Why does FDA have three systems in place to address and handle risk for medical devices? Each system serves a different purpose, but are all three actually necessary?
In this episode of the Global Medical Device Podcast, Jon Speer talks to Mike Drues from Vascular Sciences about FDA's three systems of risk for medical devices and the unique interdependencies and distinctions between them.
Some highlights of this episode include:
- Three systems for medical device risk from FDA: product classification, significant vs. nonsignificant risk, software level of concern.
- The FDA’s classification system handles risk by classifying medical devices as Class I, II, or III. The higher the class, the higher the risk. The lower the class, the lower the risk. Yet, risk is a broad subject and there are a ton of exceptions.
- Also, classification numbers/levels used by the FDA in the United States do not translate in a linear way to those in the European Union (EU) and elsewhere. There are similar systems but different rules that are philosophically different.
- The significant vs. nonsignificant risk system focuses on clinical trials for medical devices. All clinical evaluations of investigational devices, unless exempt, must have an approved investigational device exemption (IDE).
- The software level of concern is important because it determines the level of documentation required for software development.
- The default classification for any new medical device is Class III. If it’s a new device, it’s not well-established and the benefits and risks are relatively unknown.
- When it comes to labeling, say anything as long as you can prove and support it. When technology stays the same but the labeling claim changes, risk changes.
- The determination of significant or nonsignificant risk is not made by the FDA or Institutional Review Board (IRB), but the medical device company.
- The software level of concern consists of Class a, b, and c. Does the software have no possibility of causing injury or damage to health? Is non-serious injury possible? Or, is serious injury or death possible?
Memorable quotes by Mike Drues:
“Theoretically, the classification of your medical device depends on risk. In other words, the higher the class, the higher the risk. The lower the class, the lower the risk.”
“The default classification for any new medical device is Class III, which basically means we set the bar at the highest possible level.”
“Oftentimes in regulation, we do things not because they make sense, but because that’s the way we’ve done them in the past.”
“The determination of significant or nonsignificant risk is not up to the FDA. It’s not even up to your IRB, or Institutional Review Board, it’s up to the company.”
Announcer: Welcome to the Global Medical Device Podcast, 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.
Jon Speer: On this episode of the Global Medical Device Podcast, I catch up with Mike Drues. Mike is with Vascular Sciences and a familiar voice and guest on the Global Medical Device Podcast, as well as numerous webinars that you can find in Greenlight Guru resources page. So check those out. But today, Mike and I talk about the three systems of risk that are in place and have been for a bit at FDA. Those are a classification system, the significant risk, non-significant risk approach, as well as software level of concern. And we explore how these things are connected, how they're also disconnected at the same time. So I hope you enjoy this episode of the Global Medical Device Podcast. Hello and welcome to the Global Medical Device Podcast. This is your host and founder at Greenlight Guru, Jon Speer. Joining me today is Mike Drues with Vascular Sciences. So Mike, welcome.
Mike Drues: Thank you, Jon. Always a pleasure to speak with you and your audience and nice to see you today.
Jon Speer: Yeah, good to see you too. I mean, even though we're half a country apart or probably two thirds of a country apart, it's good to be able to have a little bit of FaceTime with one another.
Mike Drues: Absolutely.
Jon Speer: So today... You and I were chatting the other day and we talked a little bit about the systems of risk, I guess, if you will, that are in place and have been for a bit within FDA. And I guess to kind of start the conversation, those systems of risk would be how a device is classified, the significant risk, non-significant risk, and then if you have a software, then it would be a software level of concern. So Mike, maybe take a little bit of a moment and kind of set the stage and talk a little bit about how risk is handled from an FDA perspective, of course, elaborating a bit on each of those different systems.
Mike Drues: Yeah. Great question, Jon, and as always, thanks for the opportunity to talk about this very important topic with you and your audience. Obviously, you and I are both into risk in many ways. And just as a reminder for our audience, I happen to be a subject matter expert for FDA at a few different areas, one of them being risks. And why I find it interesting, Jon, is that a lot of people don't realize that actually here in the United States, we have three completely separate and distinct systems for dealing with risk. The first system that you just mentioned, the classification system most of our audience is probably familiar with, class one, class two, class three, theoretically, the classification of your medical device depends on risk, but in other words, the higher the class, the higher risk, the lower the class, the lower the risk. Well, I say that theoretically, Jon, because that's what every regulatory textbook says, and that's what virtually every regulatory professional says. But there's two caveats to that. First is risk is a very, very broad subject as you and I have talked about before, that's point number one, and point number two, there are a ton of exceptions, a ton of exceptions. And for those in the audience that might be interested, Jon, just a few months ago, I did a webinar for Greenlight on classification, so for those that are interested, take a look at that webinar because I take a deep dive into that.
Jon Speer: Yeah. And we'll provide a link to that webinar for the show notes that are accompanying this. So absolutely, it was fantastic. I love that webinar its one of my favorites.
Mike Drues: Thank you, I appreciate that Jon, very kind of you to say. So the first system that we use to deal with risk is the classification system, class one, class two, class three... By the way, one other quick point to note, not to go off on too many different tangents here, but we're limiting our discussion here to the United States because in other places in the world we use what seems to be a similar classification system. In other words, a numerical classification system in EU or Japan or Canada or whatever, but suffice it to say that those classification numbers or levels do not transponder in a linear fashion. In other words,-
Jon Speer: Yeah. So like class two in US may or may not translate to a class 2A or 2B in the EU they're just similar systems, but different rules, if you will, to determine that classification.
Mike Drues: Well, the way I like to describe it, Jon, and again, this is a topic for a completely different discussion, but the classification system here in the US and the classification in the EU or elsewhere in the world, although we still use numericals class one, class two, and so on, they are philosophically different from one another. And there's no easy conversion factor to go from one to another, like from Fahrenheit to centigrade or something like that. So anyway, the first classification system that we use here in the US is class one, class two, class three, the second, as you just mentioned, Jon, is the significant versus non-significant risk system and we can talk more about that in a moment, but that's primarily important when we get into clinical trials. What medical devices are what we call IDE exempt, meaning that you do not need investigational device exemption, you do not need to go to the FDA before beginning a clinical trial, although inaudible always do anyway, versus a significant risk device where an IDE would be required and you would need to go to the FDA before you do a clinical trial. And again, I want to emphasize this point quite strongly, Jon, cause I see a lot of people, they don't get this, there is no simple conversion factor between the class one, class two, class three system versus the significant and non-significant risk system. It's not like converting from ounces to grams or Fahrenheit to Celsius. Just because something is class one, class two or class three, doesn't imply anything about whether it's significant or non-significant risk and vice versa. And then the third system that we use for risk here in the US as you mentioned, Jon, is the software classification system, the software level of concern, if you will. And the reason why that is important in a pragmatic sense is because that will determine what level of documentation that we're going to need in our software development and that type of thing. So those are the three systems that we use for handling risk here in the US, Jon, how would you like to further peel back this onion?
Jon Speer: Well, I mean, it is really interesting that to the lay person, you would think that there might be a correlation of relationship between the classification, the SR/ NSR, significant risk, non-significant risk or software level of concern. Obviously they had sort of different origins, they've evolved over time, at different points in time, but do you think that each of these has importance as far as the system of risks that, I mean, why do we need three different methods? Why can't we just have one?
