In this episode, we dissect the evolving landscape of FDA medical device regulations with regulatory expert Mike Drues.
These discussions provide a comprehensive overview of the historical context and current implications of FDA guidances, with a particular focus on 510(k) predicate selection, clinical data requirements, and the nuances of permanent implants.
The conversations address the challenges and complexities inherent in aligning medical device design and development with contemporary FDA expectations and offer invaluable insights for professionals in Product Development, Quality Assurance, and Regulatory Affairs.
Interested in sponsoring an episode? Click here to learn more!
Listen now:
Like this episode? Subscribe today on iTunes or Spotify.
Some of the highlights of this episode include:
- How the 510(k), originally an exception, has ironically become the predominant pathway in medical device regulation.
- That, despite efforts to raise regulatory compliance standards, the enforceability of current guidances is limited due to unchanged regulations.
- Discussion of FDA's non-binding guidances balance between being commendable for clarity and somewhat overreaching, reflecting a nuanced industry challenge.
- How evolving FDA thinking underscores the importance of aligning with contemporary standards, going beyond mere compliance.
- That selecting predicates with past design-related recalls for 510(k) submissions poses legal challenges and necessitates reevaluating older predicates in light of new risks.
- How clinical data requirements in medical device regulation are dictated by engineering and biological factors, not just regulatory pathways.
- How demonstrating substantial equivalence can be complex, and small technological differences must not introduce new safety concerns.
- The definition of 'permanent implant' challenges traditional understanding, emphasizing the importance of biocompatibility and safety in regulation.
Links:
Memorable quote:
"Just because you're meeting the standard, that just means that you're passing... That doesn't necessarily mean that you're making a safe and effective product." – Mike Drues
Transcript
Etienne Nichols: Hey everyone. Welcome back to the Global Medical Device Podcast. Today we're going to be talking about three guidances that were dropped from the FDA September 7. With me today is Mike Drews. I'm excited to have you with us to talk about this concept and something that you've been teaching for a long time.
Good to have you today, Mike.
Mike Drues: Thank you, Etienne, and always a pleasure to speak with you.
Etienne Nichols: Before we get into the direct guidances, though, and talking about each one specifically, if we get a chance to talk to each one specifically, do you want to give kind of a thought on why these might be coming out, what the objectives behind these might be?
Mike Drues: Yeah, great question, Etienne. And once again, thanks for the opportunity to have this discussion with you and your audience. So maybe just as a quick intro for those in the audience that might not be familiar with these recently published guidances, the titles of them, and we can provide links on the podcast website.
The first one, best practices for selecting a predicate. This is the one that Etienne and just mentioned a moment ago. The second one, evidentiary expectations for 510(k) implants, and the third one on clinical data in 510(k). As I said a moment ago, all of these are under the general area, the general umbrella of the 510(k).
In terms of your original question, what is FDA's motivation for putting these out? Well, in FDA's credit, they're trying to give industry a better understanding of the, what's involved in it and what the expectations are, which I think are all very admirable things.
But at the same time, Etienne and I have to point out the irony, almost hypocrisy of this. The 510(k) was originally created by Congress back in 1976, and we are rapidly approaching the 50-year, the one century anniversary, if you will, of the. We're still trying to figure it out.
FDA has put out a litany of guidances around the 510(k) over the last almost half century, and we're still trying to figure out, let me remind our audience, Etienne, and that although FDA has put out guidance, many, many guidances on the 510(k) over the last almost five decades, the regulation itself, for the, this is an important point.
The regulation itself has not changed, not one word, not one punctuation mark since it was originally created in 1976. And I think that's an important point. Also, another thing that folks should remember, many don't even realize is that when the 510(k) was originally created almost a half a century ago, it was meant by Congress to be the exception rather than the rule. Let me say that one more time, Etienne. The 510(k), when it was created, was meant to be an exception rather than the rule.
Fast forward nearly a half a century later, and for better or for worse, as everybody knows, the 510(k) has been the rule, has become the rule rather than the exception.
Should it be that way? That's a topic of a whole different discussion, but maybe some of those issues we'll get into. So just wanted to give everybody a little bit of a historical perspective first before we continued on.
Etienne Nichols: Yeah, that's a good point. I'm glad you bring that up, too. It's funny, I've heard you use the tax code as kind of just a parallel, and I imagine if somebody was giving a loophole, that's meant to be an exception, but it's legal to use, someone's going to start using it all the time.
And so, I guess that should have been something we would have expected. One of the things that make me wonder, though, when we talk about this, the actual regulation has not changed, but the guidances have changed. Do you think this guidance has the potential to reshape how professionals approach device comparisons? I don't know. We need to get into what's actually in the guidance.
But after looking at it, do you think it changes the best practices?
Mike Drues: Good question. Etienne and I would like to say yes, but maybe I'm getting old.
I don't know. Like I said, people have been trying to figure out the 510(k), both in industry as well as the FDA, for nearly a half a century.
When I read these three guidances that just came out last month, just like many other guidances, I'm constantly reminded of what the old French philosopher said. I can't remember his name. But the more things change, the more they remain the same.
The more things change, the more they remain the same. There's nothing in any of these guidances that is even closely new for me. For me, now, I've been doing this for a long time, and in addition to my regulatory background, I have a good biomedical engineering and a medical background as well.
So, nothing in these guidances were good for me. Sorry, new for me. I agree with most of what's in the guidance, although as we'll talk about, I think it's interesting, I think FDA is overreaching in what some of the things that they're asking for in these guidances.
Not that I think that they shouldn't ask for those things, but they can ask for them in the guidance, but it's not in the regulation. So, a company, if they wanted to, and I'm not suggesting that they should, but if they wanted to, they could easily push back.
Because as we all know, guidance is not binding. Regulation is, but guidance is not. So, it's admirable for FDA to try to start to raise the bar with these guidances, and I think they probably should have done that long ago.
But in order to really, truly raise the bar, to make it enforceable, some of these changes need to be considered to put into the regulation, not simply the guidance.
Etienne Nichols: Okay. And I know we talked about three different guidances. I think maybe we need to hone in a little bit on the selecting a predicate, just because we've kind of been alluding to that one for a little bit here.
Maybe we can address the others. Towards the end of the episode, you mentioned overreaching. And so, I can't help but curious if you want to talk a little bit about your thoughts on what aspects may be overreaching.
Mike Drues: Yeah. So again, I think I'd be happy to point out some of those what I consider to be potential overreaches as we discuss them. But again, I just want to point out, with regard to specifically substantial equivalents, I just want to remind our audience, and we've talked about this before as well, that a very significant number of 510(k) are still rejected today. In 2023, specifically because of substantial equivalence or the lack thereof?
Is it a little bit of an aside, Etienne? And I find it interesting how so many people think that this concept of substantial equivalence is so easy, such a no brainer, so easy, that a caveman can do it, as the commercial says.
But if it was, then how do we explain the fact that still right around 70 or 75% of 510(k) are rejected? And of those that are rejected, about 85% of those are rejected specifically because of substantial equivalence or the lack thereof.
So, if substantial equivalence is such an easy thing to understand, how do we explain those statistics? Back to your original question, Etienne, and in terms of what did actually FDA provide in this particular guidance? They shared some of their best practices. And I'm sure it's going to sound a bit arrogant for me to say this in a public podcast, Edie, and but it is true, many, if not all, of the best practices in this guidance, as well as the other two guidances, come right out of my Mike Drew's regulatory textbook. I've been telling this to companies for a very long time.
It's in some of my PowerPoint training slides that I've been using for many, many years. So, for example, they share basically four criteria in best practices in selecting a predicate.
Choosing a predicate that is based on, well, sorry, that is cleared using well established methods. I would like to think that all devices that are cleared, indeed all devices that come onto the market are using well established methods.
Of course, what does the phrase well established mean? We could certainly quibble about that. Predicates devices that meet or exceed expected safety and performance.
This is a potential overreach, in my opinion, Etienne, because if you look at what the regulation says, not the guidance, but if you look at what the CFR says about the 510(k), it says you have to show that your device is at least as safe.
But when it comes to efficacy or performance, that's where it gets very nebulous. Right?
And also, Etienne, here's another question to consider, and this is not unique to this guidance. We've been struggling with this for the last 50 years.
How can I argue that my device is safer or has better efficacy, and yet it's still basically the same as I. E. Substantially equivalent than another device. So, it's almost an oxymoron to try to do that.
You could certainly do that in the De novo, no problem. But in the 510(k), sometimes I hear people use the word new and five and 110K in the same sentence, and it's like how do you put those two words together in the same sentence?
The third of the criteria is predicates without unmitigated use related or design related issues. This gets to the recall question that the industry has struggled with before, and that is, should we be allowed to use a predicate that in and of itself was recalled?
And Congress did change the law on that a few years ago. But in my opinion, Congress changed the law too far, and we can talk about that more in one of the other guidances.
And then finally, predicates without an associated design related recall that's similar to the one that I just mentioned. Congress did change the law, but they didn't go explicit like that to say, if you have a design related recall, you cannot use it as a predicate.
Now, on a personal note, Edie, I think it's unconscionable if a company were to base their device on a similar design, if the other design is having problems, not to at least consider the possibility that your device might have similar problems as opposed to if it was a manufacturing defect and their device was manufactured in a method that was completely different than ours.
Right. And I would like to think that we wouldn't need regulation to tell people that basing your device on a design that was recalled that led to problems is probably not the greatest idea.
But unfortunately, Etienne, and I guess I didn't fall off the turnip truck yesterday because some people have done exactly that.
Etienne Nichols: I appreciate you going through those different things that is specified in the guidance document. One of the things that kind of was going on in my head when you were talking about those things is this.
You mentioned 1976, when the regulation first came out. Has the thinking from FDA kind of evolved to this point? And now they're saying, hey, we have gradually come to this point in our thinking, and these are the things that will likely lead to rejection.
For example, for your five hundred and ten K, or is this saying, this is what we'd like you to do, and obviously you can argue otherwise, what are your thoughts?
Can you comment?
Mike Drues: Yeah, great question.
Yes, it does reflect more current Thinking.
