Hacker News new | past | comments | ask | show | jobs | submit login
Italian hospital saves Covid-19 patients by 3D printing valves (3dprintingmedia.network)
622 points by csnover on March 16, 2020 | hide | past | favorite | 195 comments

Seems like this is a downside of single-use sterile medical parts. In most cases its much easier and safer to just grab a new single use part but in times of shortage you could sanitize and sterilize multiple use metal parts in house and get them back in use without needing to physically manufacturer more locally, which often isn't practical - even with additive manufacturing.

>in times of shortage you could sanitize and sterilize multiple use metal parts in house

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.

That is a valid argument for neurological equipment, which has the highest liklihood of infection for prions, but during this pandemic, the likelihood of prion transmission is very remote, probably to the point that it approaches zero, and is probably zero for these valves.

Not all prions need to break skin to be an issue. Kuru, one of the more infamous prions, was transmitted orally [0]. Mad Cow is also thought to be transmitted orally, but research is on going. Though a low possibility, the absolute last thing you need is an even more complicated issue running about during a world-wide triage/crisis. To save live you are trying to eliminate variables, not add to them.

[0] Kuru is not something you want to research whilst eating, fyi. Hell, none of them are.

transmitted orally from eating neural tissues not from prions in saliva that could contaminate medical equipment such as ventilators (that's my understanding at least)

Mad Cow is still active in it's research. We cannot yet rule out that salivary contact is a transmitter of prions, though I agree that it is very unlikely.


>Autoclave at 132 C for 4.5 hours

All hospitals should have autoclaves, shouldn't they?

Every biology adjacent facility I've ever heard of requires them for at a minimum biohazard waste disposal. I don't know how a hospital could not have enough given that all those plastic one time use parts still get autoclaved after disposal.

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.

I'll add Candida Auris to the list of scary things that are hard to kill in hospitals: https://en.wikipedia.org/wiki/Candida_auris

It's not to protect the patient (well also that), it's to protect the doctors. My granddad was a medical student back when they still reused syringes. That gave him a liver infection, so he could not donate blood his entire life.

One doctor can save numerous lives. Putting one out of commission is a huge blow in a crisis.

I agree with the values you're describing here, but it's still not clear to me that a re-use workflow needs to be that different from single-use. Presumably the single-use involves tossing used pieces into a biohazard disposal container to be removed later. What's the issue with tossing the same material into a biohazard container of things to be solution-rinsed autoclaved and/or UVd the hell out of, then maybe wrapped or banded so that it's clear they're clean?

Maybe it's a labor/scale issue?

Labour/scale, indeed. The facilities where these things are usually produced have the capacity to make sure they are sterile. Hospitals cannot sterilise items to the same scale.

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.

Except that sterilised items usually are repackaged in some way that removes this doubt.

I have seen many different techniques for this kind of workflow.

> 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.

All surgical steel is reused exactly this way. There are plenty of plastics that get autoclaved as well.

Yeah. For example, thinking about needles and syringes, they used to be disinfected by boiling water. That's certainly not enough to kill all existing pathogens especially after you used it in a person.

You probably need to autoclave the parts, and then some more to be sure.

This would be ok some decades ago. Now we know that prions can survive

Autoclaving is hard but it seems to be doable for Prions https://consteril.com/prion-sterilization-guide/

Simple boiling won't do much even for viruses

Depends on the nature of the contamination. Some pathogens are exceptionally difficult to de-contaminate, like Candida auris.

There aren't many things that survive high pressure autoclaving. Prions are the only possible concern that I'm aware of.

The cost and complexity of making parts that can survive autoclave is a big driver of single use parts.

Asteroid organic sample return missions actually face some of the same problems. Even moderately complex sample retrieve mechanisms can't be autoclaved. That's why so many are using passive sample collection bins and compressed gas.

From what I understand, there are some components that can't be put in an autoclave and/or have areas that aren't well suited to autoclaving. So it may not be a one-size fits all solution.

Good point

The defense in depth mindset response would be single use where it is superior, but with emergency protocols for reuse in emergencies where perfect becomes the enemy of good. E.g. rate disposable items for heat treatability even if nobody will try to disinfect them outside of an emergency shortage.

You could alternatively just have a couple reusable items on hand per machine instead of making every single disposable one reusable.

Yeah, that would be my recommendation. Make a supply of parts that can take the autoclave and keep them around for emergencies. If you had to use them because you messed up the supply chain (didn't order enough), big fine. If you ordered but the supplier couldn't deliver, no fine.

Even better! Seal them in a box labeled "break glass in emergency" if capitalism is the main objection.

This article could make one more concerned about that problem just by making them aware of it. But if you were previously concerned about it, this news should reassure you, since now you know that there is a solution available even in crisis conditions.

Only for parts amenable to 3D printing, which is limited at the current state of technology.

Yes the issue is the stability of the plastic at the temperature and pressures that sterilisation needs. I know that plastics have been developed for exactly this purpose but I don’t think they’re very widespread yet.

It appears from Wikipedia that normally you sterilize things at 121° in steam for 15–20 minutes? If that's the case, common plastics like nylon and polydimethylsiloxane should be okay, although poly(lactic acid), commonly used in FDM 3-D printing, and poly(ethylene terephthalate) won't be.

A lot of plastics will deform when autoclaving. There are enough lists for autoclave compatibility that I forget the answers, but this looks like a pretty complete chart.


This is a really comprehensive table! It lists PLA as "Poly(L-lactide)", giving it a surprisingly good "Fair" grade for autoclaving, and polydimethylsiloxane as "Silicones", rating them "Good". It also lists polycaprolactone, which melts around 70° and can be easily molded by hand while molten, as "Fair". Surprisingly, though (to me), it lists nylon (6 and 6-6) and teflon (polytetrafluoroethylene) as only "Fair", like PLA and PCL, even though at least teflon is totally insensitive to water and commonly used with service temperatures up to 260° (among other things, in FDM printer hotends).

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.

It is totally possible that different companies will have slightly different charts but usually they're about the same. The canonical way to know if to ask the manufacturer of the actual plastic you're using, these comprehensive tables are usually put out by distributors as general guidance.

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.

Note that it's 121˚C (and under pressure), not a tepid water bath.

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.

Right, I didn't mean 121° Frankenstein, and I see now that my comment could be misinterpreted as meaning that all common plastics would be fine, when actually I only meant nylon, polydimethylsiloxane, and a few others — and even for nylon I should have been more specific.

I was just trying to emphasize that it's a fairly...aggressive process.

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.

I appreciate you having clarified my comment with your emphasis! It was definitely needed.

Nylon isn't generally known to be autoclavable (at least I avoid it when I want to sterilize parts in house (R&D at a medical device company).

Thank you! The plastics compatibility chart linked above also listed it (even nylon 6-6, which it called nylon 66) as only "fair", which surprised me. Does it hydrolyze?

I have no idea what happens, maybe cracking or deformation. Most things I've read is that the max service temp of nylon is close to autoclaving temps (or just below) and its not suitable for use in boiling water either.

Given the temperature constraints, I can imagine exactly those plastics are fairly unlikely to be suited to the traditional extrusion heads on common-or-garden 3d printers.

Yes I would think moulding would be the way. One of my clients is a plastics company and I saw a presentation on it a little while back. The new plastic versions of things can be non-disposable but still much cheaper and lighter than metal versions.

We use plastic for single-use because of their low price. If the parts are to be re-used, I assume we could use silicon/glass/metal.

Is it really necessary to use a sterile valve for a new patient affected by the same strain of virus. I would expect the hospital to reuse a valve, instead of letting a patient die without the ventilator.

Would there be risk of contamination from other diseases that patients may have? I agree with your general expectation that risky reuse over high-probability death seems strongly preferable. One tricky moral dilemma is what to do if you have 1 clean valve, 1 dirty valve, and 2 patients.

> One tricky moral dilemma is what to do if you have 1 clean valve, 1 dirty valve, and 2 patients.

Why would this be trickier than having 1 clean valve, 0 dirty valves, and 2 patients?

Well it's exactly the same after you make the choice/use the first valve.

> what to do if you have 1 clean valve, 1 dirty valve, 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)."

Absolutely insane. It's barely started.

Triage as you would anyway (same prioritization), or flip a coin.

I think you'd at least want to sterilize it. Not sure if it's the case with these things, but other devices like endoscopes have enough cracks and crevices for stuff to hide so they have to go through an acid wash instead of steam, so the plastic probably wouldn't survive. Stuff like that either has to be non-corrosive metal or disposable.

Wouldn’t plunging them in a solution with high concentration of alcohol do the trick? (Not a medical professional, asking out of curiosity).

There are much better 'disinfectants': alcohol is good at killing living cells, but doesn't do a great job killing spores or penetrating into protein-rich....gunk.

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).

More emphatically: do not trust alcohol at any concentration to kill pathogens. (Likewise, freezing.)

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.

This crisis is moving technology to the forefront even more. From life saving equipment, to entertainment/communication/learning for the quarantined. There are countless ways for tech to play a large and important role.

I feel like we as the tech community need to let hospitals know we’re available to help with design and 3D printing of critical parts. But how wouldn’t the hospitals know what parts are 3D printable and how would we ever organize to contact them? I’m in the Bay Area and I could make myself available to design and print anything anyone needs. Does anyone have ideas on how to determine what parts would be helpful?

I urge makers and more professionalized 3D printers to familiarize themselves with FDA guidelines[1] on 3D printing and internalize their thinking. No doubt this will slow you down, but you have time right now to prepare before the shortages begin. This will make things safer for patients, give confidence to medical professionals that you have your head in the right place, and likely demonstrate good faith in the event that the FDA comes after you (after the COVID-19 crisis ends).

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...

it was a very good thing that someone was quick on their feet and made this happen but 3d printing is no solution to the systematic lack of these supplies.

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.

People may not realise this but we have wonderful methods to quickly mass produce plastic parts.

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.

I am in a kickstarter for an injection moulded item (a photo scanning holder) which has run 18 months late because of massive systemic backlogged production issues with the injection moulds requiring repeated re-re-re-re-fabrication.

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.

Injection moulding isn't hard but it's not uncommon for even experienced teams to take a couple of goes to get it right. "Re-re-re-re-fabrication" is excessive though and sounds like it was an inexperienced team.

Source: worked at a place that made injection moulded hardware.

its somebody else's kickstarter, not mine. Its the PIXLater, a diffuser and three or four parts which need to all fit the same size hole, to provide the right 35/128/70mm shaped holes for negatives, to be backlit by the diffuser, and sit on a tablet. The problems relate to the repeated click-unclick of the parts, which require strong corners and chamfers to insert, and friction hold. Every single iteration of design required re-design of the injection moulding. I think had the design been well worked out in advance it would have only been 2-3 cycles, but since the design had to overcome mechanical rigidity, re-insertion, keystoning the parts, over-dimension outcomes, under-dimension outcomes, it was fraught. Its finally reached production.

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)

So when you talk about getting anywhere near lego you're talking about a type of problem that's in a completely different ballpark. 3D printing can't even hope to reach those tolerances. If they were making a part that could simply be 3D printed, the type of manufacturing relevant to the article, I'm pretty sure they wouldn't have had all those problems.

If I was doing a kickstarter, I wouldn't go injection molding.

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.

This somewhat defeats the purpose of a classic kickstarter, which is to 'kickstart' the commercial production of something.

The platform is of course used for many different sorts of project.

Is this uncommon? Did you design a difficult to manufacture part? Is your manufacturing team incompetent? Or are you just unlucky?

(No offense meant but there has to be a reason, though I'm not familiar with the space)

Naïve question: can you 3D print a useful injection mould? And assuming yes, is there enough capacity to print enough moulds?

Yes and yes, though there are lots of caveats as to material types used. Whitepaper from formlabs here: https://formlabs.com/blog/3d-printing-for-injection-molding/

I think the situation may be a little more complicated than that.

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?

I do have experience in this area and all I can say is look at the market - injection moulded parts are literally everywhere. How often do you see 3d printed parts in commercial products?

I see more blow-molded parts than injection-molded parts, but there sure are a lot of injection-molded parts. And the cup I'm drinking from is made by metal-spinning, filled with a drink poured out of a deep-drawn can with a stamped top. As for 3-D printed parts, I only see them occasionally in commercial products, but remember that low-cost 3-D printing is only about ten years old — a lot of product designs, and even more product designers, date from more than ten years ago. And even ten years ago 3-D printing prices were a lot higher than today, and there was no such thing as a US$1000 SLA machine, for example, or PETG filament, or PLA filled with wood or brass.

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?

The parts that CAN be 3d printed will never need multiple back-and-forth cycles because 3d printed accuracy is terrible compared to the tolerances required for some complex injection moulded parts. In fact using the moulding method your surface finish is accurate literally down the molecular level. (Of course since we're talking about plastic that doesn't mean the overall shape and dimensions are that accurate but that problem is just due to the material properties, not the process)

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.

although I don't own this specific book, the gingery books I have are pretty great. https://gingerybookstore.com/InjectionMoldingMachine.html if you've got time and space, you can do way cheaper than $10k. You need a drill press and a metal lathe. If you're feeling ambitious, there are other books that will show you how to build a drill press and metal lathe.

