Pandemic Ventilator Project
panvent.blogspot.com
panvent.blogspot.com
This was the standard ventilator in Europe for decades. It is simple to design, does not even require electricity (mechanical solution driven by air from central compressed air system) so all parts can be sourced locally, and the design has a clinically proven track record.
Here is a video with schematic showing how it works:
Ask programmers, they’ll get you up to speed.
There's a difference between simple ventilator [1], and an anesthesia machine [2] that adds gas mixing, scavenging, etc.
ICU patients are anesthetized using IV sedation (a common regimen in the US is fentanyl/propofol), not inhalational anesthetics. Most vents only have simple inline filters to reduce contamination.
[1] A classic vent in the US is the Puritan Bennett 840. Here's its manual showing filters: https://www.medtronic.com/content/dam/covidien/library/us/en...
Seriously thinking about putting together a group of people to start producing these locally. Around campus we have a significant capacity in manufacturing precision mechanical devices.
I read this as "how to save a life with a vaccuum cleaner in extremis", and not "here's a good ventilator design". Is it a good idea? I dunno. I'll check on the vaccuum today just in case.
Also FWIW: the air path doesn't seem like the critical piece here. My (amateur) intuition tells me that the number of lives saved by amateur ventilators will be rather less than the number killed by amateur intubation devices.
Again: I'd be sort of OK with a ventilator manufactured by a HN reader. Under no circumstances do I want a software geek putting a tube in me.
I would consider the following a bare-minimum feature set for a COVID patient ventilator, any less and it would do more harm than good:
- cycle between an inspiratory and expiratory phase
- during the inspiratory phase, deliver an adjustable volume of gas (in the region of 6 ml per kg of patient's body weight) using as little pressure as possible, with an adjustable upper limit of pressure (in the region of 30 cmH2O)
- during the expiratory phase provide an adjustable pressure against exhalation (in the region of 0-30 cmH2O)
- allow blending of air and oxygen to deliver an adjustable inspired oxygen fraction
- allow the timing of the inspiratory and expiratory phases to be independently varied, thereby allowing the respiratory rate and the ratio of inspiration to expiration time to be controlled. Permit respiratory rates in the range of 8-60 breaths per minute
- measure and display the pressures and volumes within the respiratory system
- allow adjustable alarm-limits for pressures and volumes, and provide clear audible and visual alarms if these values are exceeded
For added patient safety and benefit, the following would be helpful
- measure inspired and expired oxygen and carbon dioxide content, and display both on a continuous waveform graph
- allow the patient to initiate the inspiratory phase by sensing patient inspiratory effort and providing pressure support for inhalation; ie sense when the patient inhales and deliver 10-15 cmH2O pressure for 0.5 seconds to augment inhalation
My mother is a retired RN with an "inactive" license in Washington state. She does elder care as a side gig for individual clients.
She recently had to turn someone down because the patient was intubated; it's not legal for her to care for such a patient. If anything went wrong she could be both at risk of a lawsuit and in trouble with the state for practicing medicine without a license.
(She's also in her seventies herself, putting her in the high-risk category, but that's a different issue.)
I'm at a loss as to how our legal and licensing institutions can be adapted to allow for a quick increase in the number of personnel who can care for intubated patients.
Where you, and fellow hackers can help, is by offering tech solutions that solve the issue of coordinating/figuring out where your mom’s skills could be most helpful. We will also need to collect information, such as when your mom is available to help. Could be a good time to revisit the concept of an Uber-like platform for healthcare delivery, assuming that it can be up an running in a few days. Hack-a-thon, anyone?
> revisit the concept of an Uber-like platform for healthcare delivery
Where has this been discussed before? There are lots of referral services, some of them charities, some of them private businesses. For example, in Washington state there's the Korean Women's Association: https://www.kwacares.org/
There's an immediate need for a single source of truth for all the locations where COVID-19 testing is available. Trump said Google was building it, but turns out they are weeks away from an MVP that would only cover the SF Bay Area. I propose we start by solving that problem.
Building on this, we can answer the question: what is the scope of the problem? Today, the CDC doesn't have an accurate count of the # of COVID-19 tests that have been performed because they don't have a way to collect that info from the growing number of labs that are performing the tests. Moreover, there's no way for anybody to know how many people have actually been tested, because the same person can be tested multiple times. We could solve this problem by crowdsourcing that information directly from individuals.
