Assessment of N95 respirator decontamination and re-use for SARS-CoV-2

The unprecedented pandemic of SARS-CoV-2 has created worldwide shortages of personal protective equipment, in particular respiratory protection such as N95 respirators. SARS-CoV-2 transmission is frequently occurring in hospital settings, with numerous reported cases of nosocomial transmission highlighting the vulnerability of healthcare workers. In general, N95 respirators are designed for single use prior to disposal. Here, we have analyzed four readily available and often used decontamination methods: UV, 70% ethanol, 70C heat and vaporized hydrogen peroxide for inactivation of SARS-CoV-2 on N95 respirators. Equally important we assessed the function of the N95 respirators after multiple wear and decontamination sessions.

Dear editor,

10
The unprecedented pandemic of COVID-19 has created worldwide shortages of personal protective 11 equipment, in particular respiratory protection such as N95 respirators(1). SARS-CoV-2 transmission is 12 frequently occurring in hospital settings, with numerous reported cases of nosocomial transmission 13 highlighting the vulnerability of healthcare workers(2). The environmental stability of SARS-CoV-2 14 underscores the need for rapid and effective decontamination methods. In general, N95 respirators are 15 designed for single use prior to disposal. Extensive literature is available for decontamination procedures 16 for N95 respirators, using either bacterial spore inactivation tests, bacteria or respiratory viruses (e.g. 17 influenza A virus)(3-6). Effective inactivation methods for these pathogens and surrogates include UV, 18 ethylene oxide, vaporized hydrogen peroxide (VHP), gamma irradiation, ozone and dry heat(3-7). The 19 filtration efficiency and N95 respirator fit has typically been less well explored, but suggest that both 20 filtration efficiency and N95 respirator fit can be affected by the decontamination method used (7,8). For 21 a complete list of references see supplemental information.

23
Here, we analyzed four different decontamination methods -UV radiation (260 -285 nm), 70ºC dry heat, 24 70% ethanol and vaporized hydrogen peroxide (VHP) -for their ability to reduce contamination with 25 infectious SARS-CoV-2 and their effect on N95 respirator function. For each of the decontamination 26 methods, we compared the normal inactivation rate of SARS-CoV-2 on N95 filter fabric to that on 27 stainless steel, and we used quantitative fit testing to measure the filtration performance of the N95 28 respirators after each decontamination run and 2 hours of wear, for three consecutive decontamination 29 and wear sessions (see supplemental information). VHP and ethanol yielded extremely rapid inactivation 30 both on N95 and on stainless steel ( Figure 1A). UV inactivated SARS-CoV-2 rapidly from steel but more 31 slowly on N95 fabric, likely due its porous nature. Heat caused more rapid inactivation on N95 than on 32 steel; inactivation rates on N95 were comparable to UV. 33 34 for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 24, 2020. . https://doi.org/10.1101 Quantitative fit tests showed that the filtration performance of the N95 respirator was not markedly 35 reduced after a single decontamination for any of the four decontamination methods ( Figure 1B).

36
Subsequent rounds of decontamination caused sharp drops in filtration performance of the ethanol-treated 37 masks, and to a slightly lesser degree, the heat-treated masks. The VHP and UV treated masks retained 38 comparable filtration performance to the control group after two rounds of decontamination, and 39 maintained acceptable performance after three rounds.

41
Taken together, our findings show that VHP treatment exhibits the best combination of rapid inactivation 42 of SARS-CoV-2 and preservation of N95 respirator integrity, under the experimental conditions used here 43 ( Figure 1C). UV radiation kills the virus more slowly and preserves comparable respirator function. 70ºC 44 dry heat kills with similar speed to UV and is likely to maintain acceptable fit scores for two rounds of 45 decontamination. Ethanol decontamination is not recommended due to loss of N95 integrity, echoing 46 earlier findings 5 . 47 48 All treatments, particularly UV and dry heat, should be conducted for long enough to ensure that a 49 sufficient reduction in virus concentration has been achieved. The degree of required reduction will 50 depend upon the degree of initial virus contamination. Policymakers can use our estimated decay rates 51 together with estimates of real-world contamination to choose appropriate treatment durations (see 52 supplemental information).

