key: cord-0941651-u9ckyx3l
authors: Côrtes, Marina Farrel; Espinoza, Evelyn Patricia Sanchez; Noguera, Saidy Liceth Vásconez; Silva, Aline Alves; de Medeiros, Marion Elke Sielfeld Araya; Boas, Lucy Santos Villas; Ferreira, Noely Evangelista; Tozetto-Mendoza, Tania Regina; Morais, Fernando Gonçalves; de Queiroz, Rayana Santiago; de Proenca, Adriana Coracini Tonacio; Guimaraes, Thais; Guedes, Ana Rubia; Letaif, Leila Suemi Harima; Montal, Amanda Cardoso; Mendes-Correa, Maria Cassia; John, Vanderley M.; Levin, Anna S.; Costa, Silvia Figueiredo
title: Decontamination and reuse of surgical masks and respirators during COVID-19 pandemic
date: 2020-12-24
journal: Int J Infect Dis
DOI: 10.1016/j.ijid.2020.12.056
sha: 7cf040b00793f442c31e4c904faafef766c0215c
doc_id: 941651
cord_uid: u9ckyx3l

OBJECTIVES: Coronavirus disease 2019 pandemic increased demand on personal protective equipment(PPE) causing shortages. We evaluated surgical masks and respirators reuse by analyzing their performance and safety before and after decontamination methods: oven, thermal drying, autoclave, and hydrogen peroxide plasma vapor. METHODS: 45 surgical masks and 69 respirators were decontaminated. We evaluated the visual integrity, air permeability, burst resistance, pressure differential, and particulate filtration efficiency of new and decontaminated masks. We also analyzed 14 used respirators, after work shifts, before and after decontamination by RT-PCR and viral culturing. Finally, reprocessed masks were evaluated by users as to functionality and comfort. RESULTS: The most simple and useful method was dry heat in the oven(75 °C for 45 minutes). Physical and filtration assays indicated that all methods were safe after 1 cycle. Oven maintained the characteristics of surgical masks and respirators for at least five reprocessing cycles. We detected viral RNA by RT-PCR in two of 14 used masks. Four masks submitted to viral culture were PCR-negative and culture-negative. Reprocessed respirators used in work shifts were positively evaluated by users even after three decontamination cycles. CONCLUSION: Dry heat is a safe decontamination method of surgical masks and respirators for at least five cycles and is feasible in the hospital.

Coronavirus disease 2019 is growing pandemic affecting all the world starting in December 2019. SARS-CoV2 has already infected 66,729,375 people worldwide (in December 08, 2020(World Health Organization, 2020a)). The main modes of its transmission and spread are human-to-human contact through droplets (expelled during sneezing, coughing, or speaking) and close contact. However, possible J o u r n a l P r e -p r o o f airborne transmission is a concern. The World Health Organization (WHO) recommends the use of medical masks for regular care of COVID-19 patients and respirators such as N95, FFP2 and FFP3 when aerosolgenerating procedures are performed (World Health Organization, 2020b) . Healthcare workers (HCW) are on the forefront of this pandemic, thus are the most exposed population. A Chinese study evaluated 138 hospitalized patients with COVID-19 of which 29% were HCW . Therefore, the demand on personal protective equipment (PPE) such as surgical masks and respirators has substantially increased.

This high demand leads to a rapid shortage of both PPE and raw materials for its manufacture (Chaib, 2020) .

Most of these PPEs are certified and recommended by manufacturers to be used only once. However, due to the SARS-CoV2 pandemic, it is necessary to find alternative measures. Health care workers are sometimes obliged to extend the useful life for days or weeks, without demonstrated effective and suitable methods of decontamination. Thus, it is important to evaluate the effectiveness of reprocessing masks, which is the best method, and the maximum number of decontamination cycles , (Mackenzie, 2020) . This study aims to evaluate the performance of surgical masks and respirators after using different decontamination methods that can be done in a hospital central sterile service department.

