The need for systematic quality controls in implementing N95 reprocessing and sterilization

Abstract

Background

Due to increased requirement for personal protective equipment during the coronavirus disease 2019 pandemic, many medical centres utilized sterilization systems approved under Food and Drug Administration Emergency Use Authorization for single-use N95 mask re-use. However, few studies have examined the real-world clinical challenges and the role of ongoing quality control measures in successful implementation.

Aims

To demonstrate successful implementation of quality control measures in mask reprocessing, and the importance of continued quality assurance.

Methods

A prospective quality improvement study was conducted at a tertiary care medical centre. In total, 982 3M 1860 masks and Kimberly-Clark Tecnol PFR95 masks worn by healthcare workers underwent sterilization using a vaporized hydrogen peroxide gas plasma-based reprocessing system. Post-processing qualitative fit testing (QFT) was performed on 265 masks. Mannequin testing at the National Institute for Occupational Safety and Health (NIOSH) laboratory was used to evaluate the impact of repeated sterilization on mask filtration efficacy and fit. A locally designed platform evaluated the filtration efficiency of clinically used and reprocessed masks.

Findings

In total, 255 N95 masks underwent QFT. Of these, 240 masks underwent post-processing analysis: 205 were 3M 1860 masks and 35 were PFR95 masks. Twenty-five (12.2%) of the 3M masks and 10 (28.5%) of the PFR95 masks failed post-processing QFT. Characteristics of the failed masks included mask deformation (N=3, all 3M masks), soiled masks (N=3), weakened elastic bands (N=5, three PFR95 masks), and concern about mask shrinkage (N=3, two 3M masks). NIOSH testing demonstrated that while filter efficiency remained >98% after two cycles, mask strap elasticity decreased by 5.6% after reprocessing.

Conclusions

This study demonstrated successful quality control implementation for N95 mask disinfection, and highlights the importance of real-world clinical testing beyond laboratory conditions.

Introduction

The current pandemic and global health crisis related to coronavirus disease 2019 (COVID-19) has infected more than 584 million individuals worldwide, with more than 1 million deaths in the USA alone since the winter of 2019. In the USA, more than 91.9 million cases had been confirmed as of August 2022 [1]. Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) is spread by aerosols and droplets expelled at increased rates with coughing. Studies have suggested that viable aerosols can linger in the air for up to 3 h and can spread beyond 6 feet in environments with positive pressure room ventilation, as commonly encountered in healthcare facilities [[2], [3], [4]]. Additionally, data have suggested that settled viral particles can persist and remain viable on surfaces for up to 72 h. With the risk of transmission and concerns related to the sequelae of COVID-19, the pandemic has had a significant impact on health care.

One specific area affected is the hospital supply chain. Personal protective equipment (PPE), including N95 masks, is in high demand across healthcare systems. N95 masks are rated to filter 95% of particles ≥0.3 μm, and are used to protect healthcare workers from aerosol transmission of virus. In addition to the increased demand, there is significantly decreased supply with global manufacturers, many based in China, unable to resume operations. Due to these constraints, several federal recommendations were published regarding the potential re-use and extended use of medical-grade N95 masks that are marketed and approved for single use. Prior literature related to the sterilization and reprocessing of N95 masks led to several institutions implementing programmes to prolong their N95 mask supply [[5], [6], [7], [8], [9]]. The technologies utilized include vaporized hydrogen peroxide (VHP), ultraviolet germicidal irradiation (UVGI), heat and hydrogen peroxide gas plasma (HPGP).

In considering new processes and technologies, quality systems play a critical role in the transitioning of any innovation into clinical practice. If quality systems are neglected, the lives of both the patient and provider may be placed at risk if the quality process fails to ensure that the product performs as intended and delivers the intended protection. As such, quality monitoring is a requirement of many regulatory bodies, including the Food and Drug Administration (FDA). FDA's regulations are described in the Code of Federal Regulations Title 21 Part 820 entitled ‘Quality System Regulation’ [10]. A complete listing of the requirements is beyond the scope of this article, but it is important to note that these requirements should be reviewed in the early stages of introducing any potential product or process.

From a device perspective, during normal FDA operations, the main unit that oversees medical devices is the Center for Devices and Radiological Health. There are three major pathways for regulation based upon risk; these are, in ascending order of risk, Classes 1–3. Most Class 1 devices are exempt from premarket clearance, and Class 3 devices require a rigorous process to demonstrate safety and efficacy before they can be used on patients. Alternatively, an Emergency Use Authorization (EUA) permits the use of unapproved medical products (drugs, biologics and devices) or the use of approved medical products in previously unapproved ways to diagnose, treat or prevent diseases or conditions caused by chemical, biological, radiological or nuclear agents. It is important to note that the EUA is not part of the normal pathway, and can only be utilized during emergencies. The EUA is in effect for 1 year or for the duration of declared emergency by the Department of Health and Human Services.

This article will focus on the process, quality control measures and results of a reprocessing method employed at a single institution, and will highlight the importance of continued in-hospital quality control even with methods cleared by FDA under current EUA. Although not mandated specifically by an EUA, a quality management system was adopted early at the study institution to ensure that all reprocessed masks functioned predictably to specification once distributed. In addition, during their utilization, performance of the masks was monitored so that issues were reported as they were identified and could be corrected quickly.

Methods

In the setting of an emerging pandemic and limited N95 mask supply, the Penn State Milton S. Hershey Medical Center (PSMSHMC), a 628-bed academic medical centre, immediately investigated methods to prolong the supply of N95 masks through reprocessing. A multi-disciplinary team was assembled with experts from infection control, infectious disease, pulmonary/critical care, gastroenterology, facilities management, public health, occupational medicine, environmental health and safety, sterile processing, logistics and operations support, marketing, nursing, surgery and anaesthesia (N95 Taskforce). Members included clinicians performing aerosol-generating procedures previously demonstrated to be at high risk for transmission of viable viral particles [11]. Ethical considerations included in crafting an appropriate reprocessing plan included ensuring multi-disciplinary input into the study design, maximizing equitable availability of reprocessed masks to all front-line clinical healthcare workers, and ensuring strict sterilization procedures and timely data analysis to ensure appropriate safety to participants. Participants using masks included in the study were informed of reprocessing, and assented to quality control testing during the mask fit and distribution process.

After reviewing the available literature and considering hospital resources and expertise, the N95 Taskforce decided to utilize the Sterrad HPGP reprocessing system for mask sterilization. This process had received EUA from FDA for up to two cycles of reprocessing. Additionally, this method allowed leverage of available equipment and trained personnel to deploy a new process for N95 masks. The Sterile Processing Department at PSMSHMC owns five 100NX Sterrad machines, enabling approximately 100 masks to be sterilized/reprocessed per hour. The specific sterilization procedures followed the instructions for use of the ASP Sterrad 100NX system using the 24-min express cycle [12].

To design the process, the team identified areas of high mask usage, including the operating room, emergency department, endoscopy suite and COVID-19 testing areas. The group implemented several quality control measures for extended re-use and single-use reprocessing. Policies regarding N95 mask re-use and appropriate doffing and donning procedures were shared via hospital intranet, mailings and messaging placed at mask collection areas.

The group also initiated universal mask fit testing after mask reprocessing to provide post-processing evidence that fit and filtration were not affected. Nurses trained in Occupational Safety and Health Administration qualitative mask fit testing (QFT) were assigned to a designated area where all personnel receiving a disinfected mask would undergo QFT using either a saccharine or bitter solution [13]. Testing of processed masks was performed utilizing a Draeger Accuro pump (Draeger Inc, Houston, TX, USA) with a hydrogen peroxide calorimeter tube. In addition, an ATI PortaSens II meter (ATI, Collegeville, PA, USA) was utilized to evaluate residual hydrogen peroxide levels from within the masks. Any personnel that failed QFT repeated QFT with a new mask of the same brand and size. If the repeat QFT failed, individuals were resized for a different mask that fit appropriately. Masks that failed fit testing underwent further evaluation to assess any external contributing factors, including soiling of the internal lining or visible mask deformation.

The processes designed by the N95 Taskforce are outlined in Figure 1 (mask collection and redistribution) and Figure 2 (mask sterilization).

Figure 1.

Mask collection and redistribution. This process aimed to return disinfected masks to the original healthcare worker to increase employee acceptance and to reduce additional stress on a mask by repeated deformation or fitting to different facial structures.