Really depends on how committed you are to bootstrapping and saving nickels.

Yes! I've seen several machines on Youtube that's either built from this book or pretty close to that design. You could definitely make a fully automated DIY machine at home for less than $3k no problem.

> 3d printing has an extremely small window where they are useful for bulk use parts like this.

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.

It's a corner case...albeit a corner case we're currently experiencing.

There's a limited amount of space and logistics to have 1000s of spares of everything you might possibly need.

Depends a lot on the part and on the printer. Did you see the update? Some company "massprinted" them using a powder based printer. From the image I'd say 40 or 50 pieces.

That was likely 40 from a single print run, probably under a half hour. They probably could print 200 on a single two-hour run, but were told the quicker turnaround was more valuable.

Lol there's no 3d printer that will print those in one hour.

With FDM it's difficult to impossible, yes. Even with a really fast delta & klipper.

With SLS this is however realistic. Maybe two hours for a batch.

SLS is pretty slow, not sure which machine can actually print all those parts in an hour.

Makes you shrug when the US military is dependent on India and China for a huge portion of the medicine it uses including antibiotics and other supplies like this.

We already had this discussion with masks.

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.

Article about the Texas mask manufacturer mentioned above: https://www.washingtonpost.com/business/2020/02/15/coronavir...

Very interesting read.

Since there are likely plenty of rich people reading HN:

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.

Great idea. By doing something as mundane as being a counterparty on the other side of that futures contract, one can actually be the difference between that party being incentivized to produce in a crisis driven shortage or not.

And I think this is likely to be a limiting factor in more cases. If anyone reading this has good channels to companies, organizations or individuals to pitch this idea to, please do.

I love this idea!

> This is one of those instances where the government needs to step in and store inventory.

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.

>The Strategic National Stockpile does store an inventory of masks. 13,000,000 of them.

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.

It's not supposed to be a primary source for masks. It's a reserve.

To me it seems like we're in a situation where you would use such a reserve.

Exhausting that reserve in one day would do nothing to curb the epidemic, and do nothing to protect the nation other than delay any onset by one day.

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.

It feels like we need some better way to maintain systems in the fairly long term, after most people have forgotten their significance.

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.

The US HAD a department for exactly this kind of pandemic planning.

A bunch of anti-science knuckleheads decided to eliminate it.

Sure, but my concern is that over the course of decades, inevitably there'll be some group of anti-science knuckleheads who decide 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.

Sounds to me like that manufacturer needs to contact the operators of that stockpile.

not the other way around?

I know this has nothing to do with universities and the scales are very different, but it's pretty absurd that universities stockpile and hoard billions in endowments which they don't really need, when first-world governments and hospitals aren't stockpiling things like critical medical supplies and drugs (or at least don't stockpile nearly enough of it).

Stockpiling money is much easier.

They’re not stockpiling money. That money is working. It’s invested in productive assets, stocks, bonds, real estate, venture capital. They don’t keep dollar bills in a vault.

How are you defining "productive"? That still sounds zero-sum to me, though I don't know the details of how they invest the money. If it is zero-sum, then it's equivalent to stockpiling.

Economic growth is non zero sum. Everyone can get richer at the same time as available capital increases or workers become more productive due to greater human capital. Stocks are giving a company money to invest in return for a share of future profits. VC and bonds are lending money to companies to invest for a future payment. Those companies are investing in productive capital assets or in making their workers more productive. Not zero sum.

It's not zero sum, it is negative sum: it comes at the cost of using up non-renewable resources and destroying the environment.

You can invest in companies that are building/using renewable energy and/or helping the environment.

I used to think this, but in the past few decades I became convinced that only reduction in economic activity will help the environment.

Surely either replacing a coal plant with a nuclear one, or capturing the concentrated carbon dioxide right out of the smokestack, would help the environment with no reduction in economic activity.

Money isn’t zero sum because it has velocity. Companies financed with stock buy equipment using that cash, which the equipment manufacturers use to pay their employees, who then spend it at supermarkets...

Stop. You've got it completely backwards.

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.

My reply is exactly what another commenter wrote:

>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.

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.

Also, money usually doesn't have an expiration date. Drugs and most medical supplies do.

Money doesn't convert into medicine in an instant, and in times of crisis, that money will buy much less medicine.

Inflation causes stored money to expire. It's kind of a moot point in an era of 0% interest though.

Yes, but money can be stockpiled in somewhat inflation proof ways, like investing in assets that grow, e.g stocks. You can't stockpile masks that way.

I know, but important things are often difficult.

Can we 3D print a respirator yet?

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.

Not sure about oxygen respirators, but some civil organizations in Hong Kong have been 3D-printing filter cans for 3M 6200 face-piece respirators since mid-February, amid shortage of surgical face masks. The designs of the filter cans are modified so that, instead of using 3M's particulate filters, one cuts a surgical face mask into many smaller pieces and inserts only one piece of them into the filter can in each use.

Well, I counter this with "Just do it anyways".

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.

I agree 100%. In theory, everybody cares about FDA certification and all the legal details like HIPPA and stuff.

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.

>A reanimation device consists of a reanimation bag which is placed on the chest of the patient and which has a tube connected to the lungs of the patient, e.g. by a mouthpiece. When the bag is compressed heart massage may be effected through the bag by the operator. Thus, a single person may carry out resuscitation and heart massage alternately.


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.

I think it is a poor translation of "resuscitation device" like an AED or likely a bag valve mask.

Yup. It's a not-quite-false cognate in French too (https://www.collinsdictionary.com/dictionary/french-english/...)

Was very disappointed when the reanimation class wasn't taught by Dr. Frankenstein and only covered CPR.

Huh, in Dutch as well. That might be useful knowledge if I ever get in a situation where I need an AED in an English-speaking country.

I think the idea is that you perform chest compressions and pump a bellows at the same time with the same motion so you don't need a second person to breathe air into them. Just a guess.

Was anyone able to download the 3D model of the said valve?

I don't think the model is going to be shared.

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.

Valves are not subject to copyright. These valves are patented, as noted in the article.

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.

This is as good a case for jury nullification as I’ve seen in a long time.

Italy is not a common law nation, they are a civil law nation. I'm not a Italian attorney, but I don't think a jury would ever be involved, just investigators and judge.


I don't think they have juries in civil suits

Can someone explain what a "reanimation device" is? It sounds like a bad translation. Some sort of ventilator?

In Italian hospitals, there is traditionally a difference between Intensive Care Units ("Terapia Intensiva") and Reanimation ("Rianimazione") departments. Don't ask me what the difference is. A lot of what they do, however, tends to overlap, so in many places the two concepts were merged into a single department - which is basically what we'd internationally call an ICU dept.

So in this case "reanimation device" is just "a device used in ICU".

IIRC, "Terapia Intensiva" is for patients admitted through the ER and run by ER physicians while "Rianimazione" is for patients transmitted from other departments and run by anesthesiologists.

Just another case of NIH!

I assume its a vent based on everything I've read saying that we will likely run out of those.

Makes me wonder what would it take to 3D print other life saving devices? And are there designs online in the public domain ?

I won't be surprised if original part manufacturer sues these guys, hospitals for IP infringement...

That would be the most tone-deaf lawsuit I've ever seen.

Apparently the company have already threatened them https://www.theverge.com/2020/3/17/21184308/coronavirus-ital...

wasn't there a crowd effort to make and mass 3d-print ventilators? I guess I've seen in couple of days ago on reddit

Would it be possible to make or print parts to enable a single ventilator support multiple patients?

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.

You usually "tune" the ventilator parameters to each patient, so they'd have to be fairly similar size, need, etc.

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.

The Verge: "Medical company threatens to sue volunteers that 3D-printed valves for life-saving coronavirus treatments"

In this case IP might mean Intolerable Pricks.


I except someone is already be tinkering with a fully 3d printed reanimation machine using 3d printet plastics parts and off the shelf motors.

Anybody have a link to the STL?

I have a large format 3d printer (.5m x .5m x .5m) and am in Indiana.

There’s an update at the end of article, they already have some polymer SLS version in hand. Probably not using FDM prototype at this point

I am curious if this device needs to be leak-proof - to save oxygen or prevent the release of virus -???

The parts I print on my MakerBot are very porous...

Hopefully they won't get sued for using uncertified parts. Especially if somebody dies.

Italian here: they asked for the 3d model to the company producing the valve. The company answered with a threat of legal actions.

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.

In the US, that would likely be the case. In Italy, probably not. Desperate time calls for desperate measures.

Companies should be reminded that basis for patents is:

> 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?

Yes. Most countries have a system for compulsorily purchasing a patent if it's deemed in the public interest.

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.)

Hopefully the US Gov proclaims blanket amnesty so that people don't have to think about this.

Or fix the law, so that this issues goes away in the future, whithout having to repeatedly resort to ad-hoc measures and resulting legal uncertainty.

