Italian hospital needed an $11K part for its ventilators, volunteers 3D printed them at $1 each, original manufacturer threatened to sue

boing
42

I think there’s more going on here than enforce the patent or lose it. If that were all that was going on, then the company could indeed offer then a trivially cheap license to produce copies of the part for a limited time.

But medical devices are also subject a lot of regulation and scrutiny. The manufacturer’s valves have been reviewed and tested and analyzed for patient safety, fire safety, perhaps even for its ability to be sterilized between patients. And there’s probably a manufacturing quality-assurance process to ensure the valves meet all the specs that helped them be approved.

The company may not be allowed to “bless” a makeshift PLA or ABS copycat valve with fragments of filament produced on a non-sterile 3D printer that may not be within the spec’ed tolerances. Offering the license may get them in trouble with the regulatory authorities. And if the fake valves kill someone, they may have liability if they license the design to a manufacturer without doing due-diligence on their ability to faithfully reproduce the design.

When you design a part, its design is guided by its function but also by its manufacturability. A different manufacturing process, like 3D printing versus injection molding, may require design changes, and those changes may interact with how the part operates.

Ventilators have heaters and humidifiers. Heat and humidity can have significant effect on most of materials used in small scale 3D printing. How does that effect the performance and durability of the valve? If the valve can fail in novel ways, will the rest of the ventilator detect the problem and raise the alarm? Will it aggravate the problem?

5 Likes
43

Well, to sexy it up and all, the corset case…

This is another easy to read article.

2 Likes
44

Long ago (cause apparently I’m old now), I was a summer intern for an electric company that had nuclear power plants. As companies do with interns, I got to participate in lots of training, tours, introduction activities that happened to be going on for other new real employees.

By chance one of those was following an engineer working with plant operations on upgrade of a switch on the control panel. It was a 4 way toggle switch, post that could move in 4 directions, not engaged in any when in the middle. The current switch on the panel had a thin metal post, and it they found it would bend over time. Meaning, as an operator went to move it, the post could bend and not disengage then engage again as expected. It still worked, but there was to much slop in the process. A quick move could cause 2 things to both be connected at the same time.

They found a replacement switch with a big thick solid post, with a ball on top for easy use. It failed the first round of QA though. The ball on top meant in an earthquake, if the building jumped, the extra weight could cause the switch to change positions, with undesired results. There was a second model of the same switch with a lock button carved out of the top half of the ball. That passed QA, and they were planning the installation.

It was a $100 electrical switch that cost $13,000 for the nuclear plant to buy. It needed QA certification that it passed all kinds of items. Quite the learning for an intern.

Things that have to pass life critical certification cost vastly more, and it’s almost all in the certification.

$1 vs $100 vs $10,000 depends entirely on what that certification guarantees. If it was my life, and I had the luxury, I would want the one that passed $10K worth of requirements. But, if the option was “nothing” vs “$1 worth of it’ll probably work”, sign me up for probably work every time compared to nothing.

12 Likes
45

Then they should grant temporary license to print the part. There is no way that bit of plastic costs 11 grand

2 Likes
46

Jerry-rigged is also common usage, & in the dictionary
.

3 Likes
47

And how many power plants was this switch approved for? That’s the piece that’s missing here.
Nuclear power is one area that has to have this level of quality control. But if you were producing 5k nuclear power plants per year all of a sudden that R&D for the $13k switch drops to a cost of only $2.60 per plant.

I agree with everyone though, in times of a crisis working is vastly better than not working.

48

People do misuse the term often, and so dictionaries record the usage.

5 Likes
49

Biomedical engineer here: Ok, so this is a part from a resuscitator not a ventilator. It looks like a disposable or single-use part. Ventilator parts are certainly expensive but $11k would be a very high price for any low-tech medical device part. Pricing generally reflects complexity and manufacturing cost for device parts. Pricing of devices tends to be driven by licensing and regulatory costs. So a new vent is usually $40-$80k with parts usually in the range of $100-$1000. If you bought all the parts separately, you may be able to assemble one cheaply. However, (in Australia at least) you then take on the liability as a manufacturer and you wouldn’t be able to sell it; also most hospitals wouldn’t validate it as safe for use. Compare human vents with veterinary vents to see the difference between manufacturing costs and the added price of safety, quality and risk management.

8 Likes
50

The $1 per printed part doesn’t pay for testing the part to medical standards. It doesn’t pay for the design of the part. It doesn’t pay for all the failed attempts to design the part. It doesn’t pay for part of all the other R&D for products that didn’t work out, nor for ones that have yet to be released.

A realistic price for the part is far far higher then the bare cost to produce. Especially for products produced in small quantities.

That said in a time of emergency the proper response isn’t “I’ll sue!”, but “be careful, those parts haven’t been tested and can’t be as durable, we are working hard to get the ones you have already ordered shipped to you, please switch to them as soon as they arrive”

6 Likes
51

Thank you, fellow pedant. You’re right.

2 Likes
52

Whatever happened to “No jury would convict me”?

1 Like
53

Fair enough. That said, the part is designed once. Failed attempts should factor in only so much as well, and should be a part of the overall design time/expense.

Is every part individually tested, or is it 1 in 10, 20, 50, 100?

Do they drop the price once they recoup their development costs, or do they release a new machine with different parts?

What is the average lifespan of these machines?

Do their parts truly have to cost what they do, or is it another case of having someone over a barrel and taking them for all you can, like John Deere and the farmers? You know. The shenanigans our great country was built around, in order for the few to lord over the many. In a crisis, that barrel really grows, if someone is willing to keep bending folks over it.

4 Likes
54

When it comes to life and death matters, YES, IT IS!

5 Likes
55

I have a possible solution. 3rd party re-certification, and medical EQA testing.

The nuclear industry uses something similar. I know because I was involved in the process, and worked for a company that used to provide this service for Ontario nuclear power plants… It is called EQA (Environmental Qualification Assessment) testing and certification. I used to do this for a living in order to get parts from manufacturers that did not have dedicated nuclear qualification departments. (Only a few companies, specifically some of the larger valve manufacturers actually have a dedicated department for this purpose).

The logic goes like this. There are two categories of device needed for a nuclear power plant.

a) Standard items that are not directly safety related (light switches, water piping and valves for bathrooms, electrical connectors, etc).

b) Items that MUST OPERATE CORRECTLY THROUGHOUT THEIR ENTIRE RATED OPERATING LIFE WITHIN THE PLANT IN ORDER TO PERFORM CRITICAL SAFETY RELATED FUNCTIONS.

a) Standard use items (light switches, adhesives, electrical connectors not being used in safety related jobs). Everyone needs light switches, but the switches must be qualified before they are allowed to be installed in a nuclear power plant. These items need to be qualified before installation. The nuclear power plant cannot buy the items directly but they can tell a re-certifier to buy the item, The certifier buys the order requirement plus 10% extra, tests the 10% extra to verify that the items perform as needed and do not contain prohibited materials such as teflon that fails in radiation environments, and demonstrate that the items meets minimum quality standards (insulation, lifecycle, leak proof, whatever is needed). The certifier can then sell the items to the nuclear power plant and everyone knows that statistically they are safe to use.

b) Items that MUST OPERATE CORRECTLY THROUGHOUT THEIR ENTIRE RATED OPERATING LIFE WITHIN THE PLANT IN ORDER TO PERFORM CRITICAL SAFETY RELATED FUNCTIONS. As an example, if an electrical relay is located in an area that can experience both earthquake damage and damage from a steam line break flooding the room with superheated steam. The relay has a rated operational of 25 years.

It is necessary for the relay to operate correctly over the rated 25 years, and be able to perform correctly immediately after both the earthquake and the steam-line break. If it does not, then there might be a bad outcome such as deaths or a meltdown.

