- Guru Difference
- Customer Experience
The 510(k) is the workhorse of regulatory submissions to the FDA from the medical device industry in the United States. However, 510(k)s continue to be the source of problems, delays, and rejections. It only takes common sense and simple solutions to overcome these issues.
In this episode of the Global Medical Device Podcast, Jon Speer talks to Mike Drues of Vascular Sciences about common 510(k) mistakes and how to avoid them. They also cover what has changed (if anything) in the past five years since Jon wrote the article, 7 Common Mistakes That Can Delay your FDA 510(k) Clearance.
Mistake #1 - Inconsistency with Documentation: The intended use statement often is inconsistently put into other sections of the 510(k). It’s important to understand the difference between indications for and intended use.
Mistake #2 - Not Using Checklist: The FDA’s 510(k) submission guidance provides a Refuse to Accept (RTA) checklist. Help yourself and reviewers out. The FDA states what it is going to do and what it wants, so there should be no surprises.
Mistake #3 - Not Providing All Expected Testing: The FDA rarely accepts promissory commitments to do sterilization, validation, or biocompatibility testing later. Also, you should include protocols, results, and reports to verify rigorousness of testing.
Mistake #4 - Not Doing Real-time Shelf-life Studies: FDA wants evidence to back up claims of stated shelf life and wants a correlation between real-time and accelerated aging data. Shelf life is what happens to your product in the future.
Mistake #5 - Underestimating Risk Management Requirements: Show list of requirements and provide the paper trail for your traceability. The FDA needs to conclude that your product is safe and substantially equivalent to a predicate.
Mistake #6 - Not Demonstrating Equivalence with Predicates: You need to depict your device as novel and unique, but not so different that the FDA thinks your predicate is not a good example.
Mistake #7 - Formatting Submission Incorrectly: Include page numbers and an eCopy. Also, the size of the file matters and needs to be named according to conventions.
“The intended use statement is a pretty important premise for any 510(k) submission because for all intents and purposes, that statement helps regulators and helps you determine how your product is ultimately going to be classified.” Jon Speer
“One of the most common reasons why they are rejected is because the high-level labeling when it is repeated is not repeated exactly the same way.” Mike Drues
“Intended use focuses on the device. Whereas, indications for use focuses on the patient.” Mike Drues
“The RTA checklist was intended to be a tool for the FDA reviewer, when submission received, for them to go through and make sure all the constituent parts and components and what is expected to be in that submission is included.” Jon Speer
“FDA reviewers see probably dozens, maybe hundreds of submissions in a finite period of time. Some are better than others. Well, make yours standout. Make yours easier to read.” Jon Speer
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: Hello and welcome to the Global Medical Device Podcast. This is the founder at Greenlight Guru, Jon Speer. And joining me today is Mike Drues. Mike is the President of Vascular Sciences. And if you've listened to any global medical device podcast episodes over the past, gosh, I don't know, it's been a lot of years now, there's a good chance you've heard Mike and I chat about a variety of topics ranging from quality to regulatory to CAPA and a litany of other things. But Mike, welcome back to the podcast.
Mike Drues: Thank you, Jon. Always a pleasure to speak with you and your audience.
Jon Speer: Yeah. For those listening, they may not know, you and I kind of riff back and forth in between the times that we chat and try to find topics that are somewhat evergreen or at least timely and relevant. And one topic that I think is, well, it's been timely and relevant for a long, long time and probably timely and relevant for at least the long foreseeable future, is regulatory submissions. And specifically, here in the US, the workhorse, I think, as you've referred to it before of regulatory submissions from FDA, from med device is 510(k). So, maybe we could talk a little bit more about that today.
Mike Drues: Yeah, I think, Jon, that's a great idea. And as you said, the 510( k) is without a doubt the workhorse in the United States of the medical device industry. And in spite of that or perhaps because of that, it continues to be a source of problems, a source of delays, the source of rejections with a lot of companies in this industry. And one of the most common questions I get from my customers and from people across the board is how do we avoid these common problems? And so, I came across this column Jon written about five years ago from this really oddball character named Jon Speer. I don't know. I need to meet this guy called Seven Common Mistakes That Sink 510( k) Clearances. And I thought what we could do today in today's discussion is kind of go through the seven common mistakes as you've identified them and talk about if anything has changed in the last five years. And more importantly, how can people listening today avoid these common mistakes? Does that make sense, Jon?
Jon Speer: Yeah, absolutely. I'm glad you blew the dust off that article, and I hope, I guess we'll find out here in a few moments. But I hope this article from five years ago is obsolete and irrelevant today, but my suspicion is it's probably not.
Mike Drues: I would love to agree with you, Jon. But unfortunately, that might be true in the land of rainbows and unicorns, but it's not true in the real world where I work.
Jon Speer: Yeah. All right.
Mike Drues: So, let's dig right into it and start out with the first of the seven common mistakes that you identified, and that is inconsistency with documentation. Inconsistency with documentation. And let me start out with a direct quote from your column, Jon. And that is, " The intended use statement often is screwed up," that's your words, screwed up, " by incomplete application into other sections of the 510( k)." I'm pretty sure I know what you mean by that, Jon, but for the benefit of the audience, why don't you explain what you mean by that?