Figure 2.

Mask sterilization in sterile processing. The sterilization procedure once the masks were received in sterile processing.

Concurrent with the initiation of this process, masks were sent to the National Institute for Occupational Safety and Health (NIOSH) laboratory in Pittsburgh, PA, USA for filtration evaluation after two reprocessing cycles. These masks were tested per the NIOSH Standard Test Procedure TEB-APR-STP0059 using TSI 8130 (TSI Inc, Shoreview, MN, USA), which includes testing for filtration as well as fit on standard mannequins [14]. Fit testing was performed using mannequins and the TSI PortaCount Pro+ 8038 (TSI Inc). This protocol also included testing the strength of the elastic straps of the masks.

Per NIOSH protocol, only unworn processed masks could be tested. In parallel, members of the N95 Taskforce designed a local testing rig to mimic the NIOSH testing protocol, and test filtration efficiency using a Fluke airflow meter and a Fluke 985 particle counter (Fluke Corp., Everett, WA, USA) designed to measure particles between 0.3 and 10 μm. This rig allowed for the testing of previously worn disinfected masks. Background particle counts in the air outside of the mask were compared with counts in the air drawn through the mask (Figure 3 ). This rig was used to evaluate the filtration efficiency of used masks that had been reprocessed twice and a few masks that failed QFT. The full test protocol can be found in the online supplementary material.

Figure 3.

The local mask testing rig. (A) A front view of the rig demonstrating how the mask to be tested was secured to the rig in an airtight manner using a compression plate. (B) A profile view of the rig showing the motor that generated air flow across the mask. (C) A demonstration of the testing procedure with the particle counter.

Results

After 5 weeks of implementing the process in high-use areas, 982 masks underwent reprocessing for staff re-use. Of those, 255 masks underwent post-processing QFT (ppQFT) (26%). Two types of masks were reprocessed: the 3M 1860 mask and the Kimberly-Clark Tecnol PFR95 mask or similar Halyard Fluidshield mask, which were both available in regular and small sizes. Fifteen of the tested masks were excluded from the subsequent analysis as the user required a different brand or size of mask than what was initially worn and reprocessed. Of the remaining 240 masks, 205 were 3M 1860 masks and 35 were PFR95 masks (Table I ). Twenty-five (12%) of the 3M masks and 10 (29%) of the PFR95 masks failed ppQFT. Notable characteristics of the failed masks included deformation of the mask (N=3, all 3M 1860 masks), soiled masks (N=3, Figure 4 ), compromise of the elastic bands (N=5, three PFR95 masks), and increased tightness of the mask fit on the individual (N=3, two 3M 1860 masks). The remainder of the masks that failed ppQFT did not show visible soiling or deformation, and the individuals did pass subsequent/repeat QFT with a new mask of the same size and brand. Draeger Accuro pump and PortaSens testing determined the presence of <1 part per million (PPM) (range 0–0.7 PPM) of residual or measured hydrogen peroxide at several locations on both brands of masks with at least a 1-h aeration period out of the sterilization pouch.

Table I.

Breakdown of mask testing and noted failures

Mask characteristic	N (%)
3M 1860
Reprocessed
Underwent QFT	216
Failures excluded due to incorrect initial fit	11
Failures included	25
Soiled by cosmetics	3
Visible damage	3
Felt tight	2
Elastic damaged	1
Kimberly-Clark PFR95
Reprocessed
Underwent QFT	39
Failures excluded due to incorrect initial fit	4
Failures included	10
Elastic damaged	3
Felt tight	1

QFT, qualitative fit testing.

Figure 4.

Masks showing significant soiling from facial cosmetics.

Data received from NIOSH regarding the filtration testing of PFR95 masks after two reprocessing cycles demonstrated that while the filter efficiency was intact after two cycles (>98% across all samples), there was evidence of compromised elasticity and a decrease in strap force of 5.6% for the top and bottom elastic straps after reprocessing.

Two 3M masks that had passed fit testing after personnel use and two reprocessing cycles were tested with the local testing rig. Additionally, two masks that had failed fit ppQFT were tested. Table II depicts the results of this filtration testing for each of the four masks (five tests each).