Mike Drues: Well, indeed, that's a good question, Jon and I think you probably realize who you're asking that question to because risk is obviously a very important topic. On one hand, I think it's a topic that is tremendously misunderstood. The number of people that have a gross oversimplified view of the topic of risk is just amazing to me in this industry. And yet, on the other hand, as you just pointed out, Jon, we have these three systems of risks so it kind of begs the question, are we really, in certain ways, making this a heck of a lot more complicated than it needs to be? And I'll be honest with you, Jon, as an SME for FDA in risk, I think with regard to that latter question, we are definitely making it more complicated than it needs to be. I don't want to go so far as to say that these three systems are totally unrelated to one another, but what I like to say, Jon, is they're independent, but they're also inter- dependent on one another. In other words, clearly there's a relationship, but it's not a simple relationship. As I said before, it's not just a matter of converting from one form of temperature scale to another form of temperature scale. There's a lot more to it than that, but maybe it would help, Jon, if we go into these systems perhaps a little bit more detailed and that might help illustrate.
Jon Speer: Sure. So let's dive into the classification side of things a little bit further and I'll... A recent conversation I've had with someone is they have a novel, unique product or device. If you look at the classification system, there is no regulatory regulation for this unique novel product. There is no product code. So by default, that would make it a class three product, which class three would imply a highest risk from a classification point of view. However, when you peel back this device a little bit more it is not high risk, it's super low risk. So that's where it's like," What? What's going on?" But obviously, we get into deNovo and things like that, where, deNovo is sort of a way I think that one can use on this novel unique approach from a product perspective as a way to say," Hey, yeah, I get it. It's class three, but really it should be treated from a risk perspective, more like a class two". So maybe that's a good way to start to peel back some of the nuances of classification.
Mike Drues: Sure. So obviously Jon, you point out something quite correct and that is the default classification for any new medical device is class three, which basically means we set the bar at the highest possible level. And when you think about it, Jon, that should make perfect sense I hope, to everybody working in this industry, because if it's a new device, that means it's not obviously well- established, we don't know what the risks and benefits are and so on. So it makes sense to set it at the high level, right, as we can. But I'll use the example that I used in my deNovo webinar, Jon, that I did for Greenlight, I think a few years ago now. Let's consider a band- aid. Today, band- aids are ubiquitous, but back in the day, there were no band- aids. So if we were bringing a band-aid onto the market for a very, very first time, there were no band- aids on the market before, what class would that band- aid be Jon?
Jon Speer: Well, it would be class three.
Mike Drues: It wouldn't be class three. Absolutely. But it should not take an MD or a PhD or an RAC after somebody's name to appreciate that gee, maybe it doesn't make a lot of sense to treat a band- aid as other class three devices like an artificial heart. And so that's the essence of the deNovo, Jon, we go into the FDA and we say it doesn't make sense to treat our bandaid as an artificial heart. It should be class two and here's why, or class one and here's why. So you're right that the default classification for any new medical device, there are no exceptions, is a class three. The example that I like to use, Jon, in addition to the band- aid, what I said earlier, that the class one class, two class, three system here in the US is theoretically based on risk, but there are a ton of exceptions. Let me give you one, just one of my favorite exceptions. The class two universe is a very, very broad universe. In fact, about 55% of all medical devices fall into this layer of class two. In there, we find things liked powered wheelchairs and imaging systems and so on, but we also... And we find hospital beds, believe it or not, Jon, a hospital bed is a class two medical device, but we also in that same class two universe, we find pacing leads. Now please explain to me, if classification is simply based on risk as so many people think, how do we put a pacing lead, something that goes in your heart, probably for the rest of your life and a hospital bed, something that you lie in for a few hours or a few days in the same class? There is no logic, certainly no logic based on engineering or biology to explain that. The only regulatory logic and in my opinion, Jon, it's very, very weak regulatory logic. The only regulatory logic to explain it is that oftentimes in regulation we do things not because they make sense, but because that's the way we've done them in the past and that's how pacing leads were regulated in the past prior to 1976 and they were grandfathered in. So the point that I'm trying to make, Jon, is very simple. Yes, the classification system class one, class two, class three is based on risk. That is a factually correct statement. However, it is a grossly oversimplified statement of the classification system. Does that make sense, Jon?
Jon Speer: It does. Yeah, for sure. And I guess to kind of take your band- aid example and moving into like significant risks, non-significant risk. I mean, okay, brand new, again we're going back to imagine it doesn't exist, a brand new class three from a risk perspective, from a classification, but clearly from a significant risk non-significant risk perspective from a clinical, like if we wanted to do some sort of clinical trial or something of that nature, knowing how it's used is like, okay, how would you determine the risks there associated with the bandage that had never been used before?
Mike Drues: So you know what, Jon, let's take that band-aid example a step further, right? You know, Jon, we've talked about many times I don't want to be the regulatory police. I never want to tell a company what they cannot do. Instead, I focus on telling them what they can do and there's no better example of that than when it comes to labeling. So oftentimes, a regulatory person will say to a company," Oh, you can't say this, you can't say that", I say to a company, no, no, no, you can say anything that you want as long as you can prove it. So if you want to make the claim that your new medical device cures cancer and regrows missing limbs, wonderful. I say no problem. But the first thing that you need to do to is to convince me, nevermind as a regulatory consultant, Jon, but as a biomedical engineer, how do you prove that? How do you support that? So coming back to the bandaid, let's say, hypothetically speaking that we kick up the claims on the bandaid. Let's say that I want to bring a bandaid onto the market for an inaudible of cancer. In other words, put this bandaid on and it will treat your cancer in some way. Well, if I were to do that, I pretty much can just about guarantee that your band- aid will now likely become a class three device because as a general rule, and again, I want the audience to understand this, Jon, the technology of the band- aid itself is the same, but we've just changed the claim, the labeling. In this case, treating cancer is a general rule. Anytime in CDRH, you mention the C- word cancer, now you're almost always in the class three universe because of risk. The device itself didn't change but what we said about the device did, does that make sense, Jon?
Jon Speer: It does. And I guess the cancer curing mandate, if we wanted to do a clinical investigation on this, is that significant risk or not significant?
Mike Drues: That's an excellent question. So once again, let's run with the metaphor here. So let's talk a tiny bit about the significant risk versus non-significant risk system. First of all, the determination of significant or non-significant risk is not up to the FDA, it's not even up to your IRB, your institutional review board, it's up to the company-
Jon Speer: Right crosstalk there, Mike. I think a lot of people get tripped up on this so folks tune in pay attention right now, crosstalk
Mike Drues: Well, thank you, Jon. And by the way, we did a podcast some time ago where we did a much deeper dive into significant and non-significant risks. So maybe we can put that as a reference for the audience as well. But as I started to say, the determination is not up to the FDA or the IRB, determination is up to the company. Is that like putting the fox in charge of the henhouse? Absolutely. No question about it, but that's exactly what the regulation says. So using that band- aid example with the kicked up labeling of the cancer claim, now, most likely, you would turn that bandaid, which before was, I'm sure, a non-significant risk device to a significant risk device, which means that you would likely have to go to the FDA for an IDE before you begin your clinical trial. But let me make another point about the strategy of the significant versus non-significant risk device. As I said, Jon, earlier, if you have any device that is a legitimate non-significant risk NSR device, you have no obligation to go to the FDA before you begin your clinical trial. In spite of that, I almost always advise my customers go to the FDA anyway, usually in the form of a pre- sub, even though you're not required to. Any speculation as to why you think I give that advice, Jon? Why would I encourage a company to go to the FDA to bring them their clinical trial when they're not required to do so?
Jon Speer: Well, I think there's a couple of things from my perspective. And you mentioned is this the fox guarding the henhouse? I mean, sort of, however, I think it's good to have... Go into the FDA in the form of a pre- submission it's a communication means. It's like," Hey, Mr or Mrs. FDA person. I'm doing this product where we've gone through a decision tree, we've determined it NSR, here's the rationale as to why, we're planning to do a clinical study with this number of patients, it's a great way to communicate. That way, FDA has that opportunity to say," Whoa, wait a minute, Med Device company, explain further what you mean by this", or," Hey, maybe we're not so sure" but I just think that's a great way to communicate.