Building on this, we need to know where the sickest people are right now, and predict where they could be in the near future. This is crucial information for healthcare providers (like me) and public health officials alike. Without this information in hand, it's virtually impossible to know where to direct resources (like doctors, nurses, ventilators, medical supplies, testing equipment, etc.). We need to be able to track demand AND supply in real-time --> this is the point at which the project starts shaping itself into an Uber-like platform.
Building on this, we could enable individuals to post offers to help and for others to request assistance. This would allow people like your mom to lend a hand where it is a) safe for her and b) most needed.
The ultimate goal would be to build an open-source platform that could be used by communities around the world with little-to-no deployment overhead. Ideally, it would be possible to clone the repo, customize a few parameters, and deploy instances in less than an hour.
I'll flesh this out a bit more, and submit the idea as an Ask HN topic.
Seems like that could be achieved with a static HTML website, possibly augmented with a REST API.
I doubt that information is changing very quickly. It could theoretically even be committed to a Git repo in a CSV file. However, although that would work for generating a website, it wouldn't be clone-able. So perhaps consider some canonical datastore in the cloud somewhere.
Also: I propose Apache/MIT licensing to minimize the friction of collaboration with both commercial entities and charities.
Technology wise, I don't care: it should just be something mainstream, popular and easy.
> Moreover, there's no way for anybody to know how many people have actually been tested, because the same person can be tested multiple times.
That seems like a harder problem because of patient privacy issues.
https://news.ycombinator.com/item?id=22577611
There was a word limit, so I had to comment on my own post to share full details. Sorry, not sorry HN mods ;-)
Agreed, but even a simple honor system would be better than the current situation. Perhaps verified via a simple browser-based cookie?
Separately: just dawned on me that HN has no obvious way for users to communicate directly. Any suggestions on how to share contact info without begging for spam/bots?
It's focused on non-skilled volunteering, but would seem easy enough to add and match qualifiers (e.g. medical skillsets) between organizations and volunteers.
I.e., either the governor or the legislature says "it's no longer impermissible for RNs with inactive licenses to care for intubated patients". Or even "all inactive licenses shall be considered active on request of the licensee, until DD/MM/YYYY". Fiat, done.
And it would only be a start.
Can also do: if you're currently in a program of studies XYZ and > 75% complete, you can also practice until xx/yy/zzzz and indemnified by government.
The purpose of ventilation in these severe ARDS patients is to augment oxygenation of the blood, using a combination of techniques including end-expiatory positive pressure and inverse-ratio ventilation, among others, and support the fatiguing respiratory muscles. Because the lungs of COVID-19 patients are already injured by the infection, they are very prone to further ventilator-associated injury. Modern intensive care ventilators have complicated computer-controlled 'modes', which allow precise regulation of ventilatory volumes, pressures, rates, timing, and gas blending. The ARDSnet trials in the early 2000s demonstrated the importance of carefully managed 'lung protective' ventilation, poor quality ventilation is likely to cause further lung injury and make the patient worse, not better. Therefore amateur ventilators are unlikely to be beneficial in COVID-19 unless they can provide similar lung protective ventilation, which would make them quite complicated.
Additionally, ventilation is just one component of the management of sick COVID-19 patients. First you need to pass a breathing tube into their trachea, which is a challenging and risky procedure when performed by a skilled operative in a otherwise well patient, let alone someone on the brink of respiratory failure. Once intubated you will need to keep them deeply sedated, otherwise they will strain against the ventilator and make effective ventilatory care impossible. This requires equipment, drugs, and skilled staff, all of which are going to be in short supply.
Imagine a WWII style "war bonds" effort, but instead to build and distribute say 5M ventilators over the next 6 weeks.
In addition to this, we must re-purpose closed schools (gyms, cafeterias, etc) into make-shift CoVid19 triage centers where affected patients can be put on cots with saline drip, drugs, and ventilation. This is doable, but we need to do it.
Fund raising won't help if too many people get infected. What you need right now is for those in charge to stop playing Rambo and pretend they have it in control (they don't) then grow up, tell people to stop behaving like there is nothing going on. Schools, stadiums, all non vital businesses should close and people should stay home. This virus propagates at very high speed; as for now there is no cure, and I'm sorry for believers but prayers won't work (ironically, "healing" pools at Lourdes are already closed since days). This is serious, folks. We can only slow infections to a level in which people recover at faster pace than those who get infected, otherwise ICU numbers won't be enough and a lot more people will die.
I'm way out of my lane even commenting on this, but maybe you can dramatically simplify the air delivery side of this design, if you're willing to under-engineer. A single air chamber with an inlet (blower in) and two outlets (patient out and return out). Low rpm motor slow rotating a sealed disc with holes in it balancing outflow between feed-to-patient and return. Put a blowoff mechanism inline around the flow meter. If the capacity is there and there's such a thing as a common or global delivery 'frequency', maybe it could support more than one person.
Again, I might as well be from another planet in terms of knowledge in this domain. I'm just confused at the apparent complexity around the servo, linkage, and waste-gating. I DO get that delivery can be tailored nearly entirely in software with this design. That's cool.
Keep in mind, the use of these things is that you hook this up to a patient and then leave them in a room while you go handle other emergencies. It needs to tell you if it’s having a problem (through hookups to monitoring equipment.)
I think it also needs to not try to force air into a blocked airway (i.e. to blow open the patient’s larynx the moment they try to swallow), and/or needs to let up on the air pressure in a sinusoid pattern so the patient can exhale—though I might be wrong about either/both of those, given that they’re not really problems doctors encounter with hand-pumped ventilation. But either of those, in combination with the flywheel-like momentum of a big blower fan, would explain the waste gate.
https://news.ycombinator.com/item?id=22573926 "Low-cost portable ventilator (2010)"
https://news.ycombinator.com/item?id=22573656 "Low-cost ventilator wins Sloan health care prize (2019)"
The problem is that ventilators require tubes to be inserted into the patients lung. This requires a very skilled doctor and has a risk of infection and injury.
So I am making an wooden lung - like an iron lung but made from plywood. I can share the plans and sourcing for the parts if you want
https://en.wikipedia.org/wiki/Rotary_woofer
That lets you use both positive and negative pressure. You could even support fast pulses for clearing out gunk from the lungs.
https://hackaday.com/2020/03/12/ultimate-medical-hackathon-h...
I believe a ventilator uses a negative pressure or vacuum to assist with exhalation and air exchange.
There are now portable iron lungs that fit like a turtle shell. They do both inhalation and exhalation. That should be the only acceptable standard of care.
- NIV (Non invasive ventilation, CPAP is essentially a form of this) doesn't typically perform well (on a patient outcomes, mortality basis) for pneumonia with hypoxia compared with invasive ventilation. However, this is thought to partly be because NIV delays the decision to proceed with intubation and ventilation. If there is no ventilator available, that might change the value of NIV.
- NIV will also likely cause aerosolisation of the virus facilitating spread if there are others in the area. Most sleep apnoea CPAP masks are vented which would probably make this worse (cf unvented masks commonly used on ventilators in an ICU setting)
- Many sleep apnoea CPAP machines don't allow entrainment of supplemental oxygen, which would be likely to be needed in critical COVID-19 infection
Source: I'm an intensive care specialist
Edited to add: Spoken as the Dad of a 28 week premie who spent 8 weeks on a ventilator in a NICU.
Since you are a professional in the field, I am asking your help for brainstorming useful, contributory things for people like me (and others) who are feeling the need to DO something.
If hacking up ventilators (like in the OP) is not a great idea from your perspective, can you suggest anything else?
And, THANK YOU for doing the job you do, from someone whose life was affected by people like you!
Once that is done...
If your Ph.D. was in the life sciences, or you can find others with that expertise, I'm curiously exploring a tech tree and precursors breakdown of what I need to actually carry out my own Covid-19/RT-PCR testing, notwithstanding the significant hurdles to do so pointed out by others [1]. That led me to a fellow HN'r pointing out the OpenCovid19 group [2], which has tons of links to other similar hacking initiatives.
It looks like some baseline logistical equipment is required, medical grade freezer, refrigerator, autoclave and centrifuge, along with some glassware. Someone did some work on that [3] earlier, at the moment I'm scrounging my local area to see what pre-owned gear I can pick up for myself (I've always wanted a bio-hacking section in my workshop), and comparing against what I think it would take for example, to fabricate centrifuge parts with a lathe, drill press, soldering iron, and scrounged second-hand shop items like a blender motor, for underserved communities like in the poor sections of the US and in the Third World.
Based upon what I can understand from the CDC testing protocol [4] for example, I need molecular grade water as a precursor. In my position in the US, I'm fortunate enough that I can simply purchase this. But were I to create a procedure for people who have more smarts and time than money, apparently making your own molecular grade water is quite the undertaking, especially if you undertake making your own DEPC. If my tech tree starts at "common 21st century household items, urban community scrounged second-hand equipment, access to trades or Maker shop, and Internet access with modest ($1000 USD) funding for online ordering", even getting the logistical precursors in place looks pretty daunting. The more money I allow into the picture, the easier it gets.
I'm currently researching what logistical train I need after I get a Bento Lab and primers (not necessarily Covid-19, likely not enough time to do that) to conduct my own RT-PCR tests.
I'm fully aware it sounds like I'm asking the equivalent of "teach me to program in 30 days". I have patience, I understand that what I do now likely won't pay off to help with Covid-19. What I really want is the pointers to resources that lets me read and understand how to breakdown the CDC protocol by tech and procedure layer. It sounds like RT-PCR testing could be a really useful general technology to have on hand in the future. So I want to put together the write up on say Hackaday that could drive this kind of conversation.
"First you need a centrifuge, of this minimum rpm, this many slots per patient, etc."
"Oh, you have to source a second-hand one? Link here to a list of characteristics to look for to find a good one. This is a list of good brands."
"Oh, you can't source pre-owned ones? Here is a way to use common machine shop tooling to fab one, link here to the kind of electric motor you want to find, here is how to test the motor."
This has piqued my curiosity, I'm feeding it to see where it leads me.
[1] https://news.ycombinator.com/item?id=22570801
[2] https://app.jogl.io/project/118#about
[3] https://makezine.com/2017/04/11/how-to-set-up-your-own-lab/
Not sexy. May be irrelevant in 3 weeks at crisis standards. Right now with the supply chain as is and attempting to surge standard of care capacity this is the thing that would be useful. If we cannot manage ppe lack of providers will make lack of anything else moot.
https://www.cdc.gov/niosh/npptl/topics/respirators/disp_part...
As an Earth scientist, I could imagine a mineral based replacement for the filtering part -- e.g. something like a zeolite -- but the lack of mechanical flexibility and the risk of silicosis are two serious problems with that right off the top of my head.
How about 3D printing of filters? Can N-95 -- 95% of particles greater than 0.3 microns -- be achieved?
Edited to add: It appears somebody is doing this for liquid filtration: https://liquico.com/products/
https://www.theguardian.com/society/2020/feb/27/coronavirus-...
No such austerity has happened, and now is not the time for this petty bickering.
Denying the fact that the NH is under pressure isn’t helpful either... we need to be realistic about what to expect when we underfund our healthcare system.
That seems low relative to US medical costs in general. I'm sure the average Bay Area software engineer would happily pay that much to save a dying family member.
I've seen no one is suggesting that these DIY/hacking efforts of all stripes replace the existing healthcare infrastructure. What I see is concern the existing infrastructure is not scaled to handle what we anticipate will happen even in a moderate scenario. We have in this very thread a real MD who confirms that post-ICU ventilation is not being addressed but still needs to be, yet that confirmation seems to fly by those who say that DIY isn't advisable.
My position on the mitigation however, starts from these two premises:
In Extremis, any action is better than no action at all.
In Extremis, don't let perfect be the enemy of good.
When asked, I believe many will opt for a 1% chance of living on a DIY ventilator intubated by recently-out-of-license nurses emergency-authorized during the pandemic when no other intubation staff are available in the time required to attempt saving a life, than a 100% chance of dying with no ventilation option whatsoever.
