The Bleeding Edge

[drawoh]

I am watching this documentary on Netflix. This is all about implanted medical devices not working, and not being adequately tested. I am a mechanical designer. Let's look at this as an engineering problem. If I am designing a device that will be implanted in a human being and that will remain in the human being for the next fifty years, how do you test it? How do you anticipate how human tissue and fluids will interact with your device over the next fifty years, and how do you test them, and get something out to manufacturing in the next two years?

--
JHG

 

[JohnRBaker]

I'e never worked in this field, but I have to say that I hope the people who designed and manufactured my artificial heart valve, which was installed 18 months ago, knew what they were doing. So far so good

John R. Baker, P.E. (ret)
EX-Product 'Evangelist'
Irvine, CA
Siemens PLM:
UG/NX Museum:

The secret of life is not finding someone to live with
It's finding someone you can't live without

 

[3DDave]

Implanting in pigs. And monkeys. And some guessing. I don't see how anything happens in just two years unless it's a tiny change to an existing product. For a number of items there isn't a 50 year need - either the patient won't live that long anyway or, like pacemakers, they will need replacement before then. The trade-off is that the patient is already in so much pain or so close to dying that the expected outcome doesn't have to be much better to be used.

For example, a buddy of mine had a knee replacement; he had been in some pain and they looked at it with a tiny camera and found that not only was the cartilage gone, so was the bone outer surface. He had worn through to the marrow. No amount of rest or physical therapy would ever fix that.

That said, some of the stories I've seen, particularly in hip replacement, seem stupidly naive. Recall the "metal on metal" versions? How is it reasonable to expect to replace a self-lubricating material like living cartilage with bare metal of any kind? And who fails to recognize that metal wear products aren't going to be broken down and removed safely?

It seems like a tough business. There's a lot of effort in the vast interconnected network of cells and chemistry that is dedicated to ejecting things that aren't part of the network even if the attempt is self-destructive via allergic reactions. Compound this with the ability of the neighboring parts to renew in the event of damage while mechanical substitutes cannot.

And there's enough demand for an answer, any answer, that there's a lot of room for quackery.

 

[GregLocock]

One positive is that if a material is found to be acceptable in a previous device that has a respectable lifespan then the same material should be OK in a new design in the same environment. So we'll slowly build a library of suitable materials.

As to the more conventional parts of reliability testing, accelerated durability tests are a known concept, and of course IF they have decent, correlated FEA models then they can simulate 8 lifespans of activity very quickly. 8 lifespans is about a 50% error in stress estimates for a surprisingly large variety of materials.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[EdStainless]

And they keep parts on long term testing.
Sometimes this is in animals, but usually it is in synthetic environments.
One example is the leads for pacemakers. Out of each production run they have to put a group on long term test and they are only allowed a specific failure rate.
You can't get a new implantable device to market in two years unless it is a simple upgrade to an existing one. The basic testing and certification takes longer than that.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[oldestguy]

Hey, I have two hernia fixes using mesh. Well it does not make me feel toO confident when I now see on TV, (If your mesh is causing problems, call us) of course an attorney outfit. At least they hold better now than before the mesh was available. PROGRESS

 

[JStephen]

Mentioning pacemakers reminds me of my pacemaker story. My brother worked for a pacemaker manufacturer. They would put leads on their pacemaker that the surgeon would trim to length when installing the thing, so it didn't really matter how long they were. Except that for gov't approvals, they had to show that length very accurately. So the phrase there was "Measure it with a micrometer, mark it with a piece of chalk, cut it with a hatchet."

 

[MintJulep]

For a process, start with a comprehensive Design Failure Mode and Effects Criticality Analysis (DFMECA). Use that to define your test and validation plan.

List every component.
For every component, list every source of stress it could possibly be subjected to. Design loads, abuse loads, materials compatibility, fluid compatibility, thermal conditions, thermal cycling.
For each of those list how it could cause a failure and the criticality of each failure.
For those where the criticality is sufficiently important, write down how you can test to establish that the component is sufficiently immune to that stress source.

Tedious.

 

[drawoh]

MintJulep,

That sort of thing is challenging when you know what you are doing. An interesting issue is the removal of the implanted equipment. You will appreciate this if you watch the documentary.

--
JHG