In total, forty-five new surgical masks (Descarpack, São Paulo, Brazil) and sixty-nine new respirators (KSN, São Paulo, Brazil) were evaluated ( Figure 1 ). Each mask was used for 4 hours, to evaluate the effect of use, 20 minutes, to evaluate several decontamination cycles, or during at least 12 hours shifts and submitted to decontamination by one of four methods: dry heat in the oven (Fanem 502, 75ºC for 45 min -230 L capacity); thermal drying machine (Getinge, drying cycle: 84ºC for 40 minutes -300 L capacity ); autoclaving (134ºC during 4 minutes for decontamination and 10 min for drying; 24min to get to 134ºC -960 L capacity Getinge HS 6620 -Sweden), and hydrogen peroxide plasma vapor (H2O2; standard cycle for 47 minutes -temperature below 55°C -double tray, 100 cm deep, Sterrad 100 NX, USA). A maximum of 30 masks/respirators were tested by cycle in each equipment. After each cycle, the masks were inspected visually and submitted to another cycle (maximum of 10 cycles). Additionally, five different brands of respirators: Deltaplus PFF2 and PFF3, Maskface, Tayco and Proteplus (with a headquarters in São Paulo, Brazil) were also analyzed after three cycles of decontamination in the oven. It is important to note that decontamination methods were performed in the central sterile service department of Hospital das Clinicas, Sao Paulo, SP -Brazil. Hospital das Clinicas is a tertiary hospital with 2,400 beds and the reference hospital for COVID-19.

Besides visual inspection (cleanliness, nose clip and elastic functionality) the surgical masks and respirators were evaluated for permeability to air, burst resistance, breathability (pressure differential), particulate filtration efficiency and DNA retention capacity. Additionally, a pilot study was done to evaluate the workers' responses (n=33) considering sealing and breathing performance after working regular shifts using reprocessed respirators.

Furthermore, 14 used respirators from two hospitals were submitted to SARS-CoV2 RNA detection and four of them were submitted to viral culturing. These 14 respirators were collected at different times of the first month of the pandemic, the first four were collected at the beginning of the pandemic and the five last were collected in the last week of the first month. This is important since various adaptations happened during the first month of the pandemic, for example, during the first week face shields had not yet been distributed in the hospital and the PPE training of new HCWs was still in progress. By the end of the first month all frontline HCWs were trained and had access to complete PPEs, including face shields.

The physical integrity and particulate filtration of the masks were evaluated for (i) air permeability (l.m -2 .s -1 ), following ISO 9237:1995 -Textile -Determination of the permeability of fabrics to air. The air permeability test was performed at five different points in each mask unit, for each decontamination condition. The measurement area was 5 cm² and the applied pressure drop of 100 Pa; (ii) burst resistance (bar) following ISO 13938.1: 1999 -Textiles -Bursting properties of fabrics. Measurements were made at five different points in a mask unit, for each decontamination condition. The measurement area was 5 cm²;

and (iii) The pressure differential ∆P (mmH2O/cm 2 ) following Annex C of EN 14683:2019 -Medical face masks -Requirements and test methods.

Particulate filtration efficiency. The particle filtration efficiency was evaluated by measuring the particle size distribution that passed through the mask as a function of the total amount of NaCl particles generated through an ATM226 aerosol generator (TOPAS, Saxe, Germany) through an electronic particle detection system (Scanning Mobility Particle Sizer) 3080 from TSI, Minnesota, USA), coupled to a condensation particle counter (CPC nanoparticle counter 3771 TSI, Minnesota, USA). The number of particles with and without the masks was measured from 20 to 800nm at a distance of 15 cm from the aerosol source. The J o u r n a l P r e -p r o o f ratio between number of aerosol particles reaching the detector after the mask and the total number of particles suggests the efficiency of the mask. This experiment was carried out with five new surgical masks, repeating the procedure up to four times. In addition, surgical masks were decontaminated four times in an oven and respirators twice in an oven and twice in H2O2.

DNA retention capacity of masks after dry heat decontamination. In order to evaluate the masks ability to retain small molecules, DNA of Staphylococcus aureus was extracted using QiAamp DNA mini kit (Qiagen, Hilden, Germany) according to manufactures recommendations. A 47 mm/0.45 µm cellulose membrane (Merck Millipore, Burlington, Massachusetts, EUA) was placed in a petri dish behind the mask at a distance of 1cm. DNA was dosed in NanoDrop 2000c (Thermo Fisher, Waltham, Massachusetts, EUA) and diluted in water. Water (negative control) or 50 ng DNA was sprayed 4 times on the mask's surface from 5 cm. DNA or water sprayed on the membranes without the mask protection were also used as positive and negative controls, respectively. The membranes were transferred to a 1.5ml Eppendorf and 50µl of water was added and incubated for 10min at room temperature. Then, 5 µl was used as template for a PCR reaction spa (Staphylococcus protein A) gene using the following primers (Forward-TAAAGACGATCCTTCGGTGAGC and Reverse-CAGCAGTAGTGCCGTTTGCTT). PCR products were applied into agarose (1.2%) gel and a visible band in the gel were considered positive. This test was performed with seven surgical masks (one knew, three five times decontaminated and three ten times decontaminated) as well as three respirators (one new and two seven times decontaminated).