What I'm not seeing are any alternative solutions offered by those who want to stick to the "official" script. Let's in about 6-8 weeks from now our healthcare infrastructure is overwhelmed by severe respiratory distress cases, starting in the June timeframe. What do you suggest we do to prepare for that starting now instead of any of the DIY efforts you deem too risky/non-compliant?
I'm honestly asking in response your specific post, now. Even an answer like "triage non-comorbid over comorbid patients, let the comorbid patients die" is acceptable. Even if your response is, "I'm no domain expert, I'm leaving it all up to the experts", that works for me, too. I'm just trying to figure out what your mitigation angle is.
What doesn't help improve our probabilities nor mesh with my personal philosophy is giving criticisms with either no alternatives or no position statement of what you believe is an acceptable protocol going forward. It's in my nature to seek answers and address my curiosity, even if it only leads to a partial answer. If that's not you, then you live your best life, and thanks for the note of caution, it is definitely a concern to manage as best as we can.
What is the most promising home made gear I might be able to scrap together in the worst case scenario?
More recently, many have a large oxygen concentrator machine on site which just takes it from the air.
It would be noisy but would allow the design to incorporate negative pressure of the vacuum side to assist with exhaling.
Maybe the shop vac could be plugged into a speed controller to improve duty cycle and reduce noise but controlling an AC motors speed is quite a bit more complicated(expensive) than controlling a DC one.
> controlling an AC motors speed is quite a bit more complicated(expensive) than controlling a DC one.
That type of motor is so noisy because it's a universal motor[1] - they can be controlled with light dimmer type circuit[2], but they're basically always noisy and inefficient.
A cordless drill would be a good starting point for a variable speed power train.
[1] https://en.wikipedia.org/wiki/Universal_motor
[2] If you're actually using a light dimmer to control a universal motor, use the highest current rated dimmer possible. The motor will be more inductive than the typical lighting circuit that the dimmer is designed for.
From https://app.jogl.io/project/121#about
>> Current Status of the project
>> The main bottleneck currently (2020-03-13) is organization / management.
>> […] This is an organization of experts and hobbyists from around the globe.
(edit) here's the link to their 'Ventilator' document: https://docs.google.com/document/d/1RDihfZIOEYs60kPEIVDe7gms...
the air version of that https://www.youtube.com/watch?v=d1RoEgxax60
If a doubling takes 7 days, that’s just 2 weeks.
R0 (reproduction rate) overwhelms everything. The only way out safely is to greatly reduce R0:
- social distancing: 6-9 feet away from everyone
- avoiding crowds
- Never share utensils; no buffets
- hygienic routines: wash hands, use wipes, close toilet lid before flushing (see Amoy Gardens), etc.
And convince others to do the same.
Italy closed schools when they had just 150 confirmed cases but still in a tragic crisis because people kept socializing.
See a timeline & analysis here: https://www.luca-dellanna.com/how-bad/
If anyone's interested, please drop me an email: rls at hwyl.org
If you have an idea to share, share it. Don't make people jump through the hoop of sending email.
If you are concerned about others stealing your idea, then your comment is essentially spam, because you're using HN as a recruiting tool to build your team. Only YC companies can do that here.
Your motivation might be something other than either of these possibilities. It's easier to downvote and stop thinking about it than to figure out your true intentions.
Reposted:
Could a voluntary user-tracking app help reduce the R0 level of the COVID-19 virus?
One of the main challenges with this virus is that carriers of the virus become infectious well before they show symptoms. Consequently, they pass on the infection before they can be isolated.
Apps such as Google Maps already track users' movements to a high level of precision. A similar app could be created that tracked users' movements and notified people they had interacted with if they got infected.
It would work like this:
- Users' movements are tracked
- Upon developing a fever (etc.) the user notifies the app
- The app checks their movements during the estimated period they could have been infectious
- Every other user who was in their presence for X minutes gets notified and requested to isolate themselves
- Recursively check those users' movements to see if they may have passed on the virus
- Requires simultaneous voluntary use by a large proportion of a regional population
- Possibility of false positives by bad actors (would an occasional false positive be so bad)?
I see this same or similar idea has been written up as a paper (which I've just submitted as a separate story): https://www.medrxiv.org/content/10.1101/2020.03.08.20032946v...