Table II.

Mask filtration results (all 3M 1860)

	Passed QFT		Failed QFT
Test #	Mask 1 efficiency (%)	Mask 2 efficiency (%)	Mask 3 efficiency (%)	Mask 4 efficiency (%)
1	96.5	97.1	84.5	80.9
2	96.5	97.4	85.6	82.1
3	97.1	97.2	85.6	82.3
4	96.3	97.1	83.8	81.5
5	96.6	97.1	84.0	81.8
Average	96.6	97.2	84.7	81.7

QFT, qualitative fit testing.

Discussion

Since the onset of the COVID-19 pandemic, several sources have described acute health system shortages of N95 masks, and rapid implementation of existing and novel sterilization methods to meet demand [7,[15], [16], [17], [18], [19], [20], [21], [22], [23]]. However, few have described the quality control process critical to successful implementation, scaling, quality assurance and improvement of such systems.

In addressing the shortage of N95 masks, the quality management systems at the study hospital focused on two components when reprocessing masks. The first component was quality assurance, in which the authors ensured prospectively that the chosen sterilization method was not only scalable, but also designed to allow the packaging, labelling, documentation and ancillary services to be supported within the hospital system. The second component was quality control, in which, following the implementation of a sterilization method, the authors tested and confirmed that the specifications of the original product were still being met prior to releasing the product to hospital practitioners.

Furthermore, quality audits were designed and implemented to ensure that the quality system followed regulatory requirements (Quality System Regulation Section 820.22) [10]. The importance of this process can sometimes be lost when trying to deliver innovations rapidly to an end user. Some even view regulatory requirements as burdensome, rather than as a mechanism to improve product effectiveness. On the contrary, the authors found that they were able to test rapidly and ensure a quality product, despite monitoring and assessing a large quantity of samples. The authors started with a model based on the reprocessing methods being performed at other institutions, and this was then adapted to institution-specific needs and requests.

Initially, the process was designed such that masks would be sterilized and placed into a general pool. However, stakeholders and mask users were unwilling to wear a mask previously worn by another individual, despite sterilization. Additionally, the authors were concerned that the lifespan of a mask would be shortened if it had to fit different facial structures after each sterilization. The process consequently evolved, allowing masks to be identified with single users through each sterilization. Similarly, other findings identified during real-time monitoring, such as aeration requirements based on residual hydrogen peroxide, led to changes consistent with the continuous quality improvement model [24]. The authors successfully modified the process in real-time to accommodate user need and safety.

Importantly, the process implemented highlights the effectiveness of quality control measures when performed under real-world conditions. Bessesen et al. recognized the opportunities for error in mask disinfection, and how well-designed standard operating procedures only provide adequate quality control when followed precisely [25]. Likewise, studies by Liao et al. and Ou et al., determining the effectiveness of methods of N95 mask disinfection under laboratory conditions, noted that factors associated with real-world clinical use can alter laboratory-determined contamination rates and filtering efficiency [17,26]. They posited that users' repeated donning of masks and associated fit and leakage changes might impact filtering efficacy and reusability. These findings were demonstrated under real-world conditions, as some of the post-processing mask failures in this study were due to incorrect fit, deformities or elasticity changes after repeated use (Table II).

In the laboratory setting, studies have demonstrated that changes to masks can vary by sterile processing modality. For example, while Liao et al. observed no qualitative change in mask elasticity or fit after heat treatments, they found that ultraviolet treatment of masks could lead to improper fitting [18]. Price et al. noted that previous studies of UVGI mask sterilization technology conducted fit tests on static mannequins, while dynamic testing on human subjects could expose fitting inadequacies not seen in artificial controlled settings [18]. The Sterrad HPGP-based reprocessing system was chosen for use in the present study due to evidence from past studies demonstrating the effectiveness and superiority of VHP-based sterilization compared with other modalities, with little impact on mask fit or filtration efficacy [23,27,28]. VHP's high filtration efficacy maintained after early cycles of sterilization was corroborated by NIOSH filtration testing of the study masks after undergoing two reprocessing cycles, which showed 96–97% mask efficiency in those masks that passed QFT after sterilization [28]. However, as noted, the masks undergoing NIOSH testing were not used under clinical conditions. Previous studies of HPGP have shown a reduction in filtration efficiency by >5% in some masks associated with degradation of mask components; this issue was seen among the masks that failed QFT in the present study (Table II, Masks 3 and 4) [22,23,27,28].