Mike Drues: I agree with you, Jon. And that in fact is the first problem that I'm trying to avoid, which is what you've just described. And that is make sure that the FDA sees it the same way that I do, specifically that it's non-significant risk. In other words, it is possible and I've seen this happen before, where the company determines that their device is NSR, they get an IRB to agree that this is NSR, they do the clinical trial, they submit to the FDA and FDA says, gee, we don't know what you are smoking, you must be smoking your socks. We don't see this as NSR. And now you're in a whole set of problems, right? So that's problem number one. The second problem, which is even more common, and I see this happen to some of the largest medical device companies on earth and when this happens, I just laugh. It's just such an elementary school mistake. A company does a clinical trial for an NSR device. They don't take it to the FDA first, because they're not required to. They collect a certain amount of information. They submit that to the FDA and then the FDA turns around and says, gee, we would like to see these additional pieces of information, these additional clinical endpoints. And now the company has to do a whole clinical trial over again to collect this information, Jon, can you say ka- ching, ka- ching, ka- ching? It is amazing to me how frequently this happens. And I just laugh because it is such an elementary school mistake that you can greatly mitigate, if not completely eliminate by, as you just suggested, Jon, communicating with the FDA in advance. Even though you're not required to, in most cases, when I'm working on an NSR device, I still do it anyway in the form of a pre- sub.
Jon Speer: It's so smart. I mean, a pre- sub... I mean, it's not free as far as the time, effort and energy that goes into it. But I mean, in the grand scheme of things, it's minimal, especially considering the negative ramifications. If you didn't crosstalk you have to go back and do a new study. I mean, that's like you said, ka- ching, ka- ching, ka- ching.
Mike Drues: Correct. Well, you and I have talked about pre- subs many, many times Jon, and even though there are some limitations and some challenges, I'm a huge fan of the pre- sub process and you said that there's no cost associated with it. You're right. There's no cost in terms of an FDA user fee although give Congress a little time, I'm sure they'll add one. But the way I look at it, Jon, is it's cheap insurance. In other words, how delays and problems can you avoid later on by taking it to the FDA in advance? So the significant versus nonsignificant risk determination is one of many reasons why I use a pre- sub. I know a lot of people in our industry, Jon, like to have templates. I put together an NSR versus SR template that I use with my customers. It's pretty easy to do, so if people have questions on that particular determination, I'd be happy to get into it and more specifics with it.
Jon Speer: Okay. So for sure. And actually that's a great time to take a brief pause. Want to remind folks I'm talking with Mike Drues. Mike is with Vascular Sciences. When it comes to creative regulatory, yeah, I know it seems weird to maybe to use those two words in the same sentence, creative regulatory strategy, Mike is your guy. He understands how to navigate and play this poker game I would say probably better than anyone, frankly, that I've come across and he thinks about this way differently than a traditional regulatory approach, which I think is super refreshing. So Mike's a great guy to have in your corner, speaks a lot about these, he's mentioned a couple of the webinars and podcasts that we've done. So you can go to Greenlight resources page and find all sorts of different webinars and podcasts and things of that nature. He also writes and does podcasts at other places too. But if you just search for Mike Drues Vascular Sciences, you'll get a whole library of really exciting content to consume. Talk a little bit about Greenlight Guru, www.greenlight.guru, we are the only medical device success platform on the market today designed specifically and only for the medical device industry by actual medical device professionals. So that's really exciting. And great news, we just launched the Greenlight Guru Academy. Basically, it's a way to take your interest and desire to learn more about certain topics that are facing you as a medical device professional and just go a little bit further with that. We've got all sorts of free courses that are available today. Just go to www.greenlight.guru/ academy and sign up for some of our courses. And we'd love to get your feedback on other courses that you might like to see in the Academy going forward. So, all right, Mike, so the last system getting back into the conversation, the last system that we can peel back and dive a little bit deeper is the software level of concern. I mean, I'm trying to imagine, or I was trying to think about how to take the bandaid, the beefed up bandaid example, maybe, a little bit further but I suppose you could. You could put a little chip on it and it could... Sending data to my smartphone about surface temperature or whatever the case may be. So how does software level of concern? I mean, what are the pros and cons, the good, the bad about this as a different system of risk?
Mike Drues: Great question, Jon. So the software level of concern or what some people refer to as the software classification system, which is instead of using the vernacular class one, class two, class three, we call it class A, class B, and class C. There are three questions. If you will, that go into helping to determine your software level concern. And by the way, for the... Back to the SR versus NSR discussion for a moment, there are four questions that will determine whether your device is significant risk or nonsignificant risk, in software there are three. Class A is basically if your software has no possibility, no chance of causing injury or damage at all, no possibility whatsoever. As you can imagine, Jon, that is a very hard place, quite high place to set the bar because to use the metaphor example, you know I have this pen in my hand, you could put an eye out with this pen, right? So if you're going to make a Class A determination, be, as Elmer Fudd said in a Bugs Bunny cartoon, be very, very careful because the regulation is written either purposely or not. And sort of very absolute concerns. No injury or damage to health is possible. That's a very high place to set the bar. Class B is when we have non serious injury is possible. Again, we get now get into that very fuzzy area of what is serious versus what is non serious and so on and so on. And finally, we get to class C where either serious injury or death is possible. So theoretically, just like we talked about earlier with class one, class two, class three, the higher, the risk, the higher the class, similarly class, A, B and C for software, the higher, the risk, the higher the software class, but just like in the class one, two, three world, it's not nearly so simple. As I said earlier, Jon, and I would love to hear your thoughts on this unlike classification in the one, two, three sense of the word, which will determine among other things, how much regulation are you subject to, whether it's just general controls or general plus specific controls or beyond that for class three, in the software world class A, class B, class C. What I tell my customers is it primarily dictates the level of documentation that you need, especially on the quality side of your development. Not so much on the regulatory, but on the quality side. But it doesn't really, I don't want to say it doesn't impact, but it doesn't tremendously impact the type of testing that you have to do or the degree of testing, because that's still going to be determined largely by the numerical class of your device. Would you, I don't know if I'm explaining that very well, Jon, would you agree with that explanation or how do you think we could explain it to audience so it's a little more clear?
Jon Speer: Yeah. And I think this is, at least from my perspective, this is where the rubber sometimes hits the road is that I think the... Well, I don't want to over- generalize here, but a lot of companies would say," Oh, well, if I got to pick, I got to choose how my product would be perceived from a risk perspective, I want class one, I want NSR and I want, if it has software, I want a class A, because those are lowest risk and I think sometimes the perception is lowest risk means easy, easier to do, less documentation and that's where thing where I'm like, I'll get on a soapbox for a moment. The thing that drives me crazy is somebody is like," Oh, I'm class one. It means I don't have to do design controls" and I'm like," I mean, okay, sure. From a pure black and white literal interpretation of regulations, I see why you say that however design controls are about demonstrating your product's safe and effective and that it works. Why wouldn't you want to do that?"
Mike Drues: You're not going to use my phrase, Jon, prudent engineering?
Jon Speer: Prudent engineering! But I think this is the game that sometimes people get that they see it as a game, I think they get trapped. Because there's this perception," Oh, if I'm this or that, then that means I have to do more work or less work" and I think that's a slippery slope.
Mike Drues: I agree with you, Jon, and not to go too far into design controls because you and I have talked about that many times before as well, but let me give a quick example of how there is no simple conversion factor between say class one, class two, class three versus a class A, B, and C for software. So it is definitely possible. And I can think of one or two devices off of the top of my head that are class one exempt, which basically means they're still regulated by the FDA, but they're regulated at the lowest possible level. The only thing lower than that, if for those in the audience that have watched either my webinar on classification or my webinar on pathways to market, I present my medical device pyramid, where we have class three at the top of the pyramid class two below that class one below that and at the bottom of the pyramid, Jon, do you remember what's below class one?
Jon Speer: I do. It's class zero, which is an unofficial designation, but it makes a lot of sense because there's a lot of products that are ambiguous. They are medical devices in some respect, but not necessarily from a classification.
Mike Drues: Correct. So it's what I call class zero. FDA does not use that terminology, but it's what I call class zero or better known as the wellness devices. So going back to the example that I just was going to share a moment ago, we can have a class one exempt device, which basically means that it's regulated at the lowest level by the FDA. The only thing less than that as a wellness device, which is not regulated at all, but that class one exempt device, if it's an electronic device that has software, it could have software that's regulated as a class C in software. In other words, the highest risk of the software and yet the lowest risk of the device and the explanation, and again, this is a topic I think of a much more advanced discussion, Jon, but the explanation at a certain level is fairly simple. It could be that the class one exempt device is indicated for a very, very simple condition, a very benign, class one exempt level condition. However, if the software fails, the level of risk could cause serious harm or possibly even death to the patient. So it's a perfect example of, at least in my view, Jon, there is no simple conversion factor between the two, many people would assume that if a device is class one exempt, or maybe even class two, that it must be a non-significant risk device. It doesn't necessarily have to be, it could be a significant risk device. And they would assume that it must be, on the software side, class A or possibly class B and definitely not class C, but you know what they say about making assumptions, Jon. So just be, as Elmer Fudd said, just be very, very careful. And speaking of templates, I do have a template that I give to my customers that help with the software classification determination as well.
Jon Speer: Yeah. And so maybe a good way to... Getting back to the why... Why are there three systems. I guess my way of interpreting it is each of the system has a different purpose when it comes down to it. With the classification is more about," Okay, this dictates or is a determining factor in what I need to prepare and submit to FDA in order to make my case for getting clearance or approval or whatever the case may be. Whereas the NSR/ SR, as we've talked about is really more about pre- approval or pre- market clinical use, I guess you could even do post- market even as well but about risk from a clinical perspective, whereas software level of concern obviously its, I don't know, I guess it's its own beast a little bit too, just from," what do I need to do as a software developer from that perspective", but it is confusing because I think a lot of times I, as Med Device professional, I want to lump all these things into one bucket, but they're really three separate buckets.
Mike Drues: Well, I would agree with you, Jon, that each of these systems that we're talking about today, they do serve a different purpose. They do approach risk from a slightly different perspective, but here's my... Instead of asking the question," Do we need these three systems?" because in my opinion, we don't, I think it's definitely possible to integrate all three into one. However, I think the classification system that we have, the class one, class two, class three, is as I think I said earlier, a gross oversimplification. In other words, I don't know if you would agree or disagree with me, Jon, but I think today, especially in 2021, for us only to have three levels of classification is just nuts. Where you use the three buckets and I'm a big fan of the three buckets, but those three buckets are way too big. I think we need to have, maybe I don't want to have a hundred buckets, what maybe six smaller buckets, than three big buckets. And in a sense, Jon, you mentioned this- crosstalk
Jon Speer: I mean, if you start to look at classification and you've got three choices and you got NSR/SR, you have two choices. I'm not a math crosstalk but I think this is right, so I've got two choices there. And then from software level of concern, if that's applicable, I have three.
Mike Drues: Three. So you got to do the multiplication, Jon, you've got to do a lot of different perturbations there, but one last thing to mention, and then I think we can start to wrap this up. You mentioned very quickly at the beginning of our discussion, FDA protocols and protocols are in a sense, how do you want to say, the classification system on steroids. In other words, for those that are not familiar with FDA protocol, the way I like to describe it is imagine the entire medical device universe is a big pie and in that pie, you have wheelchairs and band- aids and EKG monitors all the way through to totally implantable artificial hearts. And it doesn't make sense to treat them all the same. So what we do is we slice that pie up into hundreds and hundreds of different, smaller slices. And each of those slices will contain devices that are similar to one another. All the wheelchairs will be together, all the artificial hearts we'll be together, all the different kinds of catheters will be together and we assign each one of those a product code. So for classification, I'm not suggesting that we go quite that far in terms of having the number of product codes, but I just think there's, we're oversimplifying this risk system and the classification system. So just throwing out as a topic of discussion, instead of having three classes, class one, class two, class three, we may have five or six and we lump into it the SR versus NSR system, such that for example, any devices that are, again, I'm going to just make up a number here are, say class three or less, you will be IDE exempt and any devices that are class four or more, you will be IDE non-exempt, and depending on the class, that will determine your software level of concern as well. The point is I think that these can be consolidated and much more importantly, I think we can define them better in terms of engineering and biology, as opposed to regulatory and quality terms, which I think is what the emphasis should be here, Jon.
Jon Speer: Yeah. I guess what I'm hearing you say is today, as we talked more towards the beginning of the conversation, is they're not completely autonomous, the classification, the NSR/SR, and software level of concern, they're not completely autonomous but they're certainly not synchronous either.
Mike Drues: They're independent, but they're also interdependent on one another.
Jon Speer: So what I'm hearing you say is, the question is, is there a way to make them dependent and analogous and, and attributes really of one another rather than completely separate systems,
Mike Drues: It certainly could be done. It's definitely worthy of discussion. And I know, Jon, we have lots and lots of listeners that listen in to our podcasts, including many of my friends at FDA, although I'm sure they would never admit to that publicly, but I know that they're listening. So maybe they'll take that into account and they'll think, is there a way that we can simplify, but more importantly, make our approach to risk here in the United States more inaudible as I think there's a lot of areas of improvement that can be made here.
Jon Speer: For sure. All right. Any other final thoughts on the topic of systems of risk that FDA uses before we call it a wrap today?
Mike Drues: Yeah. So just to wrap up, be aware of the different systems of risk. Obviously, most people are already aware of class one, class two, class three, but that's not the only one. Be aware of the significant versus non-significant risk determination, especially if your device is going to use clinical data and more and more devices are requiring clinical data. And finally, if your device contains software, obviously be aware of the software classification system and bring these things into your discussion as early in the product development cycle as possible, and the last thing that I'll recommend as Jon pointed out earlier, and I could not reiterate this strong enough, whatever you're doing, whether it comes to significant versus non-significant risk or something else, take it to the FDA in advance of your submission. The first time you should be talking to the FDA should not be at the point of your submission. You should be talking to them before. In spite of the delays now happening with COVID, that's a completely different story, take your product to the FDA, it can save you so much time and headache. We've done many podcasts, I've done a webinar for Greenlight specifically on the pre- sub process, I've got a ton of pre- sub experience, I'm happy to help. So those resources are all out there available to you. You can lead a horse to water, Jon, but you can't make them drink. So it's just cheap insurance.
Jon Speer: It totally is. I mean, and Mike, I appreciate all of your insights on this topic and more. Folks, if you've been listening to the Global Medical Device Podcast for any period of time, you've heard the sage wisdom and advice from Mike Drues. And I'll repeat what I said a moment ago, he's the best in the business when it comes to creative regulatory strategy and being able to navigate all these twists and turns. So definitely a guy that you want in your corner. Reach out to him, he'd be happy to have a conversation with you. And again, Greenlight Guru, we're also here to help with our medical device success platform, help you manage design controls and risk and document management and all of those quality events that may, hopefully never, but may come up at some point in time during your product life cycle journey. Go to www.greenlight.guru to learn more about the medical device success platform. And as always, thank you so much for being loyal listeners of the Global Medical Device Podcast. The number one podcast in the medical device industry today. Continue to spread the word to your friends and family and colleagues and so on and so forth. As always, this is your host and founder at Greenlight Guru, Jon Speer, and you have been listening to the Global Medical Device Podcast.
ABOUT THE GLOBAL MEDICAL DEVICE PODCAST:
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.