Jon Speer: Sure. And folks, we'll be sure to provide a link to that article from five years ago, so that you can read along after the fact. But nonetheless, so the intended use statement is a pretty important, I think, premise for any 510( k) submission because for all intents and purposes, that statement helps regulators and helps you determine how your product is ultimately going to be classified, what you can claim, and all these sorts of things about the product. But that statement is, I don't remember exactly, you probably know exactly, but I know it is stated several times throughout a 510( k) submission. My memory recalls there's at least a handful of times where throughout a 510(k0 where that statement is stated. I believe there is one section that is completely dedicated to the intended use, and there is a form I believe that one fills out to identify what that is. But there's several other places, other sections within a 510( k) where this statement is also listed as well. What I mean by screwed up is, sadly, I've seen this a lot. I suspect you probably have a time or two as well, where that statement is not the same. It is not consistent throughout one's own 510( k) submission. It's stated one way in one place and maybe slightly differently in another place and so on and so forth. I got to believe that as that submission is sent to FDA for review, especially during the refuse to accept 14- day window, that's the easiest thing for a reviewer to say, " This is not the same throughout, so reject, send it back," which it makes sense. If I'm a reviewer, I want that statement to be the same verbatim word for word throughout that 510( k). And if it's not, that says to me as a reviewer, maybe these people don't know what they're doing.
Mike Drues: I agree with you 100%, Jon. In fact, one of the most common reasons why submissions, not just 510( k) submissions, but submissions across the board, de novos, PMA and so on, are rejected is because of what you just said. Because the labeling and what we're talking about here is what I call the high level labeling. The intended use of the indications for use are when they are repeated, you're exactly right, in all regulatory submissions, the high level labeling is repeated several times. And by the way, one of my frustrations with so many submissions being so long is because the same information is repeated over and over and over again. But that's a topic of a different discussion. But one of the most common reasons why they're rejected is because the high level labeling when it's repeated is not repeated exactly the same way. And as we talked about before, when FDA is doing their job, and let's be honest, I know we're on a recorded podcast here, but FDA doesn't always do their job. They will scrutinize the high- level labeling word by word, character by character, punctuation by punctuation mark. Other parts of this submission, I know for a fact are never read by anybody in some cases, but the high level labeling is very much an exception. So, you really need to make sure that you're consistent. And I know there are a lot of tools that will help companies prepare their submissions and make sure that, that doesn't happen, but it still happens anyway. And here's a simple piece of advice. Most people today have Microsoft Office or something the equivalent of that. I'm a big fan of using, not just simply copy and paste, but copy and paste link. Copy and paste link is a feature in Microsoft Office that you create your high level labeling in one document instead of simply copying and pasting it into your submission or anywhere else, you do a copy and paste link. In that way not only is the information exactly the same every time, but if you want to change it, you only have to change it one time and that is automatically changed in all of the other documents. So, there are very simple solutions. We're not talking about rocket science here, Jon. They're very simple solutions to a lot of this. So again, one of the most common mistakes that people make is when they repeat their labeling, it's not repeated exactly the same.
Jon Speer: Yeah, absolutely.
Mike Drues: Another thing that you brought up in that same section, Jon, is a common mistake is confusion between statements that are more about indicated use than intended use. And this is something that you and I have talked about before, the difference between indications for use and intended use. So maybe, we should spend a minute or two and just remind our audience the difference between those two statements.
Jon Speer: Yeah, absolutely. I know you have a very clear and concise explanation on the difference between indications for use and intended use, so I can't say it better than you. So, if you don't mind reminding folks how to, I mean, there's obviously a lot of similarities, but there are some important distinctions between those two statements as well.
Mike Drues: Yeah. Well first of all, Jon, thank you for those kind words.
Jon Speer: Yeah, you're welcome.
Mike Drues: I'm happy to recap what I see is the difference, and I'll just remind our audience that you and I have talked about this topic in detail in other podcasts and even in some of my webinars, so maybe we can refer to some of those for our audience. But in a nutshell, here's the big difference between intended use and indications for use. Intended use focuses on the device, whereas indications for use focuses on the patient. So, taking that one step further, when I say intended use focuses on the device, that means what does the device do? How does it work? It's mechanism of action, and so on and so on. Whereas, when I say indications for use focuses on the patient, what I mean by that is what disease injury or condition is the device intended to prevent, diagnose, or treat. So, intended use focuses on the device, what it does, how it works. Indications for use focuses on the patient. What disease, injury or condition is the device intended to prevent, diagnose, or treat? Now, taking that just a step further, I've seen in several regulatory submissions that have gone through the FDA where the intended use and the indications for use are flat out wrong. And when I point out to the FDA, they say, " Yes, Mike, we agree with you. These statements are wrong, however these submissions have gone through, so there's nothing we can do about it. That's our US government hard at work. What can we do?" And one other thing, and then we'll move on to your next common mistake. Jon, one other thing about the labeling is there is no consistency across CDRH when it comes to intended use and indications for use. It drives me nuts when I hear the politicians at FDA say, "Well, it doesn't matter what part of FDA you're dealing with or what reviewer you're dealing with because all of the reviewers are following the same regulation the same way." Well, with all due respect, well, you know what? I mean, does anybody really believe that? Here's the thing. In some groups in CDRH, they like to have what I call the traditional intended use and indications for use separate and distinct, which is actually my personal preference. However, in other groups within CDRH, they're okay with what I call blended intended use indications for use statement. What you need to do before going to the FDA and always do this before a pre- sub Jon, is find out, since we're talking about 510(k)s here, clearly there's precedent for this. Find out where in your particular section of CDRH that you're dealing with. Do they tend to prefer separate intended use and indications for use or a blended statement? And then give them what it is in that format. Once again, Jon, this is definitely a common mistake, but it's a mistake that is very easily avoided.
