Since March 2020 our team at APIL has redirected the bulk of our technical and manufacturing capacity to developing, evaluating and manufacturing medical devices as part of the hospitals pandemic response.
This work has been possible through collaboration of a wide network of individuals, groups and private sector partners; access to world class infrastructure at UHN (both at APIL and beyond), and financial support from the foundation which provided us with the capacity to undertake projects rapidly and secure further grant funding.
The UHN Departments of Anesthesia and Pain Management; Medical Engineering, Respiratory Therapy, IPAC, MDRD, and Animal Lab.
Glia Inc, a medical device development company affliated with London Health Science Center;
U of T’s Department of Mechanical and Industrial Engineering;
General Dynamics Land Systems - Canada
Our main projects addressed two broad areas in which multi-level contingency plans were required in the event that our main resource base became overwhelmed:
At the same time, shifting towards device manufacturing required significant modification and adaptation of our physical and administrative infrastructure in order to ensure regulatory compliant quality control and manfacturing practices. This included set-up of a 3D printing cluster that allows centralized control and monitoring of the entire printer process in order to identify unanticipated events that may impact the quality of manufactured parts. This is part of our long term goals of creating a medium scale manufacturing facility that can rapidly adapt to institutional needs, including in the event of major crisis causing supply chain disruptions. Such a facility will be a significant contributor to institutional resilience in difficult situations.
Our ventilator projects pursued three different strategies to address the potential ventilator shortage:
This project was initiated with Glia in collaboration with General Dynamics Land Systems Canada, which during non-pandemic times makes armoured personnel carriers.
They offered their facilities and an incredible team of highly disciplined engineers, designers, and other experts. We met for 15 minutes every morning, 7 days a week.
The device went through a few iterations. It became clear that modifying an existing firefighter’s mask to connect to standard ventilator connectors was the simplest solution.
We had to design a 3D printable adapter and figure out how to print it reliably so that it did not leak and passed quality control consistently.
BIPAP masks are not generally comfortable and this one is no exception. So far the patient feedback has been encouraging, but further human factors testing is required.
Mouth piece on maks is not transparent, so patients who are week need to be monitored closely for vomitting.
Leak reduced by a factor of 12-30 compared to standard NIV masks, but still around 1L/min, but directed towards the back of the patient (i.e. pillow, bouffant cap) instead of laterally.
Specifically see simulator at: https://ventilator-simulator.now.sh/
Diagram is on the page and I can send the original vector file if your team wants to edit it.
There are numerous splitters that have been publicized during the pandemic. Most of them basically used Y-connectors to split the limbs of the ventilator. Most of these systems are very dangerous, even for brief use. The Cerebrus system, while still limited in many respects, addresses all the key safety issues for short term use.
Respiratory rate has to be the same for both patients so volume control only (with pressure relief valve), requiring deep sedation for paralysis. Therefore only appropriate for brief intervals until patients could be transfered to a proper ICU vent.
Prototype was created and tested over three versions. Device fulfills original specification goals and meets most regulatory requirments for emergency ventilators. However the current architecture does not allow ready addition of key additional safety features that while not strictly required by regulators at this time, would be highly desirable and clinically and would greatly extend the useability of the device.
We are currently collaborating with U of T’s department of Mechanical and Industrial Engineering to develop other prototypes.
With support from NSERC we developed an evaluation framework for open source emergency ventilators that synthesizes regulatory requirements from Canada, US, UK, and Australia into detailed specifications and checklists to guide development and testing of future ventilator designs.
The framework will be openly shared upon publication. Manuscript in preparation.
See for images and data
Main technical problems:
The team tested a variety of materials for adequate filtration.
Several performed well, but making a reliable system for holding filters without leaking but allowing replacement proved too complex for our timeline. So we decided to use HME filters (>N95, low resistance, bidirectional) to start and focus on making a well sealing mask.
Several iterations of the design, test, revise. Testing consisted intially of manual seal check and wearing mask for 30 minutes during desk work, followed by quantitative testing following NIOSH/CSA standards.
We were able to quickly get to a prototype that passed quantitative testing with a high score on team members. The final prototype was cast out of silicone, from 3D printed molds and used Intersurgical airguard filters. We tested 40 subjects on both their official fit-tested N95 and on the final version of the SSM.
Disposable N95s perform quite poorly. The resuable silicone mask performed very well.
Working on Health Canada approval and further testing for comfort, usability and communication.
One manuscript in review. Another manuscript in preparation.
Great seal and filteration, Filtered expiration, Reasonable Comfort (comparable to disposable),
Weakness: communication challenging in current version (voice muffled)
See slide deck for images and data