A University of Toronto and University Health Network APIL collaborative, with CIGITI-SickKids, Queen’s University, Lakehead University, NOSM, and USASK.
The COVID-19 pandemic has unveiled widespread shortages of PPE across many sectorsof essential services worldwide. This has left healthcare and other frontline workers feeling exposed, vulnerable, and anxious. Canada-wide, volunteer-led PPE drives have successfully redirected small volumes of PPE from the public to healthcare facilities where they are intensely needed. Similarly, manufacturing industry leaders have answered the call by rapidly repurposing and dedicating their labour and capital to the production of face-shields, masks, and other PPE. These efforts have shown the resilience of our country in unwavering support for those charged with caring for patients with COVID-19 and the Canadian community.
However, the shortage of N95 respirators has posed a great challenge. While 3M continues to produce masks as rapidly as possible — they are dealing with hyper-demand worldwide, at times with government imposed restriction to exports. This has led us, among many across the world, to focus on developing a reusable, stop-gap N95-caliber mask solution, should supply disruptions of 3M N95 face-masks and respirators continue. Over the last 2 weeks, we have gathered a team of physicians, designers, materials and fluid dynamic engineers, and printing experts. This team traverses the community, four universities and respective teaching hospitals, including Toronto General Hospital and SickKids, and multiple Faculties & Departments within these university-hospital systems. We have resolved to collaboratively design, develop, validate, and disseminate viable options that can be up-scaled for widespread production and use by frontline workers. (a full list of collaborators can be found below)
Starting with readily available open-source designs, we are using an iterative approach to 3D print prototypes, followed by testing for form-fit and filtration-function by negative pressure particulate counts (“portacount”), which is followed by immediate remodification as informed by the previous round of data and feedback. We have sourced and quantified filtration efficiency of accessible filtering materials including commercially available anesthesia circuit Heat and Moisture Exchangers (HME), medical-grade bacterial and viral filters, various MERV-rated vacuum filters, HEPA filters, surgical wraps, and replaceable 3M filters.
“Montana mask” and derivatives (I3D, Kingston). This 3DP mask body with filter cap has gained most traction world-wide as a surgical mask stop-gap. It was designed to encase any sort of flat filtering material for wearers. The mask rim-contours required modification to fit most face-types and has significant seal deficits (see section Seal). With double rubber lining around the rim-contours by Billings clinic, combined with NIOSH N95 filter material, the originator team was able to get adequate seal and filtration, and passed fit-factor testing (see section Fit Factor Testing). Our UHN APIL experience is mixed: we think these are great stop-gap surgical face masks, but seal and filter functions need vast improvements to become N95 options.
NIH Stop-gap Face Mask aka “SFM” (NIH 3D exchange). This open-source and clinically tested mask body with filter cap is available on the NIH 3D exchange. We printed these in nylon in original Small and Medium sizes, and found significant leaks around the rim-contour when worn. To overcome this, silicone inlays were mould-casted with noticeable improvement on seal and comfort. This mask body still requires an X-Small size for most face types, and new versions have not been uploaded.
Montana derivatives, SFM, and other simple filter solutions were readily 3D-printable, but all suffered from seal deficiency and filter surface area deficiency — we will return to this very important point below.
“Simple Silicone Respirator” MB-ON (Christian Petropolis HSC -> UHN APIL Toronto). Originating from Manitoba, Canada, this is a unibody silicone mask, that was specifically aimed at N95 respirator replacement. The mask is fixed by a printed and thermoformed harness, and strapped onto the wearer by silicone straps. This stop-gap solution’s current configuration has a silicone opening for direct adapter connection to most commercial medical-grade pleated filters, which greatly enhances ease of breathing by increasing effective filter surface area.
To address the poor seal of the rim-contour to the face, we first attempted a DIY solution by adhering window weather stripping around the “Montana” Mask’s rim-contour, with gaps filled with silicone. Although the seal improved, we could not eliminate the leaks for all users. For the “SFM” mask, we custom-designed a mould for Small and Medium silicone inlays, that would allow multiple points-of-contact with the wearer’s skin. The inlays were cast Shore 20A–30A grade silicone, which improved but did not remove all air-leaks. The most common points of failure are in order: the nose bridge, the naso-labial folds, and lastly the labio-mental folds. Excessive rim pressure and post-wear marking on the face were noticeable with thinner mask rims, which means the mask would be tolerable for minutes but not hours of wear.