Jon Speer: Yeah. Absolutely.
Mike Drues: Okay, should we move on to the next one?
Jon Speer: Yeah, sure. Let's do it.
Mike Drues: Okay. So, the next common mistake that you identify in your column is not using the checklist, and you're referring specifically to the refuse to accept or RTA checklist. By the way, there's an RTA checklist obviously for the 510(k), but there's also one for the de novo and for the PMA as well.
Jon Speer: And I think there's also different versions for 510( k) s depending on if it's a traditional or a special or an abbreviated too, right?
Mike Drues: Absolutely correct. There are different versions of the RTA checklist. If you just look at the 510(k) RTA checklist guidance, there are different versions for each of those different 510( k) types. There is no checklist available yet for the newest version of the 510(k), the safety and performance, that's 510( k), whether we need one or not, I'm not sure because it's so similar to the abbreviated one. But you're exactly right, there are versions. And as I think I pointed out in previous discussions when we talked about the RTA in more detail, Jon, ironic as it might sound, the abbreviated RTA checklist for the 510( k) is actually longer than for the traditional 510(k). And comparing the 510(k) to the de novo, to the PMA RTA checklist. Ironic as it sounds, the PMA checklist is actually much shorter than either the 510(k) or the de novo checklist. So anyway, that's just a little bit of regulatory trivia for you and our audience to chew on. But specifically, in your column under this common mistake of not using the checklist, you say follow the checklist exactly. Then give your a reviewer a heads up by referring to the relevant pages in your document next to the checklist items. Can you explain to the audience, Jon, exactly what you mean by that? Again, I think I know what you mean, but let's hear it in your words.
Jon Speer: Yeah, sure. The RTA checklist, I think was intended to be a tool for the FDA reviewer when submission received for them to go through and make sure all the constituent parts and components and what is expected to be in that submission is included. So, it's a guide, if you will, for that reviewer. But as submitter, as one who might be submitting a inaudible, it should also be a guide, or at least one of those final checklists that I review as well. The glorious thing about regulations a lot of time is within reason anyway. FDA tells you what they're going to do, what they're going to look at and that sort of thing. Now, sometimes you get into the details and there might be some nuances or some specifics that you can't plan or anticipate, but that's fine. But the RTA checklist tells FDA what they're looking for and it tells you as submitter, this is what FDA is expecting. Now, there's also the 510( k) submission contents that's pretty well defined and articulated as well. There certainly are guidances for this thing. But the FDA, they've added this step that they refused to accept the step that's been in place now for a bit, and I think that was in direct response to the industry. Because the industry is like, " Oh wait, we don't know what happens when we submit our 510( k). We don't hear back from FDA. Well, they've implemented this step so that within the first, I think, 14 days, if I recall, you'll hear back from FDA. " Yes, we received it. It looks good to go. And now, it's moving on to the substantiative review. Well, you can make life easier for yourself and for your FDA reviewer if you can put a little annotated reference, I always advise people, make a copy of that refuse to accept checklist, list in the comment section where these things are addressed within your submission, what section, what page. So, it becomes a table of contents or at least an index for that reviewer so they know where to find it and they can flip to the appropriate page. They say, " Aha, here it is. Looks good to go." And include that when you submit your 510( k). I think trying to keep in mind, FDA reviewers see probably dozens, maybe hundreds of submissions in a finite period of time. Some are better than others. Well, make yours stand out, make yours easier to read, help that reviewer understand the story you're telling and where they can find the important contents and components that are identified on that checklist. FDA thinks that checklist is important, so put some value into it, help them out a little bit, help yourself out a little bit.
Mike Drues: I agree, Jon. And I give very similar, in fact, substantially equivalent advice to my customers all the time. As one of my customers likes to say, " You have to chew FDA's food for them." And I now repeat that to virtually all of my customers. Don't expect. I think this is a totally unreasonable expectation for a reviewer to jump around and to look through hundreds or maybe thousands of pages of information to find the information that he or she is looking for. Quite frankly, that ain't going to happen. And I have customers that tell me, " Well, we've already described how we do this in a publication. How about if we just cite the publication or provide the publication as an appendix at the back of the document?" Well, why would you expect somebody at FDA to read through some article and extract from it whatever information that you think is the important information?
Jon Speer: Exactly.
Mike Drues: You literally have to chew FDA's food for them. And I don't mean that to be patronizing, I really don't. But if you want to be in that small percentage of people that have their 510( k) s cleared on the very first time and not in the very large majority where you get these RTAs or these questions back in ping pong, back and forth. Some of the tips that Jon and I are talking about today will help you do that. Moving on to the next common mistake, and that is not providing all expected testing. Not providing all expected testing. And one of the things that you mentioned in your column, Jon, is FDA used to accept" I promise statements," or in the regulatory world, what we call a promissory note. Can you explain to the audience, Jon, what you mean by an I promise statement or a promissory note?
Jon Speer: Yeah, keep in mind, I'm old. I've been doing this for a bit. I mean, the gray hair and the beard should give some of that away, but back in when I started my career, 510( k) s have been around since almost before I've been alive. So, they've been around for all time.
Mike Drues: 1976, 1976.