User evaluation: Evaluation was done in an intensive care unit with 22 beds used for suspected and confirmed COVID-19 cases. The respirators used during a work shift (12 hours) were placed in paper envelopes, identified with the HCW names, and removed daily for decontamination by oven. Each mask was reused by the same HCW who evaluated the mask before each use, by answering a questionnaire based on Occupational Safety and Health Administration (OSHA) attached to the envelope. The masks were decontaminated three times but if a mask was rejected by the user, he or she was instructed to discard the mask.

Fourteen respirators (Deltaplus PFF2) used during 3-28 days shifts were collected in two different hospitals, seven from Hospital das Clinicas (HC 1, HC 2, HC 5, HC 6, HC 7, HC 8, HC 9, HC10, HC 11, HC 12, HC 13 and HC 14) public and reference for COVID-19 and two from the private Hospital São Camilo (SC 3 and SC 4). The masks were cut, in the region close to J o u r n a l P r e -p r o o f the nose and mouth (highest exposure area), generating 16 punches of 1 mm of diameter collected from each facial mask using micro-punch tool (Harris Uni-Core™), which consists of a razor-sharp stainlesssteel cutting tip (Appendix Figure 1 ). It should be noted that we used one tool per mask. The procedures were done in Biosafety level 3 area.

RT-PCR and viral culture were performed to identify the presence of SARS-CoV2. From eight mask fragments in lysis buffer, the RNA extraction was performed using QIAamp viral RNA kit according to the manufacturer instructions. RT-PCR was assessed using the commercial RealStar® SARS-CoV2 RT-PCR Kit 1.0, develop by Altona Diagnostics and the amplification was done using the Roche LightCycler® 96

System -USA. Real-time PCR data was expressed by the Ct (Cycle threshold) value, corresponding to the initial amplification cycle. The results were reported as detectable or undetectable.

Viral culture was performed for four respirators (HC 11, HC 12, HC 13 and HC 14) using Vero cells (ATCC® CCL-81™) as described previously (Lennette, E. H., Schmidt, N. J., 1979) ; (Ammerman et al., 2009 ); (Harcourt et al., 2020) . Cells were cultured in Dulbecco minimal essential medium (DMEM) supplemented with heat-inactivated fetal bovine serum (10%) and antibiotics/antimycotics (Cultilab, Campinas, São Paulo). For virus isolation, mask fragments were inoculated in Vero Cells culture in plastic bottles (Jet biofilm, 12.5 cm 2 area, 25 mL capacity) immediately after processing. We then grew the inoculated cultures at 37°C incubator in an atmosphere of 5% CO2. Cell cultures were maintained for at least 2 weeks and were observed daily for evidence of cytopathic effect (CPEs). At least two subcultures were performed weekly. Presumptive detection of virus in supernatant showing CPEs was investigated using the Inverted microscope (Nikkon, Japan) and then confirmed by specific RT-PCR targeting E gene.

Statistical analysis was performed using GraphPad. t test was used to evaluate new and used masks. Oneway ANOVA for each decontamination method analysis and a correlation analysis using R square value for mask performance and filtration efficiency over decontamination cycle.

Visual integrity of surgical masks or respirators was analyzed after each decontamination using oven, autoclave, thermal drying, and hydrogen peroxide. Appendix Figure 2 shows the visual integrity of surgical J o u r n a l P r e -p r o o f masks after up to ten decontaminations and Appendix Figure 3 shows respirators after seven dry heat in the oven decontaminations. The steel clip of all respirators came off after the third decontamination. The strap from all the surgical masks broke after the sixth; and after the 10 th decontamination, the masks were visibly altered.

