interval crosses the accepted safety standard after 2 decontaminations with every method, so until we have larger data sets, this study tells us that no matter what technique we use, masks should probably not be decontaminated more than once. 

Finally, the authors suggest that we should test each of these masks after decontamination before they are used clinically. I think the idea would be that you apply the re-processed mask, and then do one of those qualitative fit tests with the bitter tasting spray. As long as you don’t taste the spray, you know that the mask is safe to use clinically. That certainly makes sense to me, but is a big departure from standard practice, and might require some work to implement. 

Bottom line

Based on this laboratory study, vaporized hydrogen peroxide looks like the best option for decontaminating N95 masks while maintaining their integrity. However, this study does not support any more than 1 decontamination, because after that point there is a risk that the integrity of the mask will be substandard.

More COVID-19 posts:

COVID airway management: SAS consensus statement

COVID Resuscitation Principles

COVID Ethics: Should clinicians see patients without appropriate PPE?

Chloroquine for COVID: No good evidence yet

Surviving Sepsis Campaign COVID Guidelines

Antivirals for COVID19 – Not ready for clinical use

COVID Research Roundup

Aerosols, Droplets, and Airborne Spread: Everything you could possibly want to know

Aerosol generating procedures

THE CEBM also has a good post on mask use during COVID


Fischer RJ, Morris DH, van Doremalen N, et al. Assessment of N95 respirator decontamination and re-use for SARS-CoV-2. 2020. Prepublished online

Cite this article as:
Morgenstern, J. Can we safely reuse N95 masks?, First10EM, April 29, 2020. Available at:

Leave a Reply

12 thoughts on “Can we safely reuse N95 masks?”

%d bloggers like this: