Italian hospital saves Covid-19 patients by 3D printing valves
3dprintingmedia.network
3dprintingmedia.network
Why would this be trickier than having 1 clean valve, 0 dirty valves, and 2 patients?
"Just received word from an ICU doctor at a small NY hospital: They are officially out of ventilators and are now double venting patients with COVID (using the same ventilator for 2 infected patients)."
You would probably want to use a "high-level disinfectant" or "sterilant" instead. Gluteraldyhde is fairly common choice or ethylene oxide gas, though both can be fairly nasty if not handled carefully (which is, of course, the whole point).
The way alcohol swaps are used, they move contaminants away from the needle insertion site, and the alcohol helps to dislodge them. And, of course, it does kill some kinds of pathogen, as does soap.
One doctor can save numerous lives. Putting one out of commission is a huge blow in a crisis.
Maybe it's a labor/scale issue?
You just need to be careless once for it to be a problem. And considering how many sterile items you need to use every day - particularly in an emergency - it is not unlikely that one item went back into the reuse cycle without being sterilised. It's to avoid the question, 'was this item sterilised?' If its wrapper is open, assume it is used. You can only do that safely with single-use items.
How about putting the re-usable item in a wrapper so that the same rule can apply.
I have seen many different techniques for this kind of workflow.
You probably need to autoclave the parts, and then some more to be sure.
Simple boiling won't do much even for viruses
https://www.industrialspec.com/resources/plastics-sterilizat...
So I wonder if this table might have some significant errors in it? Maybe someone interchanged some plastics without intending to. It can't possibly be correct to put PTFE and PCL in the same category, can it?
Anyway, common plastics it rates as "Good" for autoclaving include acetal (Delrin), polypropylene, silicones, and (for some definitions of "common") PEEK.
Surprisingly polyimide (Kapton), the favorite high-temperature chemically-inert plastic of aerospace and electronics, isn't in the table at all.
I'm not surprised that some plastics may not be based on temperature only. For instance, nylon might crack if it's a thin piece due to thermal expansion. Or some may chemically react with steam.
I agree that it's weird that PCL would be okay while Teflon isn't, but it could be a molecular weight or thermal cycling thing. Teflon is very soft already so it might simply get deformed.
I'd trust such a table to point me in the right direction but find a second independent source if the answer seems weird.
Of course in practice you just try it and see if it survives and is functional. lots of plastics can survive a cycle or three but not five.
Things like needle drivers and forceps hold up fine with some care, but things with delicate moving parts are not supposed to be autoclaved. Plastics depend on the material--some are fine, some deform.
You can pasteurize eggs at ~135˚F and they're practically uncooked. The autoclave isn't like that at all--15-20 min at full temp/pressure plus the rampup/down times can do a number on things.
Got one word for you: prions. Standard sanitization procedures do not work with them, and if the hospital is not setup to decontaminate for prions, they're unlikely to be able to to it during the crisis.
see https://www.memphis.edu/ehs/pdfs/deconprions.pdf for what it takes.
All hospitals should have autoclaves, shouldn't they?
In terms of overall mass, sterilizing all those parts should be equivalent to throwing them away. It would require different internal procedures though, I don't think people are used to saving and reusing things. And it would be more labor intensive. And autoclaving might destroy the parts.
Most of which seem like reasonable tradeoffs in an emergency.
[0] Kuru is not something you want to research whilst eating, fyi. Hell, none of them are.
https://www.thelancet.com/journals/laninf/article/PIIS1473-3...
I recommend reading "Technical Considerations for Additive Manufactured Medical Devices". Then start discussions in your professional societies to make contact with the FDA as a group and start to develop guidance targeted for the COVID-19 emergency.
[1]: https://www.fda.gov/medical-devices/products-and-medical-pro...
The problem here, and I think this will be much worse in the weeks to come in the US and the UK, is the institutional decline of manufacturing capacity and the reliance on long supply chains from countries which are themselves at max capacity.
Every country should be able to, in crisis, quickly switch to mass production of vital goods that guarantee the physical safety of their citizens. Globalisation and post-industrial societies might have created nice margins for some subset of the population for a while but I doubt the corona virus will be swayed by management consultants.
The entire point of doing research - devising a theory, doing an experiment, assessing the results, revising the theory - is to find knowledge that could be benefit humanity.
However, research itself NEVER guarantees up front if a theory, an idea, an experiment,... is going to yield a beneficial outcome for humanity.
It's literally: you can't say "well, this experiment is going to yield a useful vaccin, medicine, tool,..." because the very definition of experimentation and research is trial and error.
Maybe it will take a 2 billion dollars to find a vaccin, maybe it will take 20 billion dollars. Maybe it will take 3 months, maybe it will take 3 years. It's literally impossible to predict up front.
This is not just true for medicine, this is true for any and all research that takes place at Universities.
So, if you're going to scale the endowments to academic research based on the "economic productive results" their research yields, well, you're going to miss a massive amount of opportunities.
Moreover, academic researchers don't just get bags of money so they can do whatever they fancy. That's not how this works.
The funding of the vast majority of research projects requires the application of grant proposals which are then reviewed and vetted by peers and governance boards. When the results are published, the project also have to yield a balance sheet and stringent reporting on how they spend those grants.
The vast majority of research proposals are thoroughly scrutinized. You can't just whip up a proposal and a fancy experiment and think you'll get funding. Academia is a cut-throat business if you want to do any meaningful research. Ask any researcher and they will tell you that they agonize whether they'll get the funding they need to continue what they are doing next year.
Academia is a hard place to work exactly because that's how the best ideas float to the top.
As Edison famously said about inventing the electric lightbulb:
> "I speak without exaggeration when I say that I have constructed three thousand different theories in connection with the electric light, each one of them reasonable and apparently to be true. Yet only in two cases did my experiments prove the truth of my theory. My chief difficulty, as perhaps you know, was in constructing the carbon filament, the incandescence of which is the source of the light."
Exactly. Think about the Wright brothers and flight. They didn't simply set out one day, build a plane, took the hills and flew the kite. It took them many failed attempts and a big risky investment in material and time before they were even able to take off. At the time, they didn't know if they'd succeed.
So. No. "Productive" is the wrong bar to assess the value of research by.
The question is whether it is productive to maintain large investment funds unrelated to their stated goals of education and research.
I think the OP would be happier if those endowments were drawn down on research with uncertain benefits rather than just invested in natural resource extraction companies.
>The discussion isn't about university research money and whether that is spent productively, it's about endowments, which are nothing more than investment funds.
>The question is whether it is productive to maintain large investment funds unrelated to their stated goals of education and research.
>I think the OP would be happier if those endowments were drawn down on research with uncertain benefits rather than just invested in natural resource extraction companies.
I am in favor of big research budgets, if real and valuable research is being done. I know there are a lot of uncertainties there and that there isn't always a direct 1-to-1 correlation between spending and benefit for the economy/humanity.
That is still positive sum. But it's something different. I'm referring to zero sum hoarding, not research funds.
These pieces might take several hours to print which means one 3d printer could produce maybe 50 of these a week. If you have 10 printers you could do 500 in that week.
However, within one working week I can have injection dies machined - and with a small investment ($10k ish) I can buy an injection moulding machine and pump these parts out at a rate of 100s per hour and a cost of almost nothing.
3d printing has an extremely small window where they are useful for bulk use parts like this.
its a kickstarter, and undercapitalized but please don't over-state the "ease" with which injection moulding can work. Its true but, it has a high capital burden up-front to "get it right" which clearly, this current medical emergency demands and we should expect.
No, we can't just mint things with zero cost, injection moulding. It requires craftspeople to make the moulds.
(No offense meant but there has to be a reason, though I'm not familiar with the space)
Source: worked at a place that made injection moulded hardware.
A large amount of the 1960s dislike of plastics vests in the super-bad job injection moulding of the time did, with both bad plastic stock, and badly made moulds. Now, we think its fantastic but we're standing on the shoulders of giants who made it work. Lego's fab is pretty awesome, to maintain dimensional fit for click-fit on millions of parts like that.