The steel clip of all the respirators came off after the second autoclave decontamination treatment; the masks became easy to break by simple handling after the third treatment. A single complete thermal drying cycle (including washing cycle with detergent and water) destroyed the masks, so we decided to submit masks to drying cycle only. No visual differences were found after up to two cycles of hydrogen peroxide decontamination.

DNA retention capacity of surgical masks after dry heat decontamination. The DNA filtering capacity of masks was analyzed for intact masks and masks reused six and ten times, for surgical masks and seven times for respirators. Surgical masks were able to filtrate the sprayed DNA-dope aerosol even after 6 dry heat decontamination cycles. Intact and 6 times reused surgical masks were able to retain the sprayed DNA.

However, after the tenth cycle, DNA was identified in the membrane on inside of the mask (Appendix Figure 4 ).

First, we evaluated the wearing effect (donning and doffing, humidity from breathing, etc.) of the simple mask use for 4h on its performance. Figure 2 shows that simple use wear has a measurable effect on the mask performance both on surgical masks ( Figure 2BDF ) and respirators ( Figure 2ACE ), but the magnitude is very low, much smaller than the measured differences between masks brands and, therefore without practical significance. Then we evaluated the screening of one cycle of 1 use for 20min followed by decontamination with four different technologies. Figure 3 shows that treatments had no practical differences on air permeability ( Figure 3AB ), burst resistance ( Figure 3CD ), and pressure differential ( Figure 3EF ) for both surgical ( Figure 3ACE ) and respirators ( Figure 3BDF ). Thereby, once oven and autoclave are readily available in almost all hospitals, we decided to evaluate the effect of up to five cycles with these two methods.

The third cycle of autoclave decontamination did affect the mechanical resistance (burst) of surgical masks, rendering impossible to assembling them to the testing machine without damages. Air permeability of surgical masks was not affected by decontamination cycles in both oven and autoclave ( Figure 4AD) J o u r n a l P r e -p r o o f treatments. Burst resistance, by other side decreased linearly with the number of decontamination cycles, being the effect of oven treatment much lower than the autoclaved treatment ( Figure 4BE ). Pressure differential reduced slightly due to oven ( Figure 4C ) treatment, which is positive, but the autoclave results presented higher dispersion and no clear tendency ( Figure 4F ). Except for a slight Pressure differential improvement due to autoclave treatment, (Figure 4L ), no significant effect was observed in the, more robust, respirator ( Figure 4GHIJKL ).

To test if the effects were reproducible in products from other brands, five different respirators brands have been submitted to three cycles of decontamination in the oven ( Figure 5) . Even if the effect of the treatment were statistically significant, it would be but much smaller than the differences between the new masks from the various brands.

Our data showed that the number of cycles of decontamination did not affect the particulate filtration efficiency of masks. Differences observed for surgical masks and respirators are most probably due to variability between masks and experimental errors. The particulate filtration efficiency indicates that surgical masks remained with filtration capacity above 92% after decontamination and respirators above 96% ( Figure 6 ).

Among the fourteen masks (HC 1, HC 2, SC 3, SC 4, HC 5-14) used by HCW during work shifts, RNA was detected in two before decontamination, in particular the ones collected in the first week of pandemic: HC1 and SC3. One of them (SC 3, 7%) remained positive after dry heat decontamination. HC1 had a CT value of 36 before decontamination and undetectable after. SC3 had a CT value of 33 before decontamination and 32 after decontamination in the oven. HCWs were notified about the positive results and did not reported any symptoms. Four masks (HC 11, HC 12, HC 13 and HC 14) were subjected to viral culture and none of them showed cytopathic effect and RT-PCRs were negatives during the culture.

User evaluation of masks reuse. The results of the visual and functional evaluation of the respirators decontaminated in the oven, are shown in Table 1 . It is important to note that the masks were not handled by the persons involved in the process of decontamination. The masks were decontaminated without removing them from the paper envelopes in which they were placed by the HCWs. The dirt on the masks was mostly due to the user's own makeup. Clip adhesion and elastic could be the problems limiting the reuse.

After evaluating four different methods of reprocessing masks, we believe that the most simple and useful method is dry heat in the oven (75 o C for 45 minutes). Oven maintained the physical characteristics and filtering capacity of surgical masks and respirators for at least five reprocessing cycles. Reprocessed respirators used in real 12-hour work shifts in a COVID-19 dedicated intensive care unit were positively evaluated by users even after three cycles.