Good thing they're not in the US.

So a 5 cent part that probably costs $500 due to whatever insane economics determines the price of medical equipment is 3d printed, and we are supposed to think this is good? Or even ok?

This is insane.

I think maybe you missed the part about the reason for not having them easily accessible wasn't cost, but simply availability:

> 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.

If they didn't cost $500 I bet they would have more on hand...

I have no idea, but lets just say this works 98% of the 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.

I don’t really understand your reasoning, what are your arguments as to why it would be a good idea to give the choice?

You can choose as a patient between the following Scenarios:

Scenario 1: No treatment option and you will die.


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.

I think the point is more, don't even bother giving the choice. Just do the right thing which will maximize the survival of the patient, which is B in this case. People on the verge of death from illness aren't in the best place to be making logical decisions, and that's assuming that they're even still conscious or can talk (many of them are unconscious at this point and/or intubated).

Just follow the Socratic oath and don't over-complicate things. Give the necessary treatments.

I hate to be that guy but you're probably referring to the Hippocratic oath, although I wonder what a Socratic oath would entail :-P

Everyone promises to talk about and think critically of why they believe things to be true.

Y'know, this is probably a good one too!

I hate it when the doctor gives me a choice, because the answer is almost without fail “you are the doctor, please do what you think is best”. What I like is them explaining their choices to me, so I can understand them.

If there are two equal choices with cost and benefits, then maybe.

All medical procedures come with risk, and all medical devices can fail. It's not obvious that a material increase was introduced.

Lawyers like the second scenario because then the doctor isn't liable in the 2% case. Most people, especially in emergency situations prefer it when their doctor does what they think is best, I believe.

I wonder how this'll play out longer term. I'm betting the company who designed/manufactures the ones they're copying isn't super happy about the implications for this...

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?

In an emergency situation like this one, the patent owner company wouldn't risk a huge PR loss by suing or even complaining, but when the crisis will be over they might decide to act. However, if those valves are reverse engineered, copied and given away without selling them, there would be no profit. That should fall into fair use.

> That should fall into fair use.

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"

Wouldn't the owner have to show damages though? If they couldn't supply the widgets anyway, then what were the damages?

Ask this to record companies claiming losses for imaginary sales.

> the patent owner company wouldn't risk a huge PR loss by suing or even complaining,

Yeah. About that...


Fair use is not a legal concept in most jurisdictions.

In general copyright doesn't cover designs for machine parts, except insofar as they have an expressive rather than functional character, in which case they can be copyrighted as sculptural works. But you can still make replacement parts that copy their functional aspects.

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.


IP restrictions have gotten seriously out of hand anyway; academia and the public sector should be working on making "open" versions and know-how about basic essentials like this widely available. There's no excuse for keeping these designs under wraps like that, when for many of them the "basic" version is old enough that any patent rights on the underlying idea have long since expired.

IP rights can be suspended in a crisis, additionally the law of necessity might apply - you can generally break a law - such a breaking and entering, in a matter of life a death, for example saving a child out of a burning house.

Oh sure. I'm not for a minute suggesting that printing these and saving people's lives isn;'t 100% the correct and obvious thing to do here.

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.

I'm betting the company is really happy that their clients have been able to keep patients alive.

So it turns out...


They sell them for $11,000 each, and threatened to sue the guy 3d printing them for a dollar a piece...


Is this actually a patent infringement if they aren't selling the patented design? If I get some parts, maybe for my car, 3D printed and give them away is that illegal?

It's not really illegal in the "go to jail sense." But you can definitely be liable for patent infringement if you give it away, even if they aren't selling it.

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 don't know the proper legal answer to that, but you'd lose that argument in front of a judge if you gave away a locally printed copy of someone's music or writing. I wonder more about copyright here than patent infringement. I suspect Elon would lawyer-slam you pretty hard if you were 3d printing and giving away essential Tesla repair parts, no matter whether you were making money from it or not.

There is an enormous non-OEM (“aftermarket”) parts market for automobiles. It’s a little over $300 billion (with a B) per year in parts and service. Elon can pound sand if someone wants to make a Tesla-compatible aftermarket fender or brake pads.

Would fit in anticounterfeiting laws so, probably illegal, specially if the pieces aren't clearly marked as different from the original brand, and much more specially if they are designed for replacing the originals in public health purposes.

I want to see which jury will convict them for counterfeiting!

As long as is in the law (countries may differ) any jury that would not desire to be sued by prevarication. All of them.

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.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact