In his 25 years as an industrial designer, Tod Corlett has devised lighting systems and electronic signs, and guided student projects in such diverse realms as furniture and footwear. On March 18, the Thomas Jefferson University professor got an urgent call from his dean with a new kind of challenge: Can you make us a ventilator?
The university was on full coronavirus lockdown, so Corlett could not get into his lab due to the very problem he was being asked to solve. But he started the research from his home in Philadelphia, trading insights and questions with a growing online group of designers, engineers, and medical professionals.
How much air had to be pumped into a patient’s lungs? How should the exhaled breaths be filtered, to prevent the virus from spreading? Could a makeshift contraption really approximate a sophisticated medical device without years of testing?
Three days later, with special authorization to be on the East Falls campus, Corlett assembled a motorized prototype from pieces of metal, plastic, and wood. It took him six hours.
The next day, he took it apart and built version two.
“It doesn’t have to look pretty,” he said. “It doesn’t even have to work pretty. It just has to show you something you didn’t know before, so you can move on to the next one.”
Eventually, such a device would have to work “pretty” enough to pass medical scrutiny — a challenge that hundreds of labs worldwide are grappling with, at a speed none ever imagined. At the University of Florida, engineers have made a device with PVC pipe and lawn-sprinkler valves. At New York-Presbyterian Hospital, technicians have used tubes and valves to “split” commercial-grade ventilators, allowing one machine to be used by two patients. Others have explored ways to repurpose the CPAP machines used by people with sleep apnea.
Most people infected with the coronavirus experience mild symptoms and do not need hospitalization, much less the mechanical support of a ventilator. But a sudden surge in critically ill patients can quickly outstrip a hospital’s supply of the high-tech breathing machines, forcing physicians to choose which people with fluid-filled lungs can do without.
Depending on how soon a surge arrives in a given area, the innovators working on alternative devices might have weeks to get the job done.
“We’re just on an ‘as fast as possible’ timetable at this point,” Corlett said. “No one knows how long we have.”
He and his colleagues are working on several approaches, hoping that one could soon be tested by their medical counterparts at Jefferson University Hospital. And like many of the more than 50,000 who have joined the online innovators’ group, called Open Source COVID19 Medical Supplies, they pledge to make their designs freely available to the public.
The group was founded March 10 by robotics engineer Gui Cavalcanti, cofounder of Hayward, Calif.-based MegaBots Inc., who invited ideas for rapid production of ventilators, masks, and other hospital goods in short supply.
Members have been sharing data, guidelines, and advice online, including answers to frequently asked questions.
“Are 3D-printed parts sterile?” (Answer: probably not, as the materials tend to be porous and could harbor microbes. And if treated with hospital sterilization methods, they could warp, melt, or otherwise be weakened.)
“Will I get sued for making things?” (Answer: Good Samaritan laws may provide legal protection for those who volunteer their help. And the U.S. Food and Drug Administration has spelled out steps for manufacturers to seek “emergency-use authorizations” for new products.)
Caution is warranted, particularly with respect to ventilators. Forcing too much oxygen into a person’s lungs can cause permanent injury, especially when they already are inflamed with disease, said physician Greg S. Martin, a lung specialist at Emory University School of Medicine.
Physicians learned just a few decades ago that they had been overdoing it. An Australian researcher found that by decreasing the pressure and volume of each mechanical breath, his patients were able to come off a ventilator faster and were more likely to survive, said Martin, president-elect of the Society of Critical Care Medicine.
A mechanical puff of about 400 milliliters — a little more than a soda can’s worth — is now considered appropriate, though the precise amount needed varies with the patient’s body weight.
Corlett’s first two prototypes made use of an off-the-shelf device called a bag-valve mask. Typically, a clinician squeezes this type of football-shaped apparatus to deliver air to a patient when a ventilator is not readily available, such as in an ambulance.
But someone with acute respiratory distress from the coronavirus needs breathing support for days. Squeezing these bags is tiring after just an hour or two, as legions of medical students discovered when they were put to the task during the polio epidemic.
More important, it is hard for bag-squeezers to tell whether they are squeezing the right amount. And if they are in a room with coronavirus patients for extended periods, there’s risk of infection.
The solution that Corlett and others have explored: Squeeze the bag with a motor.
In Corlett’s first version, the squeezing was done with a wooden arm attached to a vertical aluminum post. It was a crude, seesaw-like contraption for which the method of squeezing was less important than determining whether the computer-controlled motor was the right size to move the necessary amount of oxygen.
“It was about figuring out forces and leverage and distance and volumes,” the designer said.
Version two, which took him four hours to build, was more elegant, with fewer moving parts. The motor rotates a short section of PVC pipe with a nylon strap attached. The other end of the strap is stapled to a wooden board placed over the bag, so that each turn of the motor delivers a precisely calibrated squeeze.
Yet for a variety of reasons, Corlett is not close to being satisfied, so he and colleagues already are at work on alternatives.
In the event that off-the-shelf ventilator bags are hard to come by, one designer is working on a device that would squeeze a plastic bellows, the kind normally used to inflate a raft. Another option is to get rid of the motor, to reduce the number of moving parts. Instead, the bag or bellows would be placed inside a larger container, so that it could be squeezed with bursts of compressed air.
The Jefferson team also is working on what Corlett and colleague Matt Umbriac call a “wearable clean room” — essentially, a bag sealed over the patient’s head, with oxygen going in one side, and exhaled air leaving on the other, through special valves and filters.
It may sound improbable, but San Diego anesthesiologist Yuri Gelland demonstrated such a device in a short video on Twitter, drawing more than 200,000 views by Friday.
Corporations also are trying to ramp up production of traditional commercial ventilators, at the urging of President Donald Trump.
But with many parts in short supply and the possible difficulty of delivering finished products from manufacturers to the hardest-hit areas, the various makeshift efforts at universities may prove crucial, said Samsun Lampotang, the engineer overseeing the University of Florida ventilator project.
“The idea is, every region should be self-sufficient in building more ventilators,” he said. “Hop in your car, and go to Lowe’s, Home Depot, and Ace Hardware.”
Though he is an engineer, Lampotang is a professor at the University of Florida College of Medicine and has ready access to physicians who can give him feedback. Ditto for the Jefferson designers, who are under the same institutional umbrella as Kimmel Medical College and the Jefferson health system.
That connection will prove essential if the ventilator crunch comes to pass, Corlett predicted. The device has to work to the satisfaction of those who will use it.
“Reliability is absolutely key here,” he said. “If it’s not going to quickly and quietly and reliably and safely do its job so our doctors can get on with patient care, it’s not something that’s worth putting in place.”