54
Our results indicate that N95 respirators can be decontaminated and re-used in times of shortage for up to 55 three times for UV and HPV, and up to two times for dry heat. However, utmost care should be given to 56 ensure the proper functioning of the N95 respirator after each decontamination using readily available 57 qualitative fit testing tools and to ensure that treatments are carried out for sufficient time to achieve 58 desired risk-reduction. It will therefore be critical that FDA, CDC and OSHA guidelines with regards to 59 fit testing, seal check and respirator re-use are followed(9, 10). 60 for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 24, 2020. . https://doi.org/10.1101  This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 24, 2020. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 24, 2020.  This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 24, 2020. . https://doi.org/10. 1101  This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

143
The COVID-19 pandemic has highlighted the necessity for large-scale decontamination procedures 144 for personal protective equipment (PPE), in particular N95 respirator masks(1). SARS-CoV-2 has 145 frequently been detected on PPE of healthcare workers(11). The environmental stability of SARS-CoV-2 146 underscores the need for rapid and effective decontamination methods(12). Extensive literature is 147 available for decontamination procedures for N95 respirators, using either bacterial spore inactivation 148 tests, bacteria or respiratory viruses (e.g. influenza A virus)(3-6, 9, 13-15). Effective inactivation methods 149 for these pathogens and surrogates include UV, ethylene oxide, vaporized hydrogen peroxide (VHP), 150 gamma irradiation, ozone and dry heat (4,5,7,9,(14)(15)(16). The filtration efficiency and N95 respirator fit 151 has typically been less well explored, but suggest that both filtration efficiency and N95 respirator fit can 152 be affected by the decontamination method used (7,8). It will therefore be critical that FDA, CDC and 153 OSHA guidelines with regards to fit testing, seal check and respirator re-use are followed (9,(17)(18)(19)(20). This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

178
Plates were removed at 10, 30 and 60 minutes and 1 mL of cell culture medium added. The energy the 179 discs were exposed to at 10, 30 and 60 min is 0.33 J/cm 2 , 0.99 J/cm 2 , and 1.98 J/cm 2 respectively. While 180 the CDC has no specific recommendations on the minimum dose, they do report that a 1 J/cm2 dose can 181 reduce tested viable viral loads by 99. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 24, 2020. . https://doi.org/10. 1101 sprayed onto the discs from approximately 10 cm. After 10 minutes, 1 mL of cell culture medium was 187 added.

188
VHP. Plates with fabric and steel discs were placed into a Panasonic MCO-19AIC-PT (PHC Corp. of 189 North America Wood Dale, IL) incubator with VHP generation capabilities and exposed to hydrogen 190 peroxide (approximately 1000 ppm). The exposure to VHP was 7 minutes, after the inactivation of the 191 hydrogen peroxide, the plate was removed and 1 mL of cell culture medium was added.

192
Control. Plates with fabric and steel discs and steel plates were maintained at 21-23°C and 40% relative 193 humidity for up to four days. After the designated time-points, 1 mL of cell culture medium was added. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 24, 2020.

211
We use <Distribution Name>CDF(x | parameters) and <Distribution Name>CCDF to denote the 212 cumulative distribution function and complementary cumulative distribution functions of a probability 213 distribution, respectively. So for example NormalCDF(5 | 0, 1) is the value of the Normal(0, 1) 214 cumulative distribution function at 5.

215
We use logit(x) and invlogit(x) to denote the logit and inverse logit functions, respectively: This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 24, 2020. . https://doi.org/10.1101 where v is the log 10 virus titer in TCID 50 , where v is the log 10 virus titer in TCID 50 , and the well is infected 231 if at least one virion successfully infects a cell. The value of the mean derives from the fact that our units 232 are TCID 50 ; the probability of infection at v = 0, i.e. 1 TCID 50 , is equal to 1 -

233
Let Y ijkdl be a binary variable indicating whether the l th well of dilution factor d (expressed as log 10 234 dilution factor) of sample ijk was positive (so Y ijkdl = 1 if the well was positive and 0 otherwise), which 235 will occur as long as at least one virion successfully infects a cell.

236
It follows from (5) that the conditional probability of observing Y ijkdl = 1 given a true underlying titer 237 log 10 titer v ijk is given by: 241 is the expected concentration, measured in log 10 TCID 50 , in the dilute sample. This is simply the 242 probability that a Poisson random variable with mean (-ln(2) × 10 x ) is greater than 0. Similarly, the 243 conditional probability of observing Y ijkdl = 0 given a true underlying titer log 10 titer v ijk is given by: which is the probability that the Poisson random variable is 0.

246
This gives us our likelihood function, assuming independence of outcomes across wells.

247
Virus inactivation regression 248 for use under a CC0 license.
This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 24, 2020. . https://doi.org/10. 1101 The durations of detectability depend on the decontamination treatment but also initial inoculum and 249 sampling method, as expected. We therefore estimated the decay rates of viable virus titers using a 250 Bayesian regression analogous to that used in van Doremalen et al., 2020(12). This modeling approach 251 allowed us to account for differences in initial inoculum levels across replicates as well as other sources 252 of experimental noise. The model yields estimates of posterior distributions of viral decay rates and half-253 lives in the various experimental conditions -that is, estimates of the range of plausible values for these 254 parameters given our data, with an estimate of the overall uncertainty(24).

255
Our data consist of 10 experimental conditions: 2 materials (N95 masks and stainless steel) by 5 256 treatments (no treatment, ethanol, heat, UV and VHP). Each has three replicates, and multiple time-points 257 for each replicate. We analyze the two materials separately. For each, we denote by Y ijkdl the positive or 258 negative status (see above) for well l which has dilution d for the titer v ijk from experimental condition i 259 during replicate j at time-point k.

260
We model each replicate j for experimental condition i as starting with some true initial log 10 titer 261 v ij (0) = v ij0 . We assume that viruses in experimental condition i decay exponentially at a rate λ i over time t.

262
It follows that: We use the direct-from-well data likelihood function described above, except that now instead of 265 estimating titer distribution about a shared mean µ ij we estimate λ i under the assumptions that our 266 observed well data Y ijkdl reflect the titers v ij (t).

267
Regression prior distributions

268
We place a weakly informative Normal prior distribution on the initial log 10 titers v ij0 to rule out 269 implausibly large or small values (e.g. in this case undetectable log 10 titers or log 10 titers much higher than 270 the deposited concentration), while allowing the data to determine estimates within plausible ranges:

271
for use under a CC0 license.
This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

283
We had 6 replicate masks j for each of 5 treatments i (no decontamination, ethanol, heat, UV and 284 VHP). Each mask j was assessed for fit factor at 4 time-points k: before decontamination, and then after 1, 285 2, and 3 decontamination cycles. We label the control treatment i = 0. So we denote by F ijk the fit factor 286 for the j th mask from the i th treatment after k decontaminations (with k = 0 for the initial value).

287
We first converted fit factors F ijk to the equivalent observed filtration rate Y ijk by: 288 289 Observation model and likelihood function 290 for use under a CC0 license.
This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 24, 2020. . https://doi.org/10.1101 We modeled the censored observation process as follows. logit (Y ijk That is, we calculate the probability of observing a value between Y + ijk and Yijk , given our parameters.

304
for use under a CC0 license.
This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 24, 2020.  Poisson single-hit model is used, the best guess is that the true titer lies somewhere below 1 TCID 50 / 352 (volume of deposited sample). How far below is determined by the number of wells. For four wells, as 353 was standard in our experiments, the first quarter log 10 titer at which 0 wells is the most likely outcome is 354 10 −0.5 TCID 50 per volume of sample. This is also the imputed Speaman-Karber titer in that case. Since we 355 used samples of volume 0.1 mL, this corresponds to a value of 10 0.5 TCID 50 /mL. So although we do not 356 use the Spearman-Karber method here (since we infer mean titers directly from the well data) we use that 357 LOD value to plot samples for which no replicate had a positive well (since the posterior distribution in 358 that case covers a wide-range of sub-threshold values). This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 24, 2020. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 24, 2020. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 24, 2020. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.