The Mint, is not exactly injection moulding, but similar: they have to replace coin dies frighteningly often because they just wear out. The CSIRO did work with the Australian mint doing ion beam vapour deposition to create hardened surfaces to try and preserve the die lifetimes.
Stamping out CDs speaks to how accurately we can do this now.
(injection moulding CDs is not a thing)
You can have a service bureau use a $250,000 nylon 3D printer to get small to medium-sized runs for something more affordable than making a mold.
The platform is of course used for many different sorts of project.
Really depends on how committed you are to bootstrapping and saving nickels.
With SLS this is however realistic. Maybe two hours for a batch.
If you can have enough printing capacity on site (per hospital, per unit) to meet demand (keep you X number of ventilators running indefinitely) it might make sense.
Still having 1000s of spares in a closet might be safer than relying on a 3d printer bein online some n percent of the time.
There's a limited amount of space and logistics to have 1000s of spares of everything you might possibly need.
For US$10k you can buy 28 brand-name Prusa Minis or 50 off-brand knockoffs on DealeXtreme. If a part takes three hours to print on one of these FDM printers, as you're suggesting, you could print 2500 of them in a week with your 50 off-brand knockoffs; if you can get the printing time down to an hour, 8400. And, unlike with an injection-molding machine, they come out of the machine sterile, and PLA is biocompatible and aquadegradable, implants having been one of its few significant uses prior to its uptake for FDM.
Such FDM mass-production is in fact how Prusa produces a lot of the parts for its printers. Aside from sterility, it has some other significant advantages over injection molding: its dimensional tolerances, shitty though they are, remain stable over the entire print volume, so you can get a 200±0.1mm dimension on your part on a typical cheap FDM printer; and the parts you print are much ligher in weight, due to the standard honeycomb infill, and (though this is not relevant in the case of duplicating an injection-molded part) they have much more geometric flexibility than injection molding — they can include voids and enclosed spaces, and even preassembled mechanical assemblies.
The article says that after the initial FDM parts, they switched over to SLS, which is a faster and more precise 3-D printing process which can also process metals, although in this case they were doing plastic SLS. SLA is another higher-speed, higher-precision 3-D printing process, and both SLA and SLS can produce several parts at once for a further gain in speed.
On the other side, I hear it's common to have to go through three or four revs of injection-molding dies before you get to the rev that works properly, and it's common for that to take months. Maybe that's a problem that can be avoided by sacrificing something else, like cycle time or die durability; do you have experience in the area?
So I think we're going to see a big move toward 3-D printed parts in the next decade or so as designers wake up to the new possibilities afforded by these processes. A few years ago the crossover where injection molding was cheaper might have been 100 parts, then 1000, and now it's 10000 — and that's when you can make the part at all by injection molding, with the extra weight, imprecision, and geometrical restrictions it adds. Maybe in another few years the crossover point will move to 100,000; but I fully expect to see designers trained in the 1980s and 1990s continuing to use injection molding even where 3-D printing would be a lot cheaper.
What do you think about the questions I asked, about why some injection dies take multiple back-and-forth cycles over several months to get to the production rev instead of the one week you cited, and how that can be avoided?
Please note that I am ONLY talking about plastic parts. Every piece of electronics has many injection moulded parts.
Show me one item in your house with a 3d printed plastic part?
Sorry but you're just wrong about the future of plastic 3d printing machines. There's zero chance it will ever beat a cheap injection moulding machine pumping out 1000s of parts per hour.
3d printing of metal is completely different. That could be a game changer because we do not have good methods to mass produce complex metal parts.
3d printing of plastic is pointless as soon as you need more than 100 parts or so.
Also note that your 'sterility' claim is wrong. Injection moulded parts are every bit as sterile as 3d printed parts. PLA can be used for injection moulding too, no problem at all.
The owner of a Texas business ramped up mask production for SARS and then almost went bankrupt when demand collapsed afterwards. He simply will not ramp up this time without money up front.
It's not about manufacturing. Capitalism doesn't always work.
This is one of those instances where the government needs to step in and store inventory.
Very interesting read.
The Strategic National Stockpile does store an inventory of masks. 13,000,000 of them. https://www.statnews.com/2020/03/10/coronavirus-strategic-na... According to https://www.washingtonpost.com/health/2020/03/05/washington-... they have plans to get another half billion.
This sort of preparation looks useful, but I don't have the greatest hope that it will really help us deal with the next outbreak - I'm sure we had large stockpiles of equipment after the Spanish Flu too. It'll likely be down to 13 million again by the next pandemic.
A bunch of anti-science knuckleheads decided to eliminate it.
It's no good railing against the specific group, because there'll always be another. The problem which feels like it needs addressing is the fact that it's possible for these groups to interrupt the solution.
That's like 1/10th of 1 day's needs...
>they have plans to get another half billion
Each week until the virus is contained, I hope.
Reserving the masks for CDC, vaccine researchers, and national defense operators would be an example of good uses of a nation-focused 'national' reserve.
Think 'trolley problem' where one side of the track is the nation-entity that contains the citizens on the other side of the track.
you can probably help your country and save lives at minor expected cost to yourself by just giving this guy a legally binding promise to pay him for unwanted masks should he ramp up production now and no one wants to actually buy them.
Basically offer him a put option for free, it's unlikely he'll need to make use of it.
If it was in the united states the patients would just die while the administration insisted on waiting for the most profitable bids from various companies while the president did a press briefing to shake the hands of the various ceos and boast about raising their stock prices.
Actually we may just get that scenario since we are on the Italian curve and so many people right now are out in bars, restaurants, etc.
This is going to be a year to remember.
When the SHTF, we individuals, clinics, and local areas won't give a shit what DC and their lackeys say or do. We're on our own, in a local collective way.
If the 3d printed part is good for 1 person, then I'll print 10. And my buddy will print 10, and they'll print 10.
In practice, when SHTF ie in a situation like 9/11, most humans beings seem to do the best they can with what they have.
https://patents.google.com/patent/US3882860A/en
Just something I found on google, not sure if this is 100 percent accurate, but it sounds like a manual, limited use heart/lung machine.
Was very disappointed when the reanimation class wasn't taught by Dr. Frankenstein and only covered CPR.
The people printing the valve have already technically committed copyright infringement.
Sharing the model with other hospitals that need to get them printed might be still borderline but still (I hope) tolerated.
But going further and sharing the model to the public could degenerate into a serious lawsuit.
https://en.wikipedia.org/wiki/List_of_national_legal_systems
Probably an equally good valve design whose patent has expired could be used instead. It is not uncommon to patent a new design just different enough, but not actually better, to avoid competition.
https://translate.google.com/#view=home&op=translate&sl=auto...
https://translate.google.com/#view=home&op=translate&sl=auto...
So in this case "reanimation device" is just "a device used in ICU".
Just another case of NIH!
A few studies suggest it could be done: https://www.ncbi.nlm.nih.gov/pubmed/18164798 https://www.ncbi.nlm.nih.gov/pubmed/16885402
Could be faster than building new ventilators.
Too much volume (or pressure, depending on the control mode) can injure the lungs, too little and you're not addressing the underlying problem very well.
In this case IP might mean Intolerable Pricks.
https://www.theverge.com/2020/3/17/21184308/coronavirus-ital...
I have a large format 3d printer (.5m x .5m x .5m) and am in Indiana.
The parts I print on my MakerBot are very porous...
> Article I, section 8, which reads "Congress shall have power ... to promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries."
It would certainly be interesting to see "promotion of the progress of science and useful arts" to be pitted against "saving lives during a national emergency" in court, where the said company refuses to provide needed replacement parts.
Something like that surely must have happened already?
With the 2001 anthrax attacks, the US government ordered Bayer to lower the cost of an antibiotic that treats anthrax, or lose their patent on the drug. (Bayer complied.)
Source: https://is.gd/PodKth
In the hospital they said: "we don't care, we need the valve". Looks like 10 of them are currently in use on patients after some tests.
This is insane.
> She explained that the hospital in Brescia (near one of the hardest-hit regions for coronavirus infections) urgently needed valves (in the photo) for an intensive care device and that the supplier could not provide them in a short time.
It sounds like a good idea to give people a choice of nothing or something that works 98% of the time. How do you pre-plan for this and accept 98%? I have had this discussion with medical device QA people and they could not acknowledge such a choice like this could ever happen.
Scenario 1: No treatment option and you will die.
or
Scenario 2: Choose between the following options
A) No treatment and you will die, OR
B) Treatment with lower-quality QA standards (with a device that works, say, 98% of the time instead of the typical 99.9% of the time).
Most people will prefer the scenario that gives them the choice.
Note: these numbers made up purely for the hypothetical.
Just follow the Socratic oath and don't over-complicate things. Give the necessary treatments.
If there are two equal choices with cost and benefits, then maybe.
I wonder if there ought to be a 3D printed copyright license scheme kind of like the one that exists for music (APRA here in .au) where you can pay for a blanket licence to print copyrighted designs, and the licensing organisation redistributes copyright payments based on usage, so you don't need to ask for specific permission for each use?
https://www.techdirt.com/articles/20200317/04381644114/volun...
They sell them for $11,000 each, and threatened to sue the guy 3d printing them for a dollar a piece...
:sigh:
I think there's something 100% valid in what you're suggesting, but it seems there's a fair bit of confusion here.
"Fair use" is a copyright thing, not a patent thing. I don't actually know if a widget (whether patented, patent expired, or never-patented) is covered by copyright - and even if it is, at least from what I know about copyright in music/literature, "fair use" wouldn't allow for making 100% complete working copies.
"not for profit" doesn't fly as a defence against either copyright or patent infringement in other areas, I suspect it wouldn't here either.
I'm reasonably sure if the "owner" of these valves started making noises about overly onerous enforcement of any patent/copyright that applies, and sane government would step in and declare some emergency-use legal exception for cases like this.
I wonder if the "owner" could make some obvious olive-branch offer along the lines of:
"Hey, that's our widget you're printing, but we totally understand that for the duration we were unable to get our version into your hands for reasons outside both our and your control. Normally we earn $x per widget when you source them through the usual channels, we are happy to waive that for the duration of the supply chain disruption, but ask that once things are back to normal, you continue to support the research and development that goes into those and other life saving products we make, and start buying from the usual supply chain, or paying our regular royalty on locally printed versions. Thanks, WidgetCompany"
Yeah. About that...
https://www.techdirt.com/articles/20200317/04381644114/volun...
Counterfeiting in clothes is one thing, counterfeiting in hospital matherial, when even perfect products can be accidentally contaminated with biological matherial and create lots of problems (or even kill people)... Well, this plays in a very different league.
And cloning devices of course opens the door for other companies to crush the original maker flooding the market with cheaper products exactly when the maker rely on benefits to recover their money (after mass producing products). I assume [1] that there are laws also against this.
[1] I'm not a lawyer and could be wrong. Your mileage may vary.
That said, there is no reason to assume a product is covered by a patent unless it's marked with the specific patent number (a patent marking on it's packaging counts too). It might be, but most products aren't.
If you are an unknowing infringer (didn't know and had no reason to believe you infringed), your damages are limited (in the USA) to a "reasonable royalty." Basically, what a the parties would have hypothetically agreed to in a negotiation. That might be a few % of revenue or a small lump sum.
I do wonder what will happen down the track, when the manufacturers/copyright-owner/patent-owner or whatever starts asking about whether or how they should have been paid their cut on these.
If you want a monopoly on copying a machine part, you need to apply for a patent, not a copyright. Among other differences, the patent only lasts 20 years, and doesn't exist automatically but must be granted explicitly after a long litigation process — the vast majority of machine parts are not patented.