Jon Speer: So, I was born in'75. But anyway, back when I first entered this industry as product development engineer, I mean obviously as we've talked about the 510(k) is the workforce. That was the most common approach that we did at the company I worked for at that time, that we pursued for FDA regulatory submissions. At that time, I'm talking late 90s, early 2000s, and this was probably a true statement up until 2010- ish or so, give or take. Sometimes there's tests that are lengthier and whatnot. And a couple examples that come to mind are things like sterilization validation, sometimes shelf life testing. In some cases, some biocompatibility tests. But back in those days, back in the good old days, we used to be able to say, " Yes, we acknowledge we need to do this. It's underway, it's in process. We promised to do this and we promised that it will be done prior to going to market." And that used to be an accepted practice back then. But my understanding, well, that's not been an accepted practice for quite some time. And I think the reason is, well, it's probably quite obvious. I think a lot of people, a lot companies made these I promise or the promissory note commitments and then they didn't do it. And so now, FDA's like, " That's not an acceptable practice anymore. If you are submitting a 510( k), thou shall do the testing. We no longer accept, I promise, promissory note commitments because people before you, companies before you, they didn't always follow through and do it." So, don't think you can go to FDA with 510(k) with some sort of promissory note to commit to do it before you go to market. It will be expected at the time of the submission.
Mike Drues: Yeah, good point, Jon. So, just to recap, so a promissory note in the regulatory vernacular is something that we say to the FDA, " Here's 95 or 99% of our submission. These few little bits, whether it's shelf life or sterility data or something like that, we will provide to you a little bit further down the road. In the meantime, will you please chew on what we've given you?" That was up until several years ago, I don't want to say common practice, but it was an accepted practice. Today, it's not so much. And the reason is you just pointed out is because regrettably, and I do not take pride about thing about this with our industry, regrettably, there have been companies and people in them who have promised to do certain things who have never gone on and do that. And so, as a result, FDA has gotten burned a few times. And so, I wouldn't go so far now, Jon, is to say that FDA won't take a promissory note because they do still take them. I've had them take them for some of the submissions that I've been involved with very, very recently, but it has become very much the exception rather than rule. A promissory note should be like going to the ombudsman. It's a path of last resort. And I have companies all the time, they ask me, " Well, Mike, can we give FDA 95% of our submission and let them start chewing on it? Like my graduate students, " Can I give you a draft of my report and you can mark it up and give it back to me so that then I'll repeat?" Well, when I teach in universities, they don't pay me enough to do that. Similarly, with the FDA. So, you cannot give them a draft. You can take it to them with a pre- sub. And if you're planning on doing a promissory note for thing like sterility or shelf life or stability testing, definitely bring it up in a pre- sub to make sure that it's okay with them. Because if you're springing it on them at the point of actual submission, as you point out very well in your column, Jon, you might be able to fake it to get it through the RTA administrative review, but they're certainly going to pick it up on the scientific review.
Jon Speer: Absolutely, absolutely.
Mike Drues: So, my advice is use it only as the last resort. But if you're going to do it, plan on bringing it up in a pre- sub in advance to make sure that everybody knows that, that's coming.
Jon Speer: Absolutely.
Mike Drues: Okay. Moving on to the next of the seven common mistakes that is, not doing real time shelf life testing, not doing real time shelf life studies. So, why don't you explain a little bit what you mean by this, Jon, and why is this such a common mistake?
Jon Speer: Sure. In shelf life, I was trying to think. I'm sure there are examples. Well, let me recant a little bit. I believe, FDA is looking for you to address the shelf life for any and all medical devices. I think it's more, well, I don't want to, I guess, put a label on whether it's more or less critical. But a lot of my experience, certainly in my formative years in the industry, I designed and developed single use disposable devices. And most of the time, not always, these devices were packaged and terminally sterilized via ethylene oxide, sometimes gamma or what have you. But nonetheless, it's manufactured, sterilized, and it's going to probably sit on a shelf either at a distribution center or at the point of use until such time that it's needed. A lot of the devices that I worked on, we usually targeted a three- year shelf life from data sterilization. We can't just make that claim arbitrary. We have to be able to corroborate that shelf life is applicable and appropriate for that product. Obviously, if I'm making, I'll just use something I'm very familiar with, a catheter that has been packaged in a Tyvek peel pouch that has been sterilized by ethylene oxide. I can't just arbitrarily say, " Yep, it's just like this other thing, slap three years on it and we're good to go." I have to be able to support that with objective evidence. I have to be able to prove that. But it's not pragmatic or practical for me to wait until I have that real time three- year shelf life data before I submit my 510( k) because obviously, I mean, 510( k) a catheter type device, I mean, I'm not stating specifics here. But generally speaking, I mean from concept to go to market, that could be an 18 to 36- month project to begin with. So, if I have to wait three years after I finalized my design to do the real time shelf life studies, that's just not pragmatic. So, what a lot of companies do is they support this with oftentimes accelerated aging and whatnot. Or maybe, they go to market with a reduced shelf life instead of the three year, maybe that's the target. Maybe they come to market with a one year or what have you. There's a lot of different approaches to this. But regardless, part of my strategy and my planning when I'm in product development, I need to basically put that into my overall plan for this product that I need to reserve quantities of product that have been packaged and sterilize and I need to keep them on a shelf over a period of time. And I need to, at certain intervals along the way, I need to pull some of those samples and I need to conduct representative testing to be able to support and demonstrate and corroborate that my product is stable and effective and safe and all these sorts of things along that intended shelf life. I've rambled a little bit, but hopefully that's a little bit clear on what I meant by that.
