Subtitles section Play video
This episode of Real Engineering is brought to you by Brilliant, a problem solving website
that teaches you to think like an engineer.
This subject needs no introduction. The entire world is talking about the same thing. Covid-19
has turned the world upside down in just a couple of weeks, and the human race is scrambling
to adapt to this new pandemic.
Covid-19 is racing through populations. For many the symptoms are mild, for others the
virus debilitates them. Their lungs' ability to transfer carbon dioxide and oxygen in and
out of the blood begins to fall. Each breath becomes more laboured, until eventually they
become too exhausted to breathe on their own. [1]
The patients that are presenting to hospitals with these symptoms need medical intervention
before they suffocate. Unfortunately we do not have a cure. The best medical treatment
currently is to simply assist their lung function with mechanical ventilation. But, hospitals
are experiencing such huge influxes of patients that they do not have enough ventilators to
cope.
This has led to a call out to any and all capable of manufacturing ventilators to join
the fight against this common enemy.
Medtronic, a company I have worked for in the past, has it's primary ventilator manufacturing
facility based in my hometown of Galway. [2]Here they produce this ventilator. The Puritan
Bennett 980 Mechanical Ventilator. A top of the line high performance ventilator that
would give doctors a huge leg up in combating the viral pneumonia that patients are experiencing,
but it's a complicated bit of machinery.
They currently produce 225 units a week and are pulling employees from the stent manufacturing
facility I worked at, to help boost production to 500 a week.
This however won't come even close to fulfilling the demand that is yet to peak. There is this
impending doom looming over us that soon hospitals in countries like the US will experience what
Italy has already experienced. An overwhelming flood of patients.
So many enthusiastic engineers who want to help out are volunteering their expertise
to develop a low cost ventilator that any manufacturing facility could adapt to build.
Every single design I have come across centres around one key bit of technology. A BVM, or
bag valve mask. BVM are plastic bags that a clinical care practitioner can manually
deflate with their hands. It's what a first responder would use if a patient wasn't
breathing, instead of giving mouth to mouth resuscitation. It's a cheap and easy way
to force air into the lungs. All these designs are basically just robotic arms that can squeeze
this bag at a set frequency endlessly.
They of course could be manufactured quickly and in great numbers, but ventilators aren't
just air pumps that force air into a patient's lungs.
One of the primary problems facing doctors currently is managing a side effect of mechanical
ventilation, barotrauma. [3]
To understand this we first need to understand how the lungs operate under normal conditions,
so let's have a quick biomechanics class on breathing with our sister channel Real
Science.
Two muscle groups typically act to control breathing. [4] The diaphragm, which is a large
muscle which separates the abdomen from the chest, and the intercostal muscles which are
the muscles which reside between the bones of your rib cage.
When you breathe in your diaphragm contracts, which causes it to move toward the abdominal
cavity, while the external intercostal muscles between the ribs also contract which lifts
the rib cage outwards. Both of these actions increase the volume of the thoracic cavity,
the cavity that your lungs reside in. The increase in volume causes a corresponding
decrease in pressure, which allows air outside the body at atmospheric pressure to fill the
lungs and equalise the pressure.
The key thing to note here is that negative pressure drives inhalation. The lungs don't
inflate like a balloon. They expand and equalize with atmospheric pressure. On exhalation the
process reverses with a small spike above atmospheric pressure to push the air out again.
Mechanical ventilation cannot work like this. It has to force air into the lungs from the
outside and essentially blow the lungs up like a balloon. If this is not tightly controlled
the air pressure could work against the diaphragm and the intercostal muscles and end up increasing
the pressure in the alveoli above their typical max pressure.
The alveoli are tiny thin air sacs in the lung that are in contact with blood vessels
to allow oxygen and carbon dioxide to diffuse between the blood and the lungs. To do this
they have to be extremely thin and because of that they are very delicate pieces of tissue.
Over expanding them will lead to inflammation at best or rupture at worst. This is what
barotrauma is. [5]
To make this worse, those suffering from acute respiratory distress syndrome, like those
affected by Covid-19, are more at risk of suffering from this side effect of mechanical
ventilation [6], as the alveoli that are filled with fluid prevent air from entering them,
causing the pressure to elevate even higher in the functioning alveoli. The last thing
we want to do is damage the healthy tissue of a patient suffering from damaged lung tissue.
That is the opposite of helping.
To avoid this doctors need to carefully choose their settings on a ventilator. The primary
guidance for this is to limit the volume and pressure of air entering the lungs. [7]
So, any low cost ventilator will need a method to control these settings. Designs like this
one, which can only vary it's volume output, as far as I can tell, by connecting the push
rod closer to the centre of rotation of the cam. There is no variable control here. This
device would likely do more harm than good. The designer's heart is ofcourse in the right
place, but if a YouTuber can spend a day reading a ventilator design book, so can they.
However, my patience goes to zero for massive multi-million dollar companies like Virgin
Orbit, who present these designs as their own. Who clearly have done zero research into
what is needed from a ventilator and just built something as quickly as possible to
get some positive PR for their company. Look, they even had time to put a big sign behind
them for the video.