In contrast, the Simple Silicone Respirator aka “SSR MB-ON” by nature of its silicone body and rim allowed contour fit around most face shapes. Dragon Skin 30, 20, then 10 were sequentially tried, with tolerable durometry at Shore 10 to 20 score. The two straps buckle into the four legs of the harness, and need to be of the right length. We found that a Shore 10 score was adequate for stretch and strength. The prototype was comfortable enough for hours of wear. The Medium size fits a small, medium, to large face by measure of naso-mental distances from 100 to 130mm.
The gold standard we used were 1) stand-alone medical grade Intersurgical HME and DAR Air Guard filter, 2) 3M 5N11 Respirator Filter replacement. Other materials were compared against these commercial products. Using an AccuFit 9000 quantitative fit testing machine, following CSA Standard Z94.4–18, we tested the “Fit Factor” of each filter material, and compared these to a piece of a commercially available N95 mask. According to the standard an N95 mask must achieve a Fit Factor of 100 — which was indeed achieved by all N95-rated material we tested. From our sourced filter material options, we found that single layers of MERV-14 and MERV-15 filters were inadequate — reaching a fit factor of only 40–50. Two layers, however, achieved fit factors exceeding 500. Another widely-considered filter option — Halyard surgical instrument wrap — proved to be inadequate (H100 and H400 tested). These only achieved fit factors of 25–40 even when tested as double layers. Of note, a single layer of Halyard H100 wrap combined with a single layer of MERV-14 achieved a fit factor of 107. Taken together, these tests have shown that among the easily sourced materials we have sampled, there are filter material options that should allow candidate masks to pass N95-standard quantitative fit testing (QNFT), given adequate mask seal and air-tight filter encasing.
Fluid Dynamics Testing
During initial particulate filtration testing, we discovered that the pressure drop across the filters used in our masks, with a surface area of <50 x 50 sq mm, resulted in significant flow acceleration due to excessive pressure drops across the filter. This increases the drag of particles across the spectrum (20 to 1000 nm) through the filter, and amplifies any imperfections in the seal around the wearer’s face. To address this, we will need to design and perform pressure-drop testing to determine the ideal surface area for each of our proposed filter materials, based on a flow rate of 30 to 80 L/min (passing NIOSH 95 testing standard) and a maximum of negative 5 cmH2O pressure drop (tailoring to human comfort). Breathing at negative 1 to 5 cmH2O represents a tolerable level of work of breathing, and not the maximal inspiratory force a healthy person can generate. For reference, during quiet breathing, we generate ~ negative 1 cmH2O through our lungs during inspiration, whilst expiration is passive.
Laboured work of breathing is not the only important effect of inadequate filtration area. As airflow takes “the path of least resistance”, if the pressure gradient is excessive through the filter, then airflow will bypass around the mask, meaning proportionately more airflow will seep around seal leakages along the mask rim-contour, instead of through the filter system. This in turn increases entrainment of unfiltered airflow and worsens the particle count within the mask, resulting in an overall filtration system failure.
This is in sum what we referred to as surface area deficiency above: inadequate filtration cross-sectional area results in greater pressure gradient, higher flow velocity, poorer filtration efficacy, as well as subjective intolerable work of breathing.
Quantitative Fit-Testing — The Clinical Keystone
We see that apart from designing and production, one of our essential charges is testing and validation of solutions. That is what we understand by the term “clinically proven”. We have used quantitative fit-testing (QNFT) as outlined above to select the most workable solution of SSR MB-ON for health care institutions, which has tested 200 in overall for fit-factor of the QNFT rubric for different face types. But we have not yet been able to endorse any current 3DP alone stop-gap option, because of seal deficits, filtration surface area limits, resulting in underperforming overall QNFT.
Qualitative fit-testing (QLFT) will further complement QNFT, and ensure that any individual fits a particular mask-type and size. The most common technique is use of Bitrex, a noxious chemical, in a subjective sensitivity test. Institutional occupational health and safety departments need to be involved in individual fit-testing, just as if the wearer were getting fitted for a 3M respirator.
Design & Build
We are refining the geometry of the “SSR MB-ON” mask to create firmer nasal bridge padding and greater rim-seal cushion. This will also help in prevention of wear and tear of the silicone rim. We want to provide the best comfort by making smoother moulds and optimizing silicone grade. The geometry of the harness could be improved for extra support at more common points of seal failure. In fact, Dr. Petropolis at Manitoba HSC suggested removing the harness altogether with incorporation of buckles into the silicone body. Silicone straps could be exchanged for economical prefabricated straps after sourcing a supplier. More importantly, significant silicone-mould & casting changes have been made to ensure a smooth mask rim that contacts the face.