In the context of pandemics and global PPE shortage such as the COVID-19 pandemic, alternatives such as reuse and adapting existing technologies are necessary. Due to the panorama of PPE shortage in most hospitals, we evaluated the possibility of reusing surgical masks and respirators. We decided to investigate dry heat in the oven and in the thermal drying machine, autoclaving, and hydrogen peroxide, which were available at the hospital. Once all methods showed acceptable performances after one cycle of decontamination, we decided to better evaluate oven and autoclave which are technologies most found in hospitals. Oven was the method that performed best for surgical masks because autoclave drastically reduced its burst resistance. The simple use of surgical masks augmented both air permeability and burst resistance. Interestingly, this effect was basically restored after heat treatment making the masks better after decontamination than without it. This phenomenon can probably be explained due to thermal shrinkage of the material. No relevant changes were found regarding respirators, since the presented alterations after decontamination cycles were smaller than the differences found between distinct commercial brands. Air permeability and burst resistance are tests for general textiles that were also included here to expand the analysis. According to the Brazilian standard for non-woven articles for medical and hospital use ABNT NBR 15052, pressure differential ΔP must be equal or less than 4mmH2O which means that all decontamination methods passed the test. The factor that prevents the reuse of respirator more than two times is detachment of the nasal clip, due to glue failure. This suggests innovation in this manufacture process to allow greater number of reuses.

We also evaluated respirators that had been used by HCWs during hospital work shifts. The viral cultures of respirators were negative after oven decontamination suggesting that it is safe. Finally, HCWs were asked to evaluate their decontaminated respirators, used during 12-hour work shifts, and considered them adequate.

The idea of recycling PPEs has already been considered in previous epidemics (Bailar JC, Brosseau LM, 2006) , (Pillai S.K., Beekmann S.E., Babcock H.M., Pavia A.T, Koonin LM, 2016) , (Lin et al., 2017) , (Mills et al., 2018) . Most of the studies regarding PPE reuse are performed with respirators, data are especially scarce for surgical masks. Among the processes, some tests on respirators under dry heat proved to be safe, with respect to integrity, decontamination, and filtration (Viscusi et al., 2009) . Previous studies have shown different methods for respirators decontamination such as hydrogen peroxide vapor (Schwartz et al., 2020) , and ultraviolet irradiation (Mills et al., 2018) . In vitro, Darnell et al.(Darnell et al., 2004) showed that SARS-Cov is inactivated by heat decontamination for 45 min at 75ºC. Because of this, we chose this method and tested other alternatives that were available in our central sterile service department. Heat treatment is a simple and accessible method to decontaminate respirators, since the virus is not stable to heat (Chin et al., 2020) , (Loh et al., 2020) , (Abraham et al., 2020) .

The particulate filtration efficiency was measured using particles ranging from 0.08 to 0.14 μm and both surgical and respirators maintained high efficiency (above 92% and 96%, respectively) after decontamination. The size of the SARS-CoV2 particle was estimated to be 0.06 to 0.14 μm which indicates that decontamination methods may be considered appropriate. According to the Brazilian ABNT NBR 15052, the particle filtration efficiency (PFE) must remain 98% or more. However, the size of the particle in the standard is 0.105 μm and we also included smaller particles in the analysis to get closer to the viral particle's dimensions. The WHO PPE specifications indicate a PFE ≥ 95% as level 1 and ≥ 98% as levels 2 and 3. It is important to note that the PFE increased (slightly for surgical masks and considerably for respirators) with the decontamination cycles.

Real Time PCR showed that SARS-CoV2 RNA could be detected only in two (14%) masks used by HCW after a work shift. Although one mask remained positive after decontamination, it probably means that RNA was present in the mask, but the virus was inactivated. Moreover, it was not possible to perform the virus culture of this mask. It is important to note that from the 14 HCWs, only the first 4 (including the 2 with positive RT-PCR) did not had face shields available because it was the first week of the pandemic and all safety measurements were not yet implemented. Also other factors may influence the contamination of masks as for example incorrect doffing. On the other hand, masks submitted to viral culture were negative suggesting that no viable virus was present after heat decontamination.