This study monitored the real-world effectiveness of mask reprocessing through QFT, and demonstrated a mask failure rate of 12.2% (25/205) for the 3M 1860 mask after reprocessing. In performing a root-cause analysis of mask deformation, soiling and/or strap elasticity (measured qualitatively) were implicated. This demonstrated the need for continued quality monitoring due to the many external factors impacting mask efficacy after repeated use, regardless of sterilization method. This rate is similar to the second-pass N95 autoclave reprocessing failure rate of 14% seen in a Canadian institution [20]. Previous studies of HPGP have shown a reduction in filtration efficiency by >5% in some masks, associated with degradation of mask components [22,23,27,28]. This was also seen in masks that failed QFT in the present study, as filtration testing of certain masks that failed QFT did show a decrease in filtration efficiency to 80–85% (Table II, Masks 3 and 4).

The finding of decreased mask elastic strap force by 5.6% in unused Kimberly-Clark PFR95 masks following reprocessing as tested by NIOSH emphasizes the need to identify contributors to reprocessing failure beyond filtration, and shows that testing filtration rates alone may be an inadequate marker of reprocessing success. This study highlights the importance of incorporating quality control when implementing innovation in healthcare systems, and the feasibility of doing so in a timely fashion, even during times of critical need.

More broadly, with the rapid emergence of many mask sterilization/reprocessing technologies to address mask shortages that may lack rigorous testing in a clinical setting, this study emphasizes the importance of internal quality assurance beyond reliance on external supplier assertions. For example, while previous testing of autoclave reprocessing of unused masks demonstrated functional efficacy through 10 cycles, one hospital system's study of real-world clinical implementation found that reprocessing was inadequate beyond a single cycle [20]. This study found similar unanticipated user-related issues contributing to mask failure, including mask deformity with contouring and use, reduced strap elasticity, and soiling due to use of facial oils and cosmetics. These represent site-specific quality assurance issues that are not accounted for in the published literature, which may also arise at other institutions and compromise mask efficacy/reusability. Importantly, many sterilization systems that received FDA EUA were authorized solely based on published laboratory-based data given the urgent need. This study demonstrates how real-world clinical data reveal differences in efficacy that must be accounted for through internal quality improvement.

The strengths of this study are based on its large sample size, and demonstration of the efficacy of the quality control system, allowing real-time monitoring of mask testing and addressing challenges that arose under live conditions. The limitations of this study include the possibility of selection bias, as qualitative analysis was only performed in a subset of processed QFT masks. However, the 255 masks undergoing analysis were chosen at random, and represent a large sample size of >25% of those in circulation. Additionally, only 3M 1860 masks and KC PFR95 masks were examined in this study, limiting the generalizability of the utilized sterilization technology to other types of PPE. While other studies of Sterrad and VHP reprocessing systems may be more broadly applicable, it is suggested that institution- and site-specific assessment should be performed to evaluate local reprocessing efficacy, and the need for alternative or additional sterilization protocols [23,28]. Further study of mask quality control in other institutions can add to general understanding of the challenges of implementing new systems during periods of critical need.

In conclusion, this study demonstrated the process of creating a reproducible and scalable quality control system for N95 mask disinfection within a large academic hospital system. The study highlighted the importance of internal quality assurance and testing of new sterilization and reprocessing protocols in real-world clinical settings, rather than solely under laboratory conditions. Quality assurance is achievable through multi-disciplinary collaboration, evaluation and rapid real-time adjustment to meet demand and address critical mask shortages during the COVID-19 pandemic.

Author contributions

NG, DG, WH, WS, CU, JS, GF, TU, NT, RB: conception and design of the study, or acquisition of data, or analysis and interpretation of data.

NG, DG, WH, JS: drafting the article or revising it critically for important intellectual content.

NG, DG, WH, JS: final approval of the version to be submitted.

All authors approved the final version of the manuscript.

Conflict of interest statement

None declared.

Funding sources

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

Appendix A. Supplementary data

The following is the Supplementary data to this article.