Oftentimes people like to use the phrase when it comes to guidance documents. As opposed to what other kind of thinking? I'm not sure, but current thinking.
I'm sorry, can you repeat that last? No.
Etienne Nichols: Yeah.
Or are there other ways to. I almost hate to bring this up because I wouldn't want someone to argue otherwise. Why would you want someone to not use a best practice? But I guess. What do they mean when they say best practices to me?
If you read, for example, like, is it Fred Crosby who talks about quality is free? His definition of quality was it meets the standard. Well, the question is, is the standard appropriate or know, there's always that thing.
Okay, do we have to go best practices, or is the standard not good enough? I don't know.
Mike Drues: Well, yeah, that's a very good point, Etienne, just because you're meeting the standard, as I've said in many of my podcast discussions before, when a company gets a 510(k) clearance, when they get a de novo granted, when they get a PMA approved, when they get ISO blah blah blah certified, whatever it is, when they get a CE mark, that's the academic equivalent of being a C student. That just means that you're passing. That just means that you're meeting the standard.
Right. You're not flunking. Right. But that doesn't necessarily mean that you're making a safe and effective product. It certainly doesn't necessarily mean that you're making a good product. My point of this discussion, Etienne.
And then we can maybe move on to some of the other questions.
Just using the last one as an example, a predicate device without an associated design. Related recall. If a company wanted to push back, and I'm not suggesting that they should, but if they wanted to, they could push back easily on purely regulatory or legal grounds, because it's not in the change that Congress made, is it in the guidance? Absolutely. Is it a Mike Drew's best practice? Absolutely. To be honest with you, if I saw one of my customers doing that, I would point it out to them.
And if I thought that it was not appropriate to do that, and if they, for whatever reason, drew their line in the sand, so to speak, I would probably say, thank you very much.
But you know what? Go find yourself a new regulatory consultant, because that is certainly not my best practice. But strictly speaking, from a pure regulatory ground here, I think that FDA is on very weak grounds. I think they shouldn't be. And I think that Congress should clean this up more officially, so to speak.
But at the moment, I think they're on pretty weak grounds.
Etienne Nichols: Right. I don't want to beat a dead horse. I don't know whether or not it's dead yet or not. But I guess in my mind, I'm just exploring this a little bit and thinking, okay, well, let's say a company wants to go down that route.
They have a predicate that had a design related recall, but they themselves have the wherewithal to think, okay, we are actually going to test and develop a device that will mitigate whatever that previous one has.
But in order to prevent going another pathway and we can't find another predicate, I suppose that would be the area or the time when it might be appropriate, depending on the company.
I don't know. That's just a thought in my mind.
Mike Drues: No, I think it's a terrific thought, Etienne, and kudos to you for having such a thought, because let me slightly clarify what I said a moment ago. I don't think anything that I said a moment ago was wrong, but let me just clarify it a little bit, because I don't want people to misunderstand.
If you're in a situation where your device is substantially equivalent to a predicate, and there's a similarity in your design to the other device, and specifically, the other device has had a known history of problems, never even mind if it led to a recall.
It could be just a known history of problems.
In my very, very strong professional opinion, not just as a regulatory consultant, but as a biomedical engineer, a company should automatically take those things into account. In other words, it's one of my best practices, as it has been for decades, Etienne. And whenever I bring any device into the market, whether it's a anything else, I will always look for similar devices, and I will always look to see what kind of history that they have.
Do they have problems with this device and so on? And then with that list of problems, ask yourself, is it possible for these problems to happen to me? If it's not, then why not? Like I said, if it's a manufacturing difference and your device is put together using a different manufacturing method, okay.
But if it's a design difference and your device has a similar design, but you have, how do you want to say it? Taken measures to either improve the design and or mitigate that risk such that the residual risk. Residual risk, by the way, is the amount of risk that remains after we mitigate what we can in terms of the risk.
If the residual risk is acceptable and we do all those things and it is still substantially equivalent, then I have no problem doing that. I think it's very appropriate to use the 510(k).
This is why I don't think that solving these kinds of problems via regulation or even guidance is the right way to do it. Solving these kinds of problems via getting people to think that's the solution here.
Present company excluded, of course. I'm sure.
Etienne Nichols: Well, I have one other question, and there's something about. You talked about the potential for overreach, particularly in the area when you said could meet and exceed performance requirements. So that being said, this going to sound like a little bit of a dumb question, but could it be that safety has changed over the years?
And I know that seems strange, but let's just say I'm going to take a predicate from 20 years ago, and that predicate was perfectly fine at the time, but now we have additional risks, perhaps cyber related risks, for example.
And that I need to, this gets into the area where I may think, well, is that device that's still on the market? Well, does that mean it's unsafe, it's cleared, it doesn't have a recall, and I need to exceed that.
I don't know if you have a thought or opinion on that.
Mike Drues: Well, I know, Etienne, and we're still relatively new in doing these podcasts together, but I've been doing them with Greenlight for a decade now. When have you ever asked me about something that I don't have an author opinion.
Etienne Nichols: Just trying to give you an.
Mike Drues: No, but I think it's a good question. What is safety? And has our bar for safety changed over the years? Yes, it has. And that would be a good thing.
