In 1941, Britain, alone and isolated, faced the grim prospect of a starving submission to German U-boat wolf packs. The only hope for survival was a huge quantity of food from the United States, which required the Americans to create countless tons of shipping capacity. But building ships is a long, complicated process and time was rapidly running out.
The American solution was simplicity itself: Liberty Ships. The complete redesign of a standard cargo ship — literally from the bottom up — gave rise to a fleet of workhorse freighters that would eventually deliver food, munitions, and victory in Europe.
To build the ships quickly, the number of individual parts was hacked back drastically. Rather than having their hulls laboriously riveted together, Liberty Ships were made out of prefabricated parts that could be welded into place, a process that required less skill and shaved weeks off production. And because the design was simple and processes standardized, ships could be built at any one of 18 shipyards. By the end of the war, 2,700 of these "Ugly Ducklings" had been produced. By the end of 1943, American shipyards were averaging a launch every six hours.
We need a similar effort now that hospitals around the country are facing a terrifying shortage of ventilators — the machines that push air into the lungs of patients beset by respiratory diseases like COVID-19. Just as the Allies redesigned ships so they’d be cheaper and easier to produce, we need to reimagine the ventilator. Rather than racing to produce more of the same kind of expensive, state-of-the-art ventilators hospitals typically use, what we need now is speed and simplicity.
Reports from intensive-care doctors in Italy and China tell the same story: even before there’s a vaccine or antiviral medications, critically ill COVID-19 patients often will pull through if there are enough mechanical ventilation machines that can buy time for them to recover.
A new ventilator is a beautifully sleek, complicated piece of electronics running nuanced algorithms to deliver a near perfect “machine breath” to the patient. That perfection comes at a price — $20,000 to more than $40,000 — and requires specialized production facilities. Money aside, production time far exceeds what we have right now.
Once patients are on ventilators, we let them rest while the machines ensure they have the oxygen necessary for the lungs to heal. Modifications to these machines’ settings are planned by the medical staff as the patients’ conditions change or as test results become available. At the moment, this might involve summoning a specialist because nobody else dares touch the bewildering display.
But at its heart, an ICU ventilator is a profoundly simple device. Here’s how it works: the patient is sedated enough to tolerate having a breathing tube secured in the trachea. The breathing tube connects to the ventilator by a hose. As the patient initiates a breath, the machine detects a slight drop in pressure, and a piston-driven cylinder pushes a measured volume of oxygen into the patient’s lungs. If the patient doesn’t or can’t start a breath, the machine will automatically do so after a few seconds. At some point, the machine either successfully delivers the required volume or reaches a maximum safe pressure and automatically cycles so that air can come back out of the patient’s lungs and through a separate hose. Pause. Repeat. Minute after minute, hour after hour. If necessary, for days.
This is not a machine that needs to cost tens of thousands of dollars. The current models have touch screens, graphical displays, and highly nuanced modes such as “neutrally adjusted ventilatory assist” and automated ventilator-weaning protocols. Those features are nice to have in a tiny minority of cases, but they are unnecessary complexities in a mass casualty situation. So while Elon Musk very well may try to convert Tesla or SpaceX assembly lines to the production of ventilators, they shouldn’t need to be produced in factories at Tesla, or Apple, or General Motors. They should be produced everywhere.
At the other end of the technology spectrum, hobbyists and survivalists are posting designs for their own ventilators. These tend toward one-off homemade contraptions. Deployable medical equipment will still need approval from the Food and Drug Administration and should not be so reliant on duct tape.
In a city with the finest medical minds and universities, we should be able to design, test, and certify a scaled-down but effective ventilator that can be built in many sites around the country and the world within weeks.
That’s my challenge to the medical, engineering, and academic community: come up with a ventilator that delivers what we need in the ICU but is simple enough to be assembled in an unsophisticated factory using established techniques from prefabricated units. It should have as few parts as possible. All parts should be commonly available. It should be repairable by a hospital technician. Electronics should be simple, sensors replaceable, and controls adequate but not excessive. An experienced nurse or respiratory therapist should be able to master it within a couple of hours. It should be robust, easy to clean, and ugly.
We have all seen the graph — two mythical bell curves and a dotted line denoting the capacity of the health system. We are all self-isolating, desperately trying to flatten the curve. Who knows if that will work? But we also can raise that dotted line by increasing our capacity to support critically ill patients. We have probably missed our opportunity to get significant respirator capacity online for the first peak, but there will doubtless be other surges later this year and into next. To increase our capacity, we must produce and run thousands of simple ventilators and train up healthcare providers within weeks.
I don’t need a Tesla. Just give me an Ugly Duckling.
Dr. Ben Moor is an anesthesiologist at Beth Israel Deaconess Hospital in Plymouth.