Due to the extreme urgency that this crisis demanded and due to the shortage of equipment, this study used a small sample of masks. However, the study was driven by the urgent need to find alternatives so that HCW could work safely during the pandemic. These results may be applicable to other emergencies in the future.

Our results show that at least 5 cycles of reuse after decontamination by dry heat in the oven at 75ºC for 45 min of surgical masks and respirators may be a practical solution to protect HCW during periods of shortage of PPE such as the SARS-Cov2 pandemic. This method can be easily done in the hospital's central sterile services department. Finally, this study finds a solution to the problem of decontaminating surgical masks. : Correlation between the Percent of particulate filtration efficiency (Pfe) and decontamination cycles of surgical masks (A) and respirators (BC) after 1cycle of decontamination in the oven (AB; blue) and hydrogen peroxide plasma vapor (H2O2; C) (red). * p < 0.05; ns= not significant; R 2 = R square.

Using heat to kill SARS-CoV-2

Growth and Maintenance of Vero Cell Lines

Reusability of facemasks during an influenza pandemic: Facing the flu

Shortage of personal protective equipment endangering health workers worldwide

Disinfection of N95 respirators by ionized hydrogen peroxide during pandemic coronavirus disease 2019 (COVID-19) due to SARS-CoV-2

Stability of SARS-CoV-2 in different environmental conditions

Inactivation of the coronavirus that induces severe acute respiratory syndrome, SARS-CoV

Isolation and characterization of SARS-CoV-2 from the first US COVID-19 patient

Diagnostic Procedures for Viral

Filter quality of electret masks in filtering 14.6-594 nm aerosol particles: Effects of five decontamination methods

Heat treatment for reuse of disposable respirators during Covid-19 pandemic: Is filtration and fit adversely affected?

Reuse of N95 Masks

Ultraviolet germicidal irradiation of influenza-contaminated N95 filtering facepiece respirators

Clinician Beliefs and Attitudes Regarding Use of Respiratory Protective Devices and Surgical Masks for Influenza

Decontamination and Reuse of N95 Respirators with Hydrogen Peroxide Vapor to Address Worldwide Personal Protective Equipment Shortages During the SARS-CoV-2 (COVID-19) Pandemic

Evaluation of five decontamination methods for filtering facepiece respirators

Clinical Characteristics of 138 Hospitalized Patients with 2019 Novel Coronavirus-Infected Pneumonia in Wuhan, China

A novel coronavirus from patients with pneumonia in China

Marina Farrel Côrtes, conception of the work, analysis, interpretation of data, writing and revising

Evelyn Patricia Sanchez Espinoza, conception of the work, analysis, writing and revising

Saidy Liceth Vásconez Noguera, conception of the work

Aline Alves Silva, conception of the work

Marion Elke Sielfeld Araya de Medeiros, conception of the work

Lucy Santos Villas Boas, conception of the work

Fernando Gonçalves Morais, conception of the work

Rayana Santiago de Queiroz, conception of the work

Adriana Coracini Tonacio de Proenca, Acquisition of data

Thais Guimaraes, study design, Acquisition of data

Ana Rubia Guedes, conception of the work and writing

Acquisition of data

Acquisition of data

Maria Cassia Correa Mendes, study design, Acquisition of data and revising

design of the work, acquisition, analysis, interpretation of data for the work and revising

study design, acquisition of data, writing, and revising

design of the work, acquisition, analysis, interpretation of data for the work and revising

All Authors agree to be accountable for all aspects of the work, approved and revised this final version

Ethical Approval: Not applicable.

All authors read and agreed with this manuscript and that this manuscript will not be submitted elsewhere while under revision in this journal. All authors stated that there is no conflict of interest to declare.J o u r n a l P r e -p r o o f

Membrane without mask sprayed with DNA; 3 New SM sprayed with H2O; 4 Membrane without mask sprayed with H2O; 5 5 times reused mask sprayed with DNA; 6 5 times reused mask sprayed with H2O; 7 10 times reused mask sprayed with 50ng/µl DNA; 8 10 times reused mask sprayed with H2O; 9 100bp DNA ladder; 10 Respirator used 7 times sprayed with H2O; 11 Respirator used 7 times sprayed with DNA; 12 negative control of the PCR reaction. This experiment was repeated 3 times.