I would like to think that overall, our standards today are better than they were 25 years ago or 50 years ago, or 100 years ago, or whatever it is. But the other thing to remember, Etienne, and is you mentioned something to the effect of if a device is cleared and is on the market, can we assume that it's safe or effective? That's sort of a loaded question. I mean, I could give you examples of devices, and to be fair, not just devices, but drugs as well, that have gotten through the FDA that most physicians never in a million years would ever use because they question the safety and or the efficacy of the device.
So, this comes back to what I said before. When you're meeting the regulatory requirements, that's the academic equivalent of being a C student. I think as an industry, Etienne, and we can, and quite frankly, we should do better than that.
Etienne Nichols: Yeah.
Okay. Any other thoughts or things we should really be focused on and looking at with this guidance, the best practices for selecting a predicate, knowing that the comment period is not up, any other things you'd like industry to be thinking about?
Mike Drues: The only thing that I would say is, not to sound self-serving here, but I did a couple of webinars specifically on the substantial equivalents for Greenlight Guru, where we go into that topic in much more detail.
The first one I did was several years ago, long before this particular guidance. If anybody is interested, please check out that webinar, because in there, I get into very specific, simple examples of substantial equivalence, and many of the practices that FDA is describing and the guidance that you and I are talking about are literally right out of those webinars and those podcasts.
And maybe this is a topic that you and I can dig into in more detail in a coming podcast discussion. I would love that.
Etienne Nichols: Okay, the next one we were talking about. So, three came out. And do you want to talk a little bit to maybe the association of the three that they would all come out on the same day?
I thought that was interesting, but, yeah, whichever one you want to dive into next.
Mike Drues: Well, first of all, I can't comment as to why all three guidances were published on the same day. I have a feeling it was probably coincidence. To be fair, FDA has been working on these guidances for a while.
And by the way, when you look at these three guidances, they're all pretty short. I think the longest one is like 26 pages. And I hate to maybe I shouldn't say this, but when you look at the content, at least half of each of those pages and those guidances are a bunch of you know what?
So, there's really only a couple of pages in each one. That's really important. Notice I'm not saying new, because as I said before, there's nothing in here that's new. And some people who might not know me might think that I'm being arrogant here, but I'm really not.
I mean, I've been saying these things for a long time. So, take a look at these guidances. The first ones that we just talked about. The second one, though, I think that the second and third ones go together.
The clinical data that may or may not be required in a 510(k).
Pardon me? And the evolutionary expectations for 510(k) implants. Maybe the clinical data question is the better one to start out with for our next part of our discussion. Etienne.
Etienne Nichols: Okay, yeah, let's look in that. What would be driving the emphasis on the use of clinical data in a 510(k)?
Mike Drues: Well, I can tell you from my own practice, one of the most common questions that I get from customers, almost on a weekly, if not daily basis, is for my new medical device.
How do I know whether I need clinical data or not? I mean, it's amazing to me how many people ask me this question, and so many people think that this is a regulatory question.
As a matter of fact, some people, Etienne, in, they make the assumption that if I'm doing a. Do not need clinical data, or alternatively, if I'm doing a PMA. I do need clinical data.
There is nothing anywhere in the regulation that says that, and nor should it. In my opinion. Whether you need to do a clinical trial or not is not predicated on the regulation that is totally agnostic to your regulatory identity or your pathway to market. 510(k), de novo, PMA, HDE, whatever it is, the decision for whether or not you need clinical data is based on the engineering and the biology. It's as simple as that.
Not the regulation, the engineering and the biology. So, for example, looking at the key points in that guidance, FDA shares four scenarios where clinical data might be required. If there's a difference in the indication for use, if there's a difference in the technological characteristics, if substantial equivalence cannot be determined by non-clinical testing alone, and finally, if there's newly identified or increased risk compared to the predicate device.
Those are the four different criteria, the four different scenarios. Again, to me, Etienne, and this is common sense, let me just give a couple of real quick examples. So, if you have a device where you're changing the indication for use compared to the predicate, first of all, you have to be a little careful, because depending on how much you change it, that might kick you completely out of the 510(k) universe. And assuming that your device is still class two or lower, that would put you into a PMA, sorry, to a de novo.
So, if you add a significantly new or different indication, let's say, and again, I'm just going to use this as an example, let's say the predicate device is cleared for a diabetes indication, and you want to use exactly the same device, but you want to use it for a cancer indication, right? That's a pretty big difference in the labeling and the indications. You're probably not going to be able to show substantial equivalents just based on benchtop or literature or something.
You're probably going to need clinical data to do that, if you can do that and keep it in the 510(k) university to begin with. The second criteria, differences in technological characteristics.
This is another one that I find fascinating, Etienne, and our industry has been struggling for 50 years since the 510(k) has been created. How different can two devices be in terms of either labeling what we just talked about, or technology, or both?
How different can they be and yet still be close enough, similar enough to be substantially equivalent? In spite of the fact that we've been using the 510(k) for a half a century, almost, in spite of all of the guidance, including the newest ones that were just published, there is absolutely no simple answer to that question, none whatsoever. So, again, to get into the regulatory minutiae a little bit more, any differences? Whether it's labeling or technology, cannot do two things. One is they cannot raise new questions of safety and efficacy.
That's requirement number one. And requirement number two, they cannot change the overall risk. And specifically, when it comes to risk, as you may remember it in, I happen to be a subject matter expert for FDA in a few different areas, one of them being risk.
When it comes to risk, specifically for the 510(k), you have to show that there are no new risks in your device compared to that of the predicate and of the known risks between your device and the predicate.
You have to show that the level of each of those known risks is the same or lower. In other words, hypothetically speaking, if FDA can identify one new question of safety and efficacy, or if they can identify one new risk, whether it's a new risk that's not in the predicate or an increased risk already in the predicate, that one little thing alone, theoretically would be enough for FDA to say, sorry, this is not 510(k) able. You got to think about a de Novo. Now, I can tell you in the 30 plus years that I've been playing this game, up until very, very recently, Etienne, and the regulation for the 510(k), which, as I said earlier, has not changed one letter, the guidance has, but not the regulation. The regulation has never been interpreted or, pardon me, interpreted or enforced that literally.
As a matter of fact, the example that I often use, and I've used this example many, many times at FDA in the past, it was possible, and that's actually happened several times, for a company to get an MRI onto the market using a CT as a predicate.
Let me say this one more time. They got an MRI onto the market using a CT as a predicate. Now, it shouldn't take an MD or a PhD in biomedical engineering or even an RAC after somebody's name to appreciate that.
G, maybe there's some different questions, maybe some potentially new questions of safety and effectiveness. Maybe there's some different risks in Mr. Which is magnets and CT, which is X ray.
But in the past, that happened on a regular basis.
Today, in 2023, I don't think there's a snowballs chance here in San Diego that that would ever happen.
Etienne Nichols: Yeah.
Mike Drues: As I said, just to close up, the last two substantial equivalents cannot be determined by non-clinical testing alone. So sometimes you need clinical testing not to show safety and effectiveness, but to show substantial equivalence.
If you can't do it on the benchtop, or via literature or computational modeling or some other way. You might need clinical data, or at the very least, you might need some usability data, or possibly both.
And finally, as I said, their scenario number four, which I found very interesting, that they put it in this guidance. Newly identified or increased risk compared to the predicate device.
Newly identified or increased risk. Well, it's clear to me that whoever at FDA wrote this particular part of the guidance either this is going to sound harsh to some people.
They don't know the difference between guidance and regulation because that is not in the regulation. That would automatically kick you out of the 510(k). Okay, that might be a little bit harsh, or maybe to FDA's credit, they are trying to raise the bar without continuing to wait around and have Congress do it.
And unfortunately, Etienne, and you know how long it takes Congress to do anything. So, I would like to think that it's the latter. I would like to give my many FDA friends the benefit of the doubt and say; this is their job.
Sorry, this is their attempt to kind of ratchet things up. Just like recently. This is their attempt to ratchet up the regulation of LDTs, of lab developed tests, which you and I have talked about on a previous couple of podcasts.
To FDA's credit, they got sick and tired of waiting around for Congress to do it, so they're trying to do it via the guidance. It's an admirable goal, but as I said earlier, if a company really wanted to, they could easily push back.
Etienne Nichols: Yeah, that makes sense.
I'm glad you brought that up. I was curious about that, because when you started talking about the identifying new risks, obviously that's in the regulation, and then this clinical requirement. So, what other parts of these clinical requirements or the clinical data and the pre-market notification recommendations do we need to be thinking this?
They laid out pretty simply, like you said, it's not a very long document. Honestly, any other things that even not in the document, maybe best practices that aren't there?
Mike Drues: Well, great question, Etienne, and let me share with you my regulatory logic, or even simpler, my commonsensical approach for preparing for today's discussion. I printed off one of my PowerPoint slides from one of my training that addresses this.
And this slide has got to be at least ten years old. And basically what's in the guidance is summarized on this slide, and I'd be happy to share it. If somebody wants to send me an email, I'd be happy to share this particular slide.
But basically, what it means is if you have a device and the regulation, not the guidance. The regulation says the device is substantially equivalent if, in comparison to the predicate, it has the same intended use as the predicate and the same and same technological differences.
Okay, that makes sense.
Sorry. Same technological characteristics, or the device is substantially equivalent if it's the same intended use and it has different technological characteristics. What the heck does that mean?
What that's saying in common sense, common English, is, as long as the labeling is the same, the technology can be the same, or it can be different. What the heck does that mean?
As I said before, how different can two devices be, either in terms of labeling or in this case, in terms of technology, and still be close enough? But the regulation, again, not the guidance.
The regulation goes on to say any differences, whether they're in the labeling or in the technology, they cannot raise additional questions of safety and efficacy, and they cannot change the overall risk.
So back to your original question.
It's very, very simple. The closer your device is in both labeling and technology to your predicate. Remember, substantial equivalence does not mean that they are exactly the same. It means that they're similar.
They're basically the same. They're close, they're substantial equivalent. The closer you are to that other device, the less likely you will be reliant on clinical data.
On the other hand, the more different you are between your device, what we call the subject device, and the predicate device, the more likely it's going to be to require clinical data.
It's as simple as that. I mean, that's the regulatory logic. Now, the reason why I'm hesitant to give that advice oftentimes, Etienne, and even though I give it all the time, is because from a regulatory perspective, that is exactly spot on, the correct advice.
But from a biomedical engineering perspective, it makes my blood pressure go through the roof when I give that device, because what is that doing, Etienne? And that's encouraging companies to bring out products that we already have, I.
#MeToo. And let's be honest, a synonym for the 510(k) is a #MeToo.
And it creates, maybe not disincentives, but it certainly does not create incentives for companies to do anything truly new or different. So, the clinical data question, and it's amazing to me, clinical data is never, ever either required or not required under the regulation.
And in my strong opinion, it shouldn't be. Clinical data, or the lack thereof, is based on the engineering and the biology, not the regulation. Remember the adage that I've shared sometimes before, Etienne? And there's an adage that I used to use for my medical students all the time, that is, the surgery went perfectly, but the patient died anyway.
Well, the regulatory equivalent of that is we followed the regulation perfectly. That is, we did all that FDA asked us to do, and yet the patient died anyway. Unfortunately, these things happen more frequently than some people would like to think.
And the solution, in my opinion, even though I'm a regulatory consultant, the solution is not to create more regulation because we've already got thousands and thousands of pages of that.
The solution gets people to think. Once again, present company excluded, of course.
Etienne Nichols: Well, I'm curious if you have any thoughts on just what, maybe when we finish the guidance documents, there's a couple of thoughts I might want to ask about. But anything else in this clinical data before we move on to the evidentiary expectations for.
Mike Drues: Those are some of the highlights. I think it would be great to take a deeper dive into this topic in a separate podcast, but for right now, that's probably sufficient.
Etienne Nichols: What about the third one, the evidentiary expectations for 510(k) implant devices? What did you see there?
Mike Drues: Great question. So, first of all, let's start out with a higher level and somewhat philosophical question, and that is, should any permanent implant be 510(k) able? And by the way, here's a little bit of regulatory trivia for you, or regulatory minutiae. Etienne, do you know what the regulatory definition of a permanent implant is?
Etienne Nichols: Not off the top of my head right now.
Mike Drues: Well, first of all, thank you for admitting what you don't know. My wife has no problem reminding me of what I don't know. So, the regulatory definition of a permanent implant here in the United States now, it's different in some other places, but here in the US is greater than 29 days. Greater than 29 days.
It is not what a lot of people think. In other words, permanent does not necessarily mean permanent in this context. It means greater than 29 days, to quote a famous politician, it kind of depends on your definition of is, what does is mean and why it's 29 days as opposed to 45 days or a year or five days or whatever, that's a topic of a different discussion. But permanent implants, even, for example, a breast implant, something that I've been involved with, both in terms of regulation as well as in terms of product liability, these were never intended to last the rest of the patient's life, even though we call it a permanent implant.
Many of the permanent implants that we use today, orthopedics, is littered with examples, are 510(k). And quite frankly, the reason why they are 510(k) is because they were on the market prior to 1976, when FDA started regulating medical devices and they were grandfathered in.
But once again, one could easily raise the question, should any device that is intended to be put in the body for more than 29 days, should it be a 510(k) before going into the specifics of this guidance at the end?
Because I know we're starting to get close on the end of time. Any thoughts on that question of should a permanent implant be allowed to be 510(k) above?
Etienne Nichols: No, it's a really good question. I guess part of me goes back to the biological characteristics or the way the metal interacts. So, I'm just thinking like a screw, that's a titanium screw, surface material all being equivalent.
I can maybe see an argument there. I don't know that I would, across the board want to say, no, they should not be 510(k) able. But yeah. What are your thoughts?
Mike Drues: Well, first of all, let me be crystal clear on this. I'm not suggesting that no permanent implant should be 510(k) able. That is absolutely not what I'm saying. I don't like regulation that's written in absolutes and black or whites, there's got to be an infinite number of shades of gray.
And you raised a perfect example. In the general area of biocompatibility, I happen to be. We talked about risk a little bit earlier, as you and your audience may remember.
I happen to be a subject matter expert for FDA in the area of biomaterials and biocompatibility. Very difficult, not impossible, but difficult to show biocompatibility for a permanent implant, or indeed any implant going into your body or coming in contact with the body without doing some sort of clinical or biological or animal testing.
Not impossible, but difficult to do.
All right, so specifically, when it comes to this guidance, FDA does break down in a little bit more detail than at least I remember seeing in any previous guidances. But it's all in my approach already.
Some of the general considerations to take into account are the indications for use, the labeling, the intended duration of the implant. Now, that scares me a little bit, because if you're only going to put a device in somebody's body for, say, an hour or a day or a week, does that mean from a mechanical or possibly even from a biocompatibility perspective, we should be treating it differently than a permanent implant?
I'm not suggesting that in some cases it would not be justified to do that. I think there might be cases, but I think that we have to be very careful how far down that road we go.
And finally, anticipated patient and physician experiences with the implant, this gets into a little bit of usability and so on. And metal on metal hip implants is a perfect example of that. But there are a number of non-clinical recommendations, and I'm just going to tick off the ones that are mentioned in the guidance, biocompatibility.
And by the way, I have to share with you a quick story, Etienne, and I wish I could say this was a make believe, a fictitious story, but it's not.
I had a customer come to me several years ago. They told me that they were working on a permanent implant. And so, after asking know my series of general know, have you done this?
Have you done that? We got to the topic of biocompatibility, and I said, where are you on your biocompatibility testing? Now keep in mind, Etienne, and this was for a permanent implant.
And they basically turned around and they said to me, what's that?
Now you're smiling, Etienne, and I'm being 100% serious. 100% serious.
I would like to think that, and again, I'm not trying to be tried here. I'm really not. But I would like to think it does not take an MD or a PhD after your name to appreciate that.
Gee, if you're going to put something in contact with the patient, certainly if you're going to put them inside their body, don't you think it might be a good idea to ask if their body, specifically the immune system, is going to react in some way to this particular device? But I guess, regrettably, Etienne, that's one of the reasons why we have thousands and thousands of pages of regulation, because we do have some people out there that apparently, well, you know where I'm going with all that. So, biocompatibility is one. Sterility and shelf life, reprocessing and cleaning, that's become a charged topic in the last several years with the whole duodenoscope fiasco and so on.
You mentioned earlier software and especially cybersecurity, electrical safety and electromagnetic compatibility, MRI compatibility, other non-clinical performance testing, that's sort of a catch all. That's just whatever doesn't fit in those above buckets.
Animal testing and implant device design considerations, all of them, with the exception of perhaps the very last one, have been on my list for a very long time. All of them are in other guidances in other places.
Okay, maybe FDA has consolidating them all into this guidance. Perhaps that might be a couple of other areas that are specifically pointed out in this guidance, Etienne, which is again, evolutionary, I'm sorry, evolutionary.
Etienne Nichols: Evidentiary. Yes.
Mike Drues: Thank you. Evidentiary. I spent a lot of time with lawyers, but I still have.
Etienne Nichols: That might have been the first time I've actually seen that word, but yes.
Mike Drues: Sadly, whatever it is, expectations for 510(k) implants. So human factor or usability testing, although I'm not sure when it comes to usability, how usability of an implant is any different than another medical device.
With one exception that usability. Well, I don't want to say there are no exceptions, but usually when we talk about usability for a device that's going to go inside of a patient, we're talking about the usability of the surgeon being able to implant it or install it into the body properly.
But there could be situations. I mentioned the hip implant problems and the breast implant problems and the vaginal mesh problems that we've had. There could be usability issues on the patient or the user side as well.
For some of these patients experience information, this is one of these newer sorts of politically correct phrases. As a matter of fact, I saw a few months ago, FDA put out a relatively new guidance in this general area of taking patient experience into account. Now, don't get me wrong, I think it's a good thing to do, but do we really need a guidance for that?
I mean, if we didn't have this politically correct, hypersensitive society, would FDA be putting out a guidance for that? I don't know. It's kind of know. Make sure know minority groups are represented in a clinical trial.
I mean, shouldn't that be common sense?
Your clinical trial has to be representative of your patient population, and if your patient population is going to include people in other minorities or whatever, then it should be in there.
And finally, labeling and other recommendations, like instructions for use and implant cards. This is important for traceability, or what I like to call trackability. If there's a problem with a device, like we talked earlier about recalls, especially if you have a device that's implanted in the patient's body, how do we know which of those patients has this particular device?
I'm involved with several medical device product liability cases where this is an issue in many of them, where in one case in particular, the largest product liability action in the history of the medical device industry, you can probably guess which one I'm referring to.
There's still some question as to whether or not all of the patients who have this particular device, it's not an implant, but whether or not all of them who have it or are using it have been notified.
Because we don't know to 100% certainty who has them and who doesn't. This is an even bigger problem when it comes to implantable devices. So, these are some of the things that are talked about again, Etienne, and in my opinion, nothing really new.
All important stuff, but nothing new to me. Most of it. Indeed, maybe all of it is, I would say, common sense.
Etienne Nichols: Yeah, one slight just question, or maybe correction. So, the patient experience information, I was reading a little about that because I was curious myself.
It said patient. Let's see here. Patients’ perspectives on living with implants are most useful. Patient experience. Here it is. Patient experience data includes patient preference information and patient reported outcomes.
I think that's referring to how they feel, function, and survive. Says as valid scientific evidence. I'm curious if there's any difference than customer feedback complaints.
Mike Drues: Yeah, great question, Etienne. And by the way, you mentioned the word feedback and especially complaints. Just as an FYI, I'm going to be doing a webinar, not a podcast, but a webinar for Greenlight, I think, in a week or two, specifically on complaints.
So, I'm looking forward to taking a much deeper dive into just that one topic. So maybe this podcast might be published after we do that webinar. But, yeah, I will be doing that webinar very soon.
In terms of this patient experience, let me answer your question in more quality terms, Etienne, and because I know you're a quality guy and you're very familiar with the design controls, this is a user need. Right? So, who's the user of a permanent implant? Well, one user is the surgeon, assuming that they have to put it in surgically, but another user is the patient.
Now the question becomes, let's not talk about the surgeon's user needs. Let's talk about the patients, because we're talking about the patient experience.
A user need for a patient, for a hip implant, might be something like, I need to be able to walk, or I need to be able to walk better without the same degree of pain and discomfort as I had before.
And at the same time, I don't want little nasty pieces of the implant to break off and float around the rest of the body and do all these other kind of things.
Etienne Nichols: Right?
Mike Drues: Again, you're laughing, Etienne, but I hope you and more our audience appreciates our not so, the use of humor. I mean, do we really need to define a user need as toxic particles not being generated from your device and causing all kinds of systemic issues?
Etienne Nichols: Right.
Mike Drues: Something to think about. But once again, coming back to the engineering, one could argue, personally, I would not use this argument myself, but one could argue that if it's not a user need, if it's not defined by the company in advance and it's not translated to a design input, then this is not an engineering failure because the engineer is supposed to make a device design a device that meets the design inputs.
If you don't have the proper design inputs because you haven't identified the proper design, sorry, the proper user needs. You see where I'm going with all this, Etienne? So again, if you understand what I call the regulatory logic, or just apply a little common sense, all of this should make sense.
Good regulation should always be based on common sense. And if it's not, either it needs to be changed or that regulation should not be there.
Etienne Nichols: Yeah. The more I talk to people and the more I hear about this, the dots really start to connect as a full lifecycle traceability situation. So, I mean, yes, you need to be thinking about things.
You get feedback from the field and need to inform your design. Your design should evolve. You should be increasing the patient experience all along.
Mike Drues: Those are very easy things to say in the QSR. Those are very easy things to put in a bullet list and a PowerPoint slide. But what does it mean in the real world?
That's where it gets interesting.
Etienne Nichols: So, I know we're at the top of the hour. Are there any last things you'd like to say? Especially, I'm curious if you have any thoughts on where the future of MedTech is going with this.
Are they signaling anything? Any future changes or should there be changes? Any advice?
Mike Drues: Well, first of all, should there be changes? Absolutely. And I think anybody that says, oh, no, we don't need any changes because everything that we do is perfect and is 100% as good as we possibly can.
Either they're naive or they're just flat out stupid. I mean, I don't know. So clearly things should be changed and improved, and we should have discussions like you, and I have been doing for a long time now about how to change things and make things better.
But my best advice is, yeah, you have to understand what the regulation says. Yeah, you have to understand what the guidance says and take the time to read these three new guidances.
As I said before, most of the stuff in there is you could just skim over. It's not important at all. There’re only a few parts in there that are really important.
But to FDA's credit, I can't say this as a matter of fact, but I would like to think that the reason why they're putting out these guidances. And as an FDA consultant, I know that there's going to be some other kinds of guidances along these lines coming out in the future.
I'll let FDA disclose them if and know we get to that time. But to FDA's credit, I would like to think, quite frankly, they're like me, getting sick and tired of waiting around, twiddling their thumbs for Congress to make some of these changes.
So, they're doing what they can. FDA cannot change the regulation. FDA can change and add guidance anytime they want, but they cannot change the CFR. That has to come with the Congress.
If that's the motivation for FDA for doing these things, then I'm 100% in support of that. Because while I think the system is a pretty good system, it's not a perfect system. While I think the. I've said this many, many times, I think the 510(k) is a pretty good pathway to market. It's not a perfect pathway to market.
And when the Institute of Medicine, the IOM, famously or infamously, came out seven, eight years ago, basically said to FDA, the entire 510(k) program should be thrown into the trash.
I was adamantly against that. That would be truly like throwing the baby out with the bathwater, and that would not be a good solution. It's a pretty good program, but there are things that we need to fix, or at least have discussions to try to make better, including some of the things we talked about today.
Interestingly enough, in the first guidance that we talked about today on the predicate selection, they don't even mention predicate creep in that particular guidance. In my opinion, that's the Achilles heel of the.
It's not even mentioned. Something to think about.
Etienne Nichols: Well, I know we're out of time. Maybe more to come on this. There's a lot more I know we could dig into. So, first of all, thank you so much, Mike. Really appreciate it. I'll include these different links to webinars and of course, the links to the guidance documents. And we will see you all next time. Everybody takes care.
Thank you so much for listening. If you enjoyed this episode, can I ask a special favor from you? Can you leave us a review on iTunes? I know most of us have never. Done that before, but if you're listening. On the phone, look at the iTunes app. Scroll down to the bottom where it says, leave a review. It's actually really easy. Same thing with computer. Just look for that. Leave a review button. This helps others find us, and it lets us know how we're doing. Also, I'd personally love to hear from you on LinkedIn. Reach out to me. I read and respond to every message because hearing your feedback is the only way I'm going to get better. Thanks again for listening and we'll see you next time.
About the Global Medical Device Podcast:
.png)
The Global Medical Device Podcast powered by Greenlight Guru is where today's brightest minds in the medical device industry go to get their most useful and actionable insider knowledge, direct from some of the world's leading medical device experts and companies.