In order to qualify this part for use in this role in a nuclear power plant, the part would be examined for prohibited materials like teflon, artificially aged (including both artificial thermal aging of polymers, and cycle testing the relay to the extent of it’s intended use in the plant, ex: 20 million cycles over 25 years). After artificial aging, the part is then put on a seismic shaker table to simulate the earthquake event (and monitored to ensure it works properly during the earthquake and doesn’t shatter or chatter), then it is placed in a superheated steam chamber to see if it will survive the steam-line break.

If the part survives all of this, then it can be qualified as usable in the plant for that specific safety related purpose, and the nuclear power plant can buy them from the manufacturer as long as the manufacturer keeps using the same production methods and processes to make the exact same part.

The medical industry needs a requalification system for parts and medicine.
items that this should be done for include
a) epi-pens
b) insulin
c) medical equipment compnents for parts that are out of use.

In the case of the valve in question, if the manufacturer cannot make the part, they can send the design to a qualified re-manufacturer, along with required specs and minimum quality requirements. The re-certifier must be able to test the parts in question to see that they meet minimum quality requirements and then can re-sell the remaining items for use in the medical industry.

This fits with the plans for medical devices being held in escrow and only released if required to qualified re-suppliers and re-manufacturers especially under crisis/good-samaritan situations. This would allow qualified, estimated safe parts to be reproduced for use in hospitals, especially during crisis times.

Parts would have to have re-rated lifetimes, but that would still extend usability of medical equipment and save costs and lives.

I’d personally love to see this done with epi-pens or insulin. Set up a contract to import the items from a country where they are competently manufactured for 10% of the price. Buy the required quantity and 10% extra. Fully test the extra 10% to verify that they met the minimum quality requirement. Then resell the rest at 20% profit undercutting manufacturers who sell them at 900% profit.

7 Likes
56

I’m glad that you are here to repeatedly to give them voice.

57

Yeah, I don’t work on medical products, or on things in small quantities. So I can’t actually talk to what an appropriate cost would be, just that $1 isn’t. I would be unsupervised for it to be anywhere from $50 to roughly what they actually charge. With my very very rough guess being around $400. Leaving a profit margin that even Apple would be shocked by…but again this is a pretty uninformed guess. There might be enough regulations involved that the real cost is actually much closer to the real price.

None of that makes the threat of a lawsuit even marginally acceptable (either from an actual PR point of view, or from a “do you have no soul?” point of view).

4 Likes
58

So now you have the update that says no threat involved, can you not amend the headline so the whole story isn’t misrepresented? If not you are perpetuating a lie and it smacks of poor journalism.

1 Like
59

Sometimes a 3D-printed part (for instance, a prototype intended only for fitment) will be suitable (single use) and other times not at all. Some years back, an in-house “customer” (right out of school) cheerfully brought us a 3D-printed part for testing. The test involved determining how much inert gas the part would allow through at different pressure conditions. Duh! The part was (given the particular ‘refinement’ of the part’s 3D process) porous; after reminding the customer of this condition – not applicable to the actual product to be – he still insisted on testing. So be it. After informing my manager of the charade to come, we plowed on ahead. Test day. The customer, his mentor, a couple of techs, and me (normally not there for that sort of test, but needing the laugh). We start the test, we can see the part on a monitor, data comes in, and I ask our operator to hold for a moment, then the real fun begins: The other tech goes into the test cell and squirts soapy water all over the outside surface of the part which then becomes a bubble machine. Big bubbles crazily developing and bursting along the part’s entire outer surface. The tech comes back in, looks at me, asks me what I want to do. I look at the kid’s mentor, who then says, yeah, let’s stop this.

5 Likes
60

UPDATE: Verge has updated their reporting on this with more accurate information:

And yet the title and its title on its facebook post is still wrong, and will apparently stay wrong, because it makes for a better click???

6 Likes
61

I prefer my juries rigged and my coffee white with 2 sugars thanks.