Mike Drues: I think that was very good, Jon. If I could just add on a little bit further. Exactly right. Real time shelf life testing is one option, but most companies don't want to wait around for that data real time. So, there are oftentimes methods that we can use to accelerate that process by putting your device in a heated and environment or with increased humidity or something like that. You need to make sure, if you're working on a conventional device, there are probably standard methods to do these kinds of things and you need to make sure that you're following these standard methods. However, it really comes down to the device itself, the design and specifically the material. So, if you're using a biostable material, if you're talking about, say, a bare metal coronary stent. Packaging issues aside, what's the shelf life of that stent? A million years, nothing's going to happen to that stent. But for example, if you're working on a bioabsorbable material, now all bets are off. It's not a biostable material, it's a bioabsorbable material. And further, if you're working with a combination product, putting a drug or a biologic or something like that on a device like a stent, doing accelerated testing becomes very, very difficult to do. Even traditional shelf life testing becomes very challenging to do. This takes this conversation multiple levels higher, Jon. But never mind a drug, putting a biologic, like a monoclonal antibody or a gene inside of a virus on the surface of a stent. The whole notion of putting that on a shelf and having it sit there for some number of weeks or months is just nuts when you think about it. So, for most of our audience that are working on conventional as a biomedical engineer, Jon, dare I say it, boring medical devices, shelf life and sterility and stability is pretty straightforward. But for those in our audience that are working on newer, more challenging kind of technologies such as the ones that I just described, now it becomes much more interesting discussion. Most important thing to remember, Jon, I know you're obviously a big fan of the design controls. One of the most basic tenets of the design controls is that you want to do your final verification, validation testing, your VMV testing, including shelf life and sterility and so on, on your final device. And that includes your package device after it's been on the shelf for whatever period of time. I want to reiterate the strategy that I also give to a lot of my customers. If ultimately, you want to have say, a three- year shelf life claim, but you don't want to wait, you wouldn't want to hold up your 510( k) submission to support that claim, especially if you have to do it in real time, you could start out with say, a six month or a one- year claim. Get it onto the market that way, and then go back later as a label expansion and increase your shelf life claim that way. That's a common strategy. And the last tidbit that I thought I would bring up here to challenge our audience a little bit, it's one thing to talk about shelf life for traditional medical devices. It's another thing to talk about shelf life for say, combination products or bioabsorbable materials. But how about the shelf life for and SaMD, software as a medical device.
Jon Speer: It's tricky, right?
Mike Drues: Most people would say, " Well, gee, how the heck is shelf life an issue when it comes to software?" Well, this is another example, Jon, of when I see people trying to understand the letter of the law, but not the spirit of the law. The spirit of the concept of shelf life, the philosophy of shelf life is what happens to your product in the future. In the case of software, if your software is running on, say, a computer with Microsoft Windows platform, what happens as future updates issued to that software? So, although the concept of shelf life doesn't maybe map directly to a device like software. Philosophically, the concepts are very, very similar, if not the same. Would you agree with that, Jon, or am I just smoking my socks?
Jon Speer: No, I do agree. And I think the summary of this topic is regardless of whether or not you're working on a called boring medical device or like an SaMD or even an electromechanical device, you need to address the topic of shelf lives. Don't just assume, " Oh, well, it's SaMD. FDA knows its software, no big deal. Obviously, there's no shelf life because it's not sitting on a shelf." Don't assume that. Explain it. Provide the rationale and the evidence as to why you've taken the stance that you do.
Mike Drues: But it is, you mentioned it's not sitting on the shelf. But let me extend that metaphor a little bit further, Jon. It is sitting on the hard drive of your computer or your phone or whatever it is. And even though your device might not be changing the environment around that device might be changing. And so, again, I would argue that the concept of shelf life, at least philosophically, is directly applicable to software as it is to anything else.
Jon Speer: Yeah. Okay.
Mike Drues: Okay. So, let's move on to the last couple of common mistakes that you identify in your column, Jon. The next one is underestimating risk management requirements. Underestimating risk management requirements. One of the pieces of advice that you give to the readers of your column, Jon, is be prepared to show your list of requirements and provide the paper trail for your traceability and your DHF, your design history file and the other quality requirements. I'm curious, Jon, if you can just explain for our audience, why is that important in the context of a 510(k)?
Jon Speer: Yeah, again, I'm dating myself a little bit, but I can recall, well definitely I can recall long before ISO 14971 that the current industry's standard for medical device product risk management was established. I also have an opinion that design control best practices in and of themselves are a risk management approach because you're defining user needs and the requirements and you're doing the verification and validation of your product and demonstrate that it works the way it's supposed to for the people that it's supposed to work for. So, that in and of itself is risk management practice. But in today's world, there is an expectation, I think, by FDA and probably many other or most other modern regulatory bodies that 14971 and risk management practices are deployed for your device. But a big component of this, like I said, it does go back to the design control piece. I mean, you're identifying what can go wrong with your particular product. And not to sound dire here, but you need to sometimes when things go wrong, people could get hurt. So, how are you preventing, or how are you controlling your device so that those bad things don't happen? Or if those bad things do happen, how are you giving the doctor, the nurse, the user of that product, a warning, an alarm or some sort of notice that, " Hey, something's not right here." But regardless, you need to be able to think about all of these different scenarios. And yeah, you think about the normal, " Read the DFU, the directions for use." But let's be real. I mean, you and I have talked about this a lot. That's probably the first thing that gets tossed in the garbage whenever someone absolutely gets a product.
Mike Drues: Absolutely.
Jon Speer: So, think about other ways. This is, as you and I have talked a lot about human factors and things in the past, and not to rehash that whole topic. But risk management is really thinking about all the ways that your product are going to be used, whether use the way you intended and expect or the way they might actually be used. So, I think it is an expected best practice these days to be able to document those types of scenario in those cases and make sure that you have the right controls with your product so that bad things don't happen.
Mike Drues: Well, Jon, as a very experienced and very capable quality professional, I understand, and I agree with the importance of documentation. The only thing that I would add to what you just said, Jon, is in the context of the 510( k) submission, a lot of that quality documentation doesn't directly go into the submission.
Jon Speer: It's true.
Mike Drues: Don't get me wrong. I'm not saying it's not important. It's very important and you have to have all of your quality ducks in a row. But a lot of that stuff comes in, in the form of a manufacturing inspection when FDA comes knocking on your door.
Jon Speer: It's true. That's true.
Mike Drues: It's not part of the actual 510(k) submission. It is sometimes part of a de novo submission, and it is very often part of a PMA submission. But because we're trying to focus on 510(k)s here, I just wanted to point that out. And specifically when it comes to risk, you and I have talked about risk many times, Jon, and as our audience probably knows, I happen to be a subject matter of expert for FDA in risk. The risk requirements for the 510( k) are the most important. It's not the risk management plan, that's important as well. But meeting the risk requirements for the 510(k) are what's most important. And just to reiterate them very, very quickly. In order for it to have a successful 510(k), when it comes to risk, there's two boxes you have to tick. The first is you have to show that there are no risks in your device that are not already present in the predicate. That's risk requirement number one. And risk requirement number two is of the known risks, you have to show that the level of those risks in your device are the same or lower than they are in the predicate. In other words, hypothetically speaking, if FDA can identify one new risk that's in your device that's not in the predicate, or if they can identify one risk or device that's higher than it is in the predicate, then that would be enough, at least theoretically, to trump everything else. And for FDA to say, " Sorry, your device is not substantially equivalent. You need to be considering a de novo or maybe even a PMA here." And I can also tell you, Jon, that in the past, those risk requirements have never been interpreted as literally as I just explained them, but they are being interpreted that way today. For the benefit of our audience, Jon and I have done a lot of other discussions specifically on risk and provide some references to go into that in much more detail. But Jon, let's move on to the last two common mistakes for 510(k)s. The next one of my personal favorites, and that is, not demonstrating equivalence to a predicate. I'm pretty sure that I understand what you mean by that, Jon. But for the audience, why don't you explain what you mean by that?
Jon Speer: Again, to remind folks, hopefully this is information that listeners already know. But the whole premise of a 510( k) is that you are demonstrating substantial equivalents to a predicate device. In layperson terminology, basically, that means I'm comparing my product to something else the FDA has already cleared, and I'm making the case that my device is just as safe with same or known risk as this other product. Hopefully, that's not too much of an oversimplification, but I need to be able to corroborate and demonstrate that. And a lot of folks are like, " All right. What do I do? Just call my competitor and the predicate that I'm planning to use and order some products and do side by side comparative testing and analysis?" Well, sometimes you could do that, sure. But a lot of times, this is some of the nice things I think about freedom of information and a lot of this being available in the public domain. Hopefully, if you're citing a specific predicate device that you're going to compare substantial equivalents to, hopefully, you've reviewed as much information about that product as you possibly can, including that devices 510( k) submission. And a lot of times when you do this, there is quite a bit of information that you can glean and extract from those submissions. I mean, go to that product's website, you'll read the literature, find as much information as you possibly can. There's lots of ways to be able to demonstrate equivalence. I think to me, I mean, certainly saying the predicate device, and this is very, very generic, and I don't mean this to be taken too literal. But the predicate device is blue. Oh, my device is blue, the predicate device is a pound. Oh look, my device is a pound. So, you're trying to basically tell the story that this predicate device has these characteristics. My device has, I'll say similar characteristics or substantially equivalent characteristics. And you have an approach to this I really, really like, so I don't want to steal your thunder on this. So hopefully that helps folks understand a little bit that the bar you're trying to get to is that my device is for all intensive purpose the same as this other thing that I'm comparing it to.
Mike Drues: Well, first of all, Jon, I'm flattered that you said you don't want to steal from me. I would encourage you to do so because there's an expression, " If you're going to steal, steal from the best." So, I'm flattered, and I'm going to steal from you in just a moment. I think you can probably guess how. But I thought it was interesting you started out, and this part of our discussion by saying that most of our audience is probably familiar with this. I questioned that. I'm not sure that the data really would support that claim. And the reason why I say that, Jon, I'll share with you a statistic that I've shared many times. It changes a little bit, but not much over many, many years. About 75% of 510(k) cases that are submitted to the FDA today in 2022 are rejected. And of those ones that are rejected, about 85% of them are rejected specifically because of substantial equivalents or the lack thereof. So, I find it fascinating how so many people in this industry, they think that substantial equivalents is such a simple concept, such a no brainer. Well, if it is, how can you explain the statistics? And part of it is, Jon, the devils are the details. You mentioned anecdotally that, well, if the competitor's device is blue and your device is blue, but what if it's not exactly the same shade of blue? What if yours is a darker blue and theirs is a lighter blue, and so on and so on. The regulation says that your devices have to be substantially equivalent. That does not say that they have to be exactly the same. And any differences, whether we're talking about differences in the technology, like for example, the blue car or differences in the labeling. The company needs to address those differences and specifically say or show that these differences, their device is light blue, mine is dark. That's not important because, A, it does not create new questions of safety and effectiveness. That's requirement number one. And B, as I just talked about a minute ago, it does not change the overall risk. Every single difference, whether it's a change in the labeling, you're using one word and they're using another word or there's a change in the color or whatever technological parameter, every single difference, you have to go through that analysis of, does it create new questions of safety and efficacy? Does it change the overall risk? And then, the last thing that I'll mention, because we've done a lot of things on substantial equivalence, I've done a webinar for Green Light on substantial equivalence. We can provide a reference to that as well as the different 510(k)s. Every form of 510( k) relies on substantial equivalents. The traditional, the special, the abbreviated, and now even the safety and performance or FSP 510(k), but substantial equivalents in a different way. So, just briefly to recap that. Traditional 510(k)s, about 75% of 510(k)s are traditional. You need to show that it's substantial equivalent to another device. A device that's already on the market, a special 510( k), about 20% of 510(k)s are special 510( k) s. You need to show that your changed device is substantially equivalent to the pre- changed device. In an abbreviated 510( k), or what is now sort of morphed into the safety and performance 510( k), you do not show substantial equivalence to another device. You instead show substantial equivalents to a standard, to a guidance document, to something like that. So, in every single type of 510( k), you need to show substantial equivalence. But how you show it to what you show it to is different. At the end of the day, the regulation, which as you pointed out earlier, Jon, the 510( k) was created in 1976. The regulation has not changed one sentence, one word, one punctuation mark since 1976. Yeah, there have been guidance documents that have come up, but the regulation itself has not changed on substantial equivalents. The regulation says that you have to show your devices substantially equivalent. It does not say how you show it. That's up to the manufacturer. Moving on to the very last one, and then we can wrap this up. Jon, is formatting of the submission incorrectly. I would like to think this is a no brainer. What you're talking about here is submissions that are rejected under administrative review. Just to remind our audience, when you submit a 510(k) or any submission to the FDA, it's basically, it's a two- step review process. The first step is the administrative review. The second step is the scientific or the substantive review. The administrative review is nothing more than when a bean counter sits down at FDA and looks at your submission and they have their version of the RTA checklist and they say, " Do you have this section? Do you have that section?" And makes sure that you're not missing anything. Make sure that you have all your signatures and so on. Make sure that you don't leave the page numbers out, which is one of the things that you mentioned in your column. Jon, I would like to think that people have graduated from elementary school. They don't need a reminder to put your name and your page number on your homework assignment. But regrettably, Jon, 38% of 510( k) s that are rejected by the FDA, 38% are rejected on administrator review. And I've said before, and I would love to hear your thoughts on this. When a 510(k) or any other submission is rejected by the FDA on scientific review, maybe FDA has a difference in your scientific methodology or your statistical analysis or something. Okay, that's a fair job. That's FDA doing their job. But when FDA rejects your submission on administrative review, it's not FDA or anybody's fault. It's 100% these company's fault. So, given that you wrote this column in five years ago, Jon, but the statistics still remain about the same, about 38% are rejected on administrative review. Why is that? And what other advice could you give to our audience to eliminate that?
Jon Speer: Why is that? I don't understand. Because to me, this is just, it's common sense. I mean, I was trying to think, well, it's basic technical writing or what have you, but it's common sense. I mean, keep, again, I am preparing a bunch of information to support my claim, my indications for use, my intended use of my product, that it's substantially equivalent to a predicate or multi predicate or whatever the case may be. We can talk about that at a different time. But I'm making the claim that it's substantially equivalent to something else that FDA has cleared on 510(k) before. It's my story. The point you made earlier, maybe my device has been published in some scientific journal. I mean, just to think, " Oh, well, FDA's going to go Google that or PubMed will find that. And that's naive. Come on, tell your story, but make it easy for the reader to follow your story. And simple things like pagination, and that is so easy to do. Every modern word processor does that these days. It's just things like that. Providing a table of contents, just making it very, very simple. I mean, this is the textbook, so to speak, for your device or the executive summary textbook for your device for FDA to review the information.
Mike Drues: You started out, Jon, by saying that this is common sense, and I agree this should be common sense. inaudible
Jon Speer: Well, 38% of things are rejected. It says it's not common sense,
Mike Drues: But unfortunately, common sense is not as common as we would like to think. So, I would go back to one of the first pieces of advice that you gave in your column, Jon, and that is follow the RTA checklist. It lists many of the things that we just identified, including things like table of contents and page numbers. So, follow that. And for those that are working in the 510(k) universe, you can at least be comforted in the fact that you're not working in the PMA universe. Because in the PMA universe, believe it or not, Jon, the level of micromanagement in the submission is phenomenal to the point where FDA specifies font sizes and margin spacing and so on and so on. I mean, that to me is insulting if not condescending. But we don't have that in the 510( k) universe. The other thing that you said, Jon, that I like very much is tell a story. Obviously, this should be written as non- fiction, but it should be like a novel. It should be something that somebody can sit down and read and make it easy for the reader to read. Don't make them jump through a bunch of hoops. I get drafts from my customers all the time that are painful to read. I literally mean painful. One last thing, and then we'll wrap this up. Jon, as you know, I work as a consultant for the FDA. So, I see submissions coming in to the agency from time to time. And some of these submissions are terrific. But other submissions, they are literally painful to read. It's like, " Did you go to elementary school? Do you know how to construct a sentence with a noun and a verb and so on?" Make sure that this thing is readable. It sounds like common sense and it is common sense. But unfortunately, it's not as common as we would think. So, what are your final thoughts, Jon? Any final pieces of advice to send our audience to ultimately try to improve the statistics that in spite of your very good column that you wrote five years ago, a lot of people seem to be running into these same problems over and over again.
Jon Speer: Folks, if you're curious or, I mean, the simple piece of advice is get somebody else's eyeballs looking at this. In my experience, a lot of times, a 510( k) submission is a collaborative effort. Sometimes, divide and conquer. Maybe I take this section, you take that section, and so on and so forth. But at the end of the day, somebody's got to put it all together. And when somebody's put it all together and you've got the final draft or a draft that's ready, go find your professor. Maybe call Mike Drues or Jon Speer and say, " Hey, before I send this to FDA, can you look over this?" I mean, that's the simplest, easiest thing that can be done. I think one part of that might be helpful is it could be somebody within your company, that's cool. But maybe they don't have the intimate knowledge. Probably, it shouldn't be somebody that's on the project team that's reviewing this because they're going to be preconditioned, so to speak. So, say, " Oh, well, I know this because I worked on the project." Find somebody that probably hasn't worked on the project. They might be a little bit familiar, but they don't have intimate details. I mean, I guess what I'm trying to say is find an objective reviewer that's not FDA before you submit it to FDA. That's the simplest, easiest thing one can do to try to correct some of these mistakes.
Mike Drues: I agree with you, Jon. And just to wrap up my part of the conversation and then we can wrap up today's discussion, I would just like to amplify and throw a little gasoline on that last suggestion that you just made. And that is get some other eyeballs on this. One of the things that I find so fascinating and quite frankly, so scary about so many medical device companies is who's the first person to see your submission outside of your company? It's usually the FDA. And to me, that makes absolutely no sense. Get somebody, whether it's Jon or myself or somebody to come in, and I don't need to be self- serving. I'm trying to share some of my inaudible.
Jon Speer: And I worked hard of talking about these mistakes, right?
Mike Drues: We are. We are. But one of the things that I do with my customers is before they go to the FDA with their pre- sub or their submission or whatever it is, they'll ask me to come in, put my FDA reviewer hat on, read through their documents, sit through their presentation, and if I can be a bit blunt, bash the hell out of it. Because the idea is, if you're going to make a mistake, better for them to make it in front of me. What do I count? I don't matter as opposed to making it at the FDA. And I don't want to go so far as to say that if a medical device company can get their submission through me, they can get it through the FDA. But I've been doing this long enough, Jon, I make a pretty good surrogate for the FDA.
Jon Speer: I'm sure.
Mike Drues: If you have somebody come in and review your submission and say, " Oh, you're doing a wonderful job," pat yourself on the back. Have a parade. They're not doing their job. You want somebody to come in and be brutally honest. And the last thing I would just remind you and our audience, Jon, is I know you're an expert when it comes to the design controls. This is not a foreign concept. What I'm describing here is the concept of an independent reviewer, right?
Jon Speer: Yes.
Mike Drues: But it's not an independent reviewer in sort of the engineering sense of the word, which is the connotation of that regulation. It's the concept of independent reviewer in the regulatory sense of the word. And so, once again, Jon, we don't need more regulation to solve these problems. We've already got thousands and thousands pages of that. What we need is more people understanding the intent of the regulation and being able to apply it in different circumstances. Whether you're talking about applying the concept of independent reviewer to regulatory review, as I just discussed, whether you're talking about applying the concept of shelf life to software as a medical device as I discussed earlier, you've got to get past this literal interpretation of regulation and think much more broadly, maybe even a little more philosophically if you will.
Jon Speer: Absolutely.
Mike Drues: Those are my thoughts, Jon. Anything that you would add and then we can wrap this up?
Jon Speer: No, I don't know. I think that's a good place to stop the conversation. And folks, please, a lot of these things is, hopefully you've picked up from Mike and my conversation on this topic. They're absolutely in your control. These are things that you can absolutely take care of yourself and with a little help, again, sometimes when I work on something, I know what's there because I wrote it. I can take it to my Mom and she's going to think that it's amazing and she's going to want to copy of it and she's going to want to put it on her refrigerator.
Mike Drues: Put that on the refrigerator.
Jon Speer: My mom, yeah, might not be the best person to ask to give a critical review. As Mike said, find an independent person, somebody that can be a little bit more critical because I want someone who's friendly to the cause, so to speak. Not that the FDA isn't, but I want them to beat it up, rip apart, mash the hell out of it before I send it off. So, then if you can't find anybody, reach out to me, reach out to Mike. We'll happily help you. It's just we wanted to change the statistics on this topic. These are in our control as an industry.
Mike Drues: I could not agree more, Jon.
Jon Speer: All right. Well, Mike, thank you. It's good to go back and sadly review an article that I wrote five years ago that is still applicable and appropriate. So, industry, listen, pay attention, change this, make this article obsolete so that we don't have to talk about these mistakes anymore. And I hope you're enjoying the Global Medical Device podcast. And until next time, we'll talk to you real soon.
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.
Etienne Nichols is a Medical Device Guru and Mechanical Engineer who loves learning and teaching how systems work together. He has both manufacturing and product development experience, even aiding in the development of combination drug-delivery devices, from startup to Fortune 500 companies and holds a Project...