The Lakehead engineering team and TEDA (UofT Toronto Emergent Device Accelerator) are working in parallel to perform fluid-dynamic testing to determine the ideal surface area for each filter type. By fixing the maximum comfortable negative inspiratory pressure aka “pressure gradient”, an estimation of minimum area for each filter type can be ascertained.
We have started a complete re-design of a press-fit or screw-top cartridge to contain potential filters with adequate surface area, based on findings above. From our research, we think that a dual-cartridge system, adapted onto a unibody silicone mask would achieve this.
Standardized decontamination protocols already exist through many face-shield PPE project groups, namely through GLIA inc., one of our not-for-profit medical device partners. A 5000 ppm of sodium hypochlorite bath (1:10 Chlorox-TM) is an example of these published solutions, and is acceptable for the decontamination of faceshields. The silicone mask, straps and 3D-printed harness can be similarly treated. Larger institutions will have autoclave and vapour-phased hydrogen peroxide treatment, which are not available to smaller groups and individuals.
We are in the process of internal advertisement, departmental sampling and volunteer testing. A cogent and workable stop-gap option needs to be presented to Directors of PPE Response, Infection Prevention and Control, Clinical Teams, who are the frontline wearers and end-users of stop-gap solutions, and also administrative parties.
As this project is currently unfunded, we depend on and owe thanks to volunteer personnel, unpaid labour and time, pro bono printing partners and materials, student-led organisation, and vital knowledge and information sharing amongst our collaborators.
Not only should solutions be clinically tested and validated, they must attain as far as possible, given current time- and resource-constraints, Health Canada requirements and comply with federal, provincial and institutional regulations with respect to novel medical device and PPE in response to COVID-19 pandemic. Both Manitoba HSC and UofT UHN teams will seek approval for Medical Devices Establishment Licensing or “MDEL” with Health Canada.
Upon arriving at final design(s) that 1) provide an adequate and comfortable seal, with easy work of breathing, 2) are readily decontaminable and reusable, and 3) potentially reach a particulate filtration efficiency of 95% for 30nm particles (as required by NIOSH), we will recruit ~ fifty volunteers to perform quantitative fit testing to ensure the mask provides an adequate fit across a range of individuals. When this is complete, we will disseminate the final design(s) for widespread open-use. If necessitated by supply disruptions and continued demand, these design(s) will be ready for large-scale production by industry partners in Ontario and beyond.
William Ng, MBBS, MMed (Faculty of Anesthesiology, UofT) and Andrew Syrett, MD, MSc (Fellow in Anesthesiology, UofT) are from the UHN Toronto General Hospital, Department of Anesthesia and Pain Management. William is also a principal investigator with UHN Advanced Perioperative Imaging Lab, APIL.
Collaborators and Partners include:
University of Toronto | UHN Toronto General Hospital
- William Ng (Lead, UHN APIL Principal Investigator, Anesthesia)
- Azad Mashari (Director of UHN APIL, Anesthesia)
- Andrew Syrett (Testing, Anesthesia)
- Matt Ratto (TEDA Director, Information)
- Eric Greenwood (SickKids Advisor, Director of Equipment)
- Devin Singh (Testing, Design Input, Anesthesia)
- James Scott (Occupational & Environmental Health)
- Alon Vaisman (UHN IPAC physician)
- Carla Luzzi (Anesthesia)
- Binu Jacobs (Research Scientist and Associate)
- Graham Roche-Nagle (Vascular Surgery)
- Dmitry Rozenberg (Respirology & Lung Transplant)
UHN Advanced Perioperative Imaging Lab (https://apil.ca)
- Joshua Hiansen (Design Engineering, Print Lead)
- Vahid Anwari (Production Manager, Medical Radiation Technologist)
UofT School of Applied Science & Engineering
CIGITI, SickKids https://www.cigiti.ca
UofT MD Coordinators
- Andy Edem Afenu (R&D)
- Jessica Trac (R&D)
- Sinthu Senthillmohan (Communications)
- Yousuf Ahmed (Communications, PR)
- Lakehead University, Engineering and Sustainability Sciences
- NOSM’s Parallel Initiative, Division of Clinical Sciences
- Queen’s Parallel Initiative (SoM)
- UofT MD Program | Hamilton SoM “3D-PPE-GTHA”
- USASK, Engineering
Many Community Partners
- Andrew Pho, Lancotek
- James Garel-Jones, VertiGrow
- Andrew and Jessica Laura Schroeder, GLIA inc., Design
- Quang Nguyen, CEO 247 Security
- Steve Cory, Objex Unlimited
- Brian Read, Ex-CEO Horizon Plastics, Cobourg
GLIA NFP partner — https://glia.org