As far as I can tell, the earliest design that proposed using these BVM was from an
MIT student project in 2010. This paper has been online that entire time and if people
are truly copying it, they are leaving out some clever design ideas that make it more
functional. [8]
Their design included a spirometer, which measures the air flow rate out of the BVM,
by integrating this value they can calculate the volume of air delivered.
This then feeds into a controller which can vary how tightly the BVM was squeezed to change
the volume of air delivered. This gave the device a nice range of tidal volumes ranging
from 200 milliliters to 750.
This is a better design, and may be useful in a do or die situation. But, it is not perfect.
The breaths per minute controller is simply set on a time based frequency, ranging from
5 to 30 breaths per minute.
This is called a mandatory breath. It's entirely determined by the machine. You will
take a breath whether you like it or not. This would obviously be uncomfortable and
requires the patient to be heavily sedated to the point of paralysis, but it can also
exacerbate barotrauma if the patient's diaphragm and intercostal muscles are resisting the
inhalation.
High performance ventilators can work like this, but they typically don't. Their breath
sequences are normally triggered by the patient. They are still able to breathe. They just
need help because they are exhausting themselves with the effort. In order to do this the machine
needs some way of triggering the breath cycle and ending it too, based on observations of
the patient. This can be done in a number of ways.
It can be pressure triggered, where a sensor detects a drop in airway pressure indicating
the thoracic cavity is expanding. It can be flow triggered, where a sensor detects airflow
into the lungs, or it can be triggered by a sensor detecting electrical activity of
the diaphragm, indicating that the diaphragm is contracting to expand the thoracic cavity.
This also requires very fast microprocessors to detect and react to the triggers.
I have seen no low cost ventilators incorporating this vital component of ventilator design.
And it truly is a vital component.
A very difficult part of the ventilation process is weaning people off it again. A ventilator
which requires someone to be sedated to the point of paralysis makes it very difficult
to get them breathing naturally on their own again.
There are a multitude of other design considerations to be made with ventilators.
I studied biomedical engineering, and there is a heavy emphasis on medical subjects like
“surgical practice”, many of which are taught by doctors, not engineers. The first
thing those doctors taught us is to start your design by speaking to the end user, the
doctors. I spoke with Rohin Francis, a doctor who runs the fantastic Midlife Crisis channel,
to get a better understanding of some of the other things we engineers need to remember
when designing these machines.
Thanks Brian. As you've already heard, COVID-19 patients frequently develop an acute respiratory
distress-like syndrome, or ARDS, which not only fills the alveoli with fluid, making
gas exchange harder, but also increases the likelihood of the alveoli collapsing shut
at the end of every breath out. This is because diseased areas of the lung don't produce
surfactant normally. Pulmonary surfactant is a clever substance produced by alveolar
cells which coats their inner surface and one of its key jobs is keeping these tiny
sacs open when the lungs are deflated, which is what happens in healthy lungs. But in ARDS,
when you breathe out, those alveoli collapse shut and sometimes whole sections of the lung
collapse, called atelectasis. Trying to force them open with every breath requires more
pressure and hugely increases the risk of barotrauma.
So, we use positive end-expiratory pressure or PEEP, to try to prevent this. I usually
explain this by saying imagine you've got your head out of the window of a fast-moving
car with your mouth open, don't do this by the way, in addition to all the insects,
you also have a constant air pressure exerted on your airway, making it ever so slightly
harder to breathe out. That's PEEP. PEEP is a constant positive pressure that prevents
those alveoli collapsing at the end of each breath and also helps open up - or recruit
- collapsed areas of the lung.
In COVID-19 we are seeing patients requiring very high levels and tight control of PEEP
to maintain their oxygen levels and protect the lungs and this is something that a basic
bag-squeezing vent cannot really achieve. From the mechanisms I've seen so far, I'd
be concerned about the possibility of baro and/or volu-trauma.
Most of these patients are on a ventilator for a few weeks at the moment. A basic bag-squeezer
might be adequate for the first day or so when a patient is deeply sedated, but simply
won't work as you try to ease off the sedatives. Additionally, your upper airways warm and
humidify air entering the lungs, but they are taken out of the equation by the endotracheal
tube which goes directly into the lower airways. Without the warming and humidifying features
of modern ventilators, lung tissue will get rapidly damaged.
This isn't even a ventilator, it's an anesthetic machine. Because the ventilators
I was going to film with are in use right now, and even this big thing can only provide
very basic ventilation.
And while we do need more ventilators, what's even more valuable are the intensive care
nurses and respiratory therapists to work them, but they take significantly longer to
produce.
So, as you can probably tell, there is a lot more to ventilation than just pumping air
into a patient. Tight regulation of Pressure, volume, , oxygen percentage control and humidification
would all require more complicated mechanics than these simply BVM pumps. Designing a ventilator
fit for purpose with cheap and easy to manufacture components is a difficult job, but I'm positive
a viable product will come to light soon. Especially as this is not a new problem. Poorer
countries have been struggling with the lack of cheaper medical supplies for years and
there are products like this one from an Indian robotics engineer, which uses an android phone
as the user interface. It was on the market long before this pandemic started, and was
designed to help poorer families treat their loved ones at home, so the need for trained