Inside the FDA Real Stories, Real Fixes Tianna Benson (leanRAQA)

Going into 2026, what are the basic expectations of, skill sets around AI and AI literacy? And this is a question that's been on my mind, and a survey that I'm gonna launch. And so what I'm asking you for, as I'm busy promoting everybody to panelists, is that when I launch this survey, next week, you help me promote it, and you help me get people to do the survey, because I think it's really important to set a baseline for Quality and regulatory professionals, particularly, around… What are you… what should you know? what does AI literacy look like? How do you define that? How do you define that for the people that you're hiring? And, you know, if anybody would like to jump in here and save me from running my mouth, I'd love it. And …

That's what I'm thinking about. So yeah, there's gonna be a survey, written the survey, have to just load it into Google Sheets, or whatever it's called, and … and promote the things. So, please help me promote this survey. We'd like to get, you know, several hundred respondents to sort of set a baseline and understand, you know. How much we need to know about this thing called AI. … Martin, thoughts?

We need to know when not to use it. So, I think we just need to know what the limitat- what our limitations are in our particular roles right now. That's the situation right now.

And what do you think?

entertainment.

Look like in 6 months.

Yeah, so, I, I have a new employee whose name is Nate, and part of what his job is, is to watch like, give me weekly rundowns of everything that happened. And, like, one week it was, like, a slow news week, and he's like, he's like, I found this clip, you know, there just wasn't much today, so it seems like that they're just using old content. And it was a post that was generated by AI, and it said, something about the… the special 510K program, parentheses, SBD, And then it goes on to use SBD to talk about the special 510K, and I'm like, SBD is small business designation. It has absolutely nothing to do with a 510K.

So, you know, there are things like that. But, you know, he's… he was brand new to this. He didn't know, and so, like, that was, like, learning for everybody, you know?

Well, I've found that AI scrapes a lot of information, and I always am checking and double-checking what it is. Because I've been doing a lot of, putting together presentations using AI, and I don't trust Even when I give it my own information, I double-check it.

Yep.

It's not a hallucination.

But we're gonna have a generation that does not know things for themselves. and is not capable of knowing when AI is incorrect.

Maybe we should host a funeral for the card catalog in the library.

So….

For those robots.

No, brother.

If you go to the stacks.

But that is actually… there's a thing in Europe which is the over-reliance on AI. There's actually a complete issue, and that's a general issue. It's not limited to med tech or pharma. There's a general issue that people are over-relying on AI, and… Here it is, it was on the news. People getting emotionally involved with their AI chat.

Oh, I'm completely emotionally involved.

Isn't there a movie about that? Like, actually, it wasn't AI, it was a cell phone.

Yeah.

I fell in love with his cell phone.

Yep.

I have….

Many of my clients have put.

in their NDAs with me. I don't… you may not be aware, but anything that goes through AI is actually you're putting that information up into a place that you can be violating your NDAs. Many NDAs explicitly say you're not allowed to use AI on their… on my project with them. So….

I think it's… yeah, I think it's very important. Yeah, be careful.

cybersecurity world.

it's very important to… to not bunch all AI. It's really important to understand your own, I suppose, tech stack, so what you're dealing with internally in your organization, what your organization's policy is. And also understand what the limitations are, and also what type of data that you're dealing with. And that may involve a process of categorizing your data into, okay, this is the stuff that I can use on ChatGPT, this is the stuff I can use and feed it into, you know, any open, AI tool, then this is the stuff I need to be more cautious with, and this is the stuff that doesn't go into, you know, anything where there's a risk of it being exposed. But that's a function of learning.

And we… I really think that the… Everyone is going to have to get… come to terms with, some form of AI literacy. I go home, I go home to where I was brought up, every few weeks, and my auntie, gives me her iPhone and says, hi, I haven't used this in a week because it's got this notification, and I'm afraid to press the button. Because she's afraid of being scammed. And then she'll give me her laptop, and I have to do the same thing. And if we are that cautious with AI, if we don't learn the actual underlying way that it works, we are going to be my Auntie Nora in… a few years, and I think we all have an Auntie Nora, or a friend who's a little bit like that. We can't be that. We need… and that doesn't mean running into it headfirst, but I do honestly think that if we

Don't learn, and if we don't prepare ourselves and prepare our organizations, then we're going to get, We're gonna… we're gonna be on to Nora.

Well, I had an interesting thing. I had written something on my website, and I wasn't really sure about it, and so it kept nagging me in the back of my head. So after a few days, I thought, okay, let's ask ChatCTP what they think about it, and I confirmed it, and that was fine, and I thought, no, no, let's see what are the sources. Well, the source was me.

….

Yeah, I'm honored, of course, but, ….

No.

Okay, if there's no confirmation, let's delete the whole thing.

I mean, you know, I really like, what, Lisa, you were talking about, right? I mean, AI is here to stay, guys. I've been involved with it very, very seriously the last 6-7 months, doing Agentic AI. So, you have to learn to wrap, you know, we as businesses are going to morph. Because it's a phenomenal tool, but it has to be controlled, you gotta learn how to manage it, you gotta protect your data, you gotta put that in… you gotta… but how businesses are going to change is… there's no doubt they're already changing, you're totally involved with it in the regulated space. I've been doing regulation. Now, I just recently left Labyrinth.

And I'm with an Indian IT services company with 750 consultants with a huge AI, agentic AI process, practice, complete with digital twins, mapping SLPs, the data's very, very secure. Yes, your new term is human in the loop. It still will hallucinate. But we're going through a maturity model and an evolutionary cycle, but the rate of change is mind-blowing. everything will change. The very structure of business will change. Now, the people with I got infuriated, I was at the Tech Week in New York, right? And there were signs, ads, stop hiring humans, hire agents, bad idea.

Right? What we're going with is, is, hey, bring efficiencies where relevant. Manage your tribal knowledge, capture your tribal knowledge, and drop it into the AI. Let that influence your AI. Right? Bring in the wisdom seekers, the keepers of your wisdom. Bring them there to validate and build AI-driven automations. LLM is just retail AI. When you get into Agentic AI, forget about it. The very structured business will change. And what we are recommending, and we're building systems for, right, is to bring Agent… to bring what I call now, I've got a whole paper about this, a go-to-market strategy for the new company, which is Agentic.

Business automation. That's where the magic's at. And you leverage the people, human in the loop. So anyway, my two cents on that. But it's a fascinating….

We have to move on, we have to move on to today's topic and today's speaker. Now, I promised you that we were gonna have Tiana Benson, Ty Benson, from LeanRAQA, and I did an awful beat and switch, and downgraded you to her assistant, Michelle, and … So, … to set up today's topic, a couple of weeks ago, I was on a call with them, and they started talking about the situation that has come up a number of times at the FDA recently, where they submitted a product for 510K approval, and it had two well-established technologies in it, being used either in a different part of the body.

Carrie, or in, in a different fashion, for a different indication… for a different indication, and what do you think happened?

Well, I think it was….

I'll let Michelle take it from here.

So we'll talk about that case study, in a minute.

You'll spare….

Somebody's got a lot of background.

Yeah, somebody's gotta mute.

John Bolton's team right now.

Okay. Alright. So, Sean got close to what we were talking about on that call, but… so we'll go through that case study in a minute. I've got a, … a lot of stories here, or a lot of slides, and I'm going to go through most of them really quickly, because I know everybody on this call is pretty senior and doesn't need the basics of of the pre-submission. So… Of course, you know, pre-submissions are a way to enable FDA to give them early access or exposure to your technology and your ideas, get them thinking about maybe whatever you're doing that might be new or novel. And so they can be really powerful ways to,

Almost, have a safety feature built into the, your review. And it's free. Still, I don't know how long it's gonna stay free with the changing Medufas. I would be surprised if they don't start charging. So, you know, common topics that you talk about in pre-sub are, you know, your… usually your regulatory pathway and your strategy. It could be clinical, it could be your benchtop performance testing. And then again, the overall goal is, familiarization, especially for devices, novel technologies, to introduce the FDA, in advance. So, keep calm, remember, especially if you're doing a meeting, you want to make sure that you're not repeating yourself or repeating just the content that's already in the pre-submission. FDA has read it over and over again at this point, so that would be a waste of time.

You can't ask the FDA what do they think? They can't consult. It's more like you tell them what you think, and then get feedback. And then in the meeting, don't introduce new topics or questions, and don't talk too much. So, FDA never forgets, doesn't matter. I have had clients who did not tell me that they had done a pre-submission 10 years ago, and I found out while we were doing a pre-submission, now. So, FDA will know that you had a previous submission, and they will look at the feedback that they gave you. So, what we have found to be successful is to be specific in building your story, and keep it focused down to 3 or 4 substantial topics in a single pre-sub, because the more you throw at the FDA, the more resources they have to get in,

involved with And so your quality of, of background kind of, sinks, or the feedback sinks based off of, having too broad and too many different topics. And so with different topics, you know, I completely isolate for pre-subs. I might have one, two, or three pre-subs, based off of the strategy and the need for, advice in these areas. So, topic… I think clinical should always be stand-alone. Your more comprehensive or technical questions, the FDA's gonna spend too many resources and be spread too thin to give, in-depth feedback.

We find it very successful to actually quote FDA guidance documents, or standards, or feature-specific questions, directly out of those guidance documents. Because I… because then you put FDA in a position to argue with itself. So, Appendix 2 of the pre-sub guidance document is really powerful because it has a list of questions that the FDA feels like will lead to the most valuable feedback. And so this is where I have had FDA tell me. that question's not appropriate for a pre-sub, it's for a 513G, and I was like, no, ma'am, it's not. Like, if you would read, you know, your Appendix 2, question 5, it's, like, right there that this is appropriate for a pre-submission.

So… Anyways, so there's a lot of examples for… for… of questions out of just about any area. And so it's really… it's… Appendix 2 is the most helpful thing in the, In the guidance document. So that was the draft that was, in March 2024. It was just finalized in May. And so, it added a section on PMA 100-day meetings, and it clarified interactions outside of the Q submission program that could be appropriate for informal interactions, rather than having to ask those questions through the pre-submission. But, it gets confusing because it doesn't tell you how to ask or who to ask about that informal, review or informal question asking, and… or clarification meetings.

So, for that informal feedback, you know, again, if you don't have a lead reviewer like you would in a pre-submission. how… who do you even send these questions to? There's also no metrics for the response time for those informal requests, so we're still kind of seeing how those are going to play out. So, what we've seen, so far in 2025, So, FDA has a new quarterly MDUFA review, and so now some information is available, you know, well before that their annual, summary report. And so you can see how the FDA is trending on its metrics a little easier.

So, for 510K review times, they recently came out with this report that says, overall, 510Ks are getting cleared at 127 days, and that the … the average FDA days to review is 72, and industry is 55. So the pre-submission review times, they, they are pretty much on target for their goals, be it the written feedback or for, to schedule a meeting. So this is an area where I'm curious for everybody's feedback, because we have seen that they are on the money for scheduling those things, or providing them on time, or even early.

So, again, their performance metrics, … For her.

How many?

Their average days of 60… 61 for a 60-day turnaround.

That's great, Michelle.

Awesome.

Yes. Agreed, Michelle.

Amen. That's good. … So I just think that that's interesting with all, you know, all the turnover from COVID, followed by all the reductions in force, followed by the rehiring of all the FDA people that were riffed, like… that the data here, like, you would think that they're high… that this is the highest functioning FDA's been in a while. So, it's just an interesting… piece of information. So, our general impressions from what we've seen from the agency is that we're seeing the quality of their feedback, especially for written feedback, has improved. And we're not sure if this is because the employees are finally becoming more efficient and better understanding actual FDA policy, as opposed to making it in, like, an academic pursuit, like it feels like it's been in the past.

years. … But definitely, I think all the reductions in force have put a greater pressure on the reviewers. … over the last several years, we were getting really odd feedback, where they just wouldn't answer questions in the pre-sub that were clearly within the scope of a pre-sub, and they just said, well, that… if you want to know that kind of information, that's the kind of thing that you have to turn in a 513G for. And we were getting that. regularly. And then now, we have… I have not seen that question in probably a year of, pre-submissions, so… As we're overall seeing more, consistency in applications of concepts, and the use of reference devices.

So, some areas where we have seen and gotten odd feedback. are, you know, FDA defines this term, we all know, intended use and indications for use. But now, if you look at, especially people that are getting warning letters, there was one that came out just a few weeks ago to a company called WHOOP, That basically says that FDA considers, an inherent use, not just an intended use and indications for use, but inherent use could have something like a material or a feature, that implies a claim inherently in the market, in people's minds, or it flat out has, say, like, a claim, and they're trying to say something like blood pressure monitoring for wellness purposes, but the public knowledge, people know that blood pressure

You know, is really related to hyper or hypertension, and so therefore the public will use this wellness device to make decision… decisions about their medical healthcare or status. So, have you guys… what kind of… Feedback or, things have you seen where people are getting, pushback about this concept of inherent use? Like, I know in COVID, like, if you had an antimicrobial feature, that automatically implies in public knowledge that it has some sort of ability to kill You know, the germs, so… Has anybody else seen this?

Oh, sorry, go ahead, Jose.

I was gonna say, that makes me think of the whole episode with WHOOP, right? I don't know if you guys have followed that, but, you know, they're providing blood pressure, and they're like, oh, we're just a wellness and fitness product, but some of the stuff they're saying around blood pressure is sort of… Inherently implies that it's meant for medical, so just one example that comes to mind.

Exactly.

It's all around… it all seems to be around the wellness and the AI and the… And the… SAMDs. There's been a lot of… Untitled letters and devices that seem to be trying to get away with it, but they're… the inherent use for the wellness devices, they're really going after those claims, and even with the wound dressings, they, you know, they reclassified them as Class 3s now, so… antimicrobials, you know, now they're a Class 3, and I had one that went in, and they said, nope, now you're a Class 3, and you need to do 3 years' worth of You know, clinicals to even.

that it'.

through, so now, you know, it was a Swedish company, and now they're doing veterinarian.

wound care. That's so….

And it was a great device.

So this brings up the idea of there being this fuzzy line between wellness devices and medical devices, and you're seeing it in Europe with, the CRA, the Cyber Resilience Act, where they're saying devices covered under MDR are explicitly exempt from CRA. However. dot dot dot. There is this continuous….

Your untitled letter?

Yeah, there's this continuum of, you know, is it a wellness device, or is it… is it a medical device? And where is that line? And that line is moving thanks to… Technology.

My favorite.

grandmother.

when people are like, we want to be general wellness, and… but we think we… we fall within this CPT reimbursement code. I'm like, no, no. You're not… you… you can't put a foot in both of those worlds at the same time. And so people, like Carrie said, I think have taken a lot of liberties with trying to say that their general wellness And I find my clients that are adamant about their general wellness, like, you give them these boundaries in which they can be, In general wellness, and they just can't stay within the bounds. They just… it's like a scope creep of their indications in marketing.

They want the… they want the reimbursement, and they're like, until we get caught, we're gonna do it, and….

Yep.

We don't want to listen to you regulatory people, so….

The other thing is, when the wellness guidance came out, I was really skeptical that that was how it was going to end up, because it just seemed too good to be true.

The marketing people aren't reading it, nor are they gonna listen.

No, of course not. They don't care. They just want to get the product out and sell it.

Well, that, you know, that leaves….

This is… this is where….

Throw marketing under the bus, though.

No, no, no.

keeping.

Sean!

He goes….

Because you have to be….

Aware of label and labeling requirements. And so, when you do your labeling description, that's really where you have to be smart as to how you pick.

Shoot.

The wording. That's what's going to make a difference between you making a medical claim and making a, let's say, a therapeutic claim, or even making a wellness claim.

No, but it's not just the labeling. So, they can do everything right with labeling. It goes into their social media, aka their marketing approach, because Like, I have a social media procedure that everybody has to get anything pre-approved through. It's very simple, but so many people, they start to make… they drift, oh, I can make this claim, I can make this claim, and I'm like, no, you can't make that claim, because it's not in your labeling. And so. you know, it's twofold. It's being aware of your claims, your labeling, and the applicable regulations, and understanding you need somebody from regulatory to approve any posts that you're putting out there, because it's easy to drift.

Absolutely right. Any promotionals that you make, if they make medical claims, or whatever, they are part of your extended labeling claims.

It's a label.

website, social media, all those things.

A lot of people, my new clients will go, oh, no, social media is not a big deal, the FDA doesn't look at it. The first thing in January, my FDA inspector told us was, I was looking at your website last night and your social media posts. I mean, that was the first thing that came out, and I'm like, see? They're looking at this before they come onsite.

Hey, Michelle.

In Europe, you have the, … it's slightly different, the consequences of doing that. You have the claims article, Article 7 in MDR, IVDR, about claims, but, you also have, upcoming, the new liability, directive. And that basically says that if someone wants to make a claim for liability, they need to prove it, and to do that, they need access to the tag file, and if the tag file Doesn't materialize, or the tag file doesn't cover the risk it should cover. Then, the claim is awarded. And that last bit is what's going to bite if you make these claims, because then you may not have the right risk management done for the particular device, for the particular use that you have claimed in your marketing, but there you go. If someone suffers damage, whatever, you have a claim.

So that's.

Came is shut.

going to work. I guess.

Hey, Michelle, quick question, because I'm just listening to this conversation. kind of apparent to me, because my… I was on the engineering side when I was doing testing for medical devices. It seems to me that just this conversation right here kind of illuminates why so many medical device companies, especially startups. go sideways with the FBI because they don't know what you're talking about right now. Is this why there's so many recalls? Because the folks at these smaller to mid-sized, or some of the large ones, are kind of trying to skirt around the regulations, thinking they can sneak through stuff because of nuances, like… like inherent use, stuff like that?

I personally think the bulk of the recalls are coming from some of the largest companies. that are trying to… that basically, they say they're too big, like, they ignore FDA. Like, what is FDA really gonna do to Boston Scientific, or Phillips, or whatever, you know? And so, they have, like, just such bad design practices. It just, for the whole history of products, like, like, infusion pumps has been a train wreck for probably.

Excuse me.

10 years, yeah, exactly, exactly.

Okay.

And, like, I know….

I saw.

some people that worked on, infusion pump 510K that the FDA was like, okay, you guys have made too many changes. And while individually it looks like you could have written a letter of defile, you need to do a 510 now. Well, they didn't even have, like, the FDA in their review, asked for, like, some prints or some drawings or something. They didn't have any print drawings for their components. So I had to go make, like, a thousand drawings for this product. I'm like, how does that… how does it get made? If you don't want the drawing, you know?

Where's your design kit, then?

Well, and a quick question. The reason I'm asking all this is I'm working with an incubator here in San Diego called Aquilius, and they support… it's all medical device companies, and they're growing, and so I'm out there fairly routinely. Just to kind of support, this incubator to grow, and so this is the stuff that I'm… I like knowing from what you're… you're presenting today, so that I can hopefully share with these… all the startups out there to avoid what you're talking about now. And it just kind of scares me that they're out there trying to bring products to market, and most of them don't know any of this stuff. They're learning maybe the basics, and then they're just gonna hit that wall.

So it….

Exactly. I see more, like you say, those startups and small companies hit the wall, rather than the first thing out is safety problems. Like, they get themselves… they put themselves out of business, almost. With the claims that they're… they're making. So somebody else was….

So, I want to make sure that we have time to get through everything. Michelle, how are we doing in terms of the slides?

Well, we knew we were gonna get, Sidetracked in a lot of conversations here, so we're fine.

Okay, good. I will move on, because I'm also interested in people's, experience and reaction here. Okay, so… Before you move on, I want to make a couple announcements. One, if you'd like to be promoted to panelist, just raise your hand electronically. I sort of lost track of who I promoted and who I didn't, and I don't want to keep asking the same people. That's number one. Number two, next week, please join us for Ari Henkin, who's going to be, talking about AI for regulatory affairs. And the solution that he built in-house and is bringing to market, and it's pretty neat. So you're not gonna wanna miss it, and there are a lot of opportunities, particularly for consultants, to,

you know, engage with him and work with him, directly on introducing it to your clients, etc. You know what I mean. So, if you're on this call, you don't need to re-register for next week's call. You'll get an invite, you'll be let in, but if you want to invite new people, they should Go to medgroup.biz slash calls and register. Which will be updated later today. I'm still working on that janky interface, I apologize for that. And we have lots of people here, so let's move on.

Okay, so this is directly out of some, pre-submission feedback that we had gotten. And, you know, this particular question was on biocompatibility, and FDA came back and said. We know that silicone is carcinogenic, mutagenic, and repotoxic substance, which adds risk to the intended patient population. And we recommend… recommend that you, justify the use of this material with biocompatibility testing as you proposed. So… Oh, and then also the FDA specifically referenced 14971.

So… … Anyhow, so we kind of pushed back on this, on this statement here, in… in a follow-up meeting. And it turns out the… it was a… the doctor that was a consultant, like a physician, practicing physician, the FDA had consult here, and she's like, oh, well. I'm not really the biocompatibility expert. That was just my thought and concern, and so it's like, okay. Stricken from the record. You know, stay in your lane, lady. She was not the… and so, to me, the FDA lead reviewer should have taken the responsibility of saying, thank you for that thought, but it doesn't…

That's not our policy here, so we're not going to put in the pre-submission. So, don't be afraid to push back on stuff, especially if it sounds kind of off the wall.

reiterate that again, please don't be afraid to push back to your FDA.

reviewers and these Q subs, and get it stricken from the record. Yep.

Because otherwise, it's going to live with your submission all the way through. Say it again, Michelle, so everybody hears that.

If somebody… if the FDA gives you a feedback that seems like it's coming from left field, you need to push back, because you don't….

Alright!

You don't know if the person who even provided that comment Is an expert in that area.

And ask them where it's coming from, and provide the black and white evidence, because that That, right there, -

Yep. And something else I've had that's been kind of weird.

Don't be afraid to push back.

The FDA's like, well, we are aware internally of another product that does such and such… well, that's not public record.

name of the product, and that's not public record. Sorry, I just said the same thing you did, Michelle. Sorry. We're in the Wi-Fi together, Gary. They can't give you that information, so therefore they can't bring that up in your meeting without giving you the black and white evidence. Therefore, provide us the black and white evidence where that needs to… because that's in… That's irrelevant to our product, so it gives the scientific evidence where you're giving us that that is relevant to our product. I could care less about another company's product.

Yep.

It has nothing to do with our product.

So, we've had the case where FDA did not understand essential performance, which is a 601 issue, and it's affected multiple clients, and so my partner and I have actually sat down and trained FDA on the issue for EMC testing, and had about 50 or 60 reviewers in the meeting. most of the MC Safety Lab, plus some other people involved, so we had about 70 people on the call, and they actually agreed with 95% of the material.

Well.

I asked you.

back for other calls where the FDA does not know about that subject, as they will ask you back as subject matter experts if they do not understand that. And that has happened to me and other subject matter experts where we know more than the FDA, and they're unclear about that topic. So, don't be afraid to push back. Or….

I think that's a good thing, that they know when they need to ask.

And don't be afraid when you know when to ask.

Just say that. Yeah. I… I… having the connections with the right people and saying something to them helps, too, because it's not just pushing back, but saying something.

Knowing Wendy.

through the right.

And Leo?

Oh.

And Leo being one of the most well-known experts on 60601.

It's one guy.

It all works.

You know what?

We're not all gonna be able to, like, pull a connection at FDA and have 70 people show up to hear us talk about something.

Oh yeah, if anybody missed his… if anybody missed his presentation on 60601, type 606601.

Good God, that was a….

Yeah, that's.

That's all I'll have to say.

into the chat, and I'll send you the YouTube link. We have a new YouTube channel, people subscribe. MedTech Leading Voice Exchange on YouTube.

Oh, cool.

Okay. … So, another thing is I… we have seen, FDA using inappropriate clinical reviewers. So, we had a product that was a catheter for men, and the only people at 3… it was 3 clinical reviewers, all 3 were not only female, all 3 only practiced female urology. They did not have a male urologist or anybody that had treated male urinary issues. On the… on the panel.

So, and they, they said, well, we learned about, just because we don't practice. on men doesn't mean we don't know. We had to learn it in medical school. I'm like….

That was how long ago, so you absolutely asked for a… Specialist who knows about male urology.

Period.

Period. Or whatever your product subject is. They need to find that specialist.

Yep. And then… but they had their own clinical reviewer on the meeting, and all they said that was one of the world's leading, male… practicing male urologists, and they just basically said they didn't care what he had to say. Like, no, this just period, we think this is high risk. So… Anyways… This is another odd thing that we've gotten a lot lately, is that there, the multi-office interactions are really challenging, where, you know, in the past, you know, if you take the concept of, like, a Foley catheter that has a temperature sensor in it.

And this, this example is right out of the guidance talking about how to use a reference device. You have a Foley catheter that has intended use to drain urine. You have a temperature sensor that has an intended use of sensing temperature. Individually, they are their own cleared devices, but you can put them together, as long as you don't change the intended use or features of the Foley catheter or the temperature sensor. You can use… in that case, you can use a reference device for the temperature sensor. and a predicate… primary predicate of the urinary catheter. So that's a guidance document example.

But lately, we have seen anything that we have that is… it's been just as straightforward, technology-wise, of… and just as simple as putting two already cleared things that are already used together, together in actually the same device. And the FDA told us, no, if we have to consult with another division, you need two separate 510Ks, because we can't… We can't cross-share the review.

And that's new.

Have you used the inter-center consult process?

Do what?

the inter-center consult process, I mean, it was specifically put in place for combination products, but it also is used for things like this, so I'm really surprised that they're not making use of it.

Okay. Can you write that in the chat? I have not, so I can… I wanted to look that up later.

Have you thought….

You, Scott Colburn about this, too?

Because I would think this is an issue he'd, be able to… Manage, too.

You think so? Because he's on the evaluation of the… the… tech… the recognized consensus standards.

Yeah, I know, but he has a lot of sway.

Okay.

He's a mover and shaker.

Okay, that's good to know. I have, … worked with him in the past, but it was, like, when I was doing… I was on some… General Feeding Councils a while back. Yeah, William County.

Michelle Daniel wrote a great comment. Mr. XFDA guy.

Okay.

Do not assume the reviewers are knowledgeable or experienced, and make sure your presentation is as simple as Possible.

Write it at a 5th grade level, right?

Daniel.

That's exactly my point, because sometimes we make these assumptions, and… When they know less than you do, they could become defensive, and then the meeting's not gonna go well.

Yep.

Because they're, they're, they're gonna be insecure, so, … This is what I have experienced, both inside and outside the FDA. The more insecure and the less knowledgeable they are, the more dangerous they become.

Well, the other thing is how invested they are. Like, one of my clients is struggling to get approved. They've actually had to challenge it, and the person on the reviewer is in their late 70s and has been in that same industry, and therefore the product that they're trying to get released directly competes with probably things he's invested in, and so that's what they're… that's the premise that they're challenging the decisions on the 510K. Because all the pre-subs were going along, everything was going great, you know, they were kind of iffy on some of the clinical data, but they're like, you know, everything was… all the precepts were positive, and then they got rejected, and they're like, wait a minute, this is not… this is not adding up to what we, you know, discussed.

So… Again, with the multi-office interactions. They said the procedures and probes are reviewed by different divisions within the FDA, the devices have different intended uses, the procedures are carried out by different medical specialties, and the two procedures have different risks. … And then… then they try to reference a guidance document on bundling. Or multiple inter… it was… it was just… it was totally in left field compared to what we were asking and trying to accomplish. And so it says, accordingly, please unbundle your IFU statement such that you submit a marketing submission specific to the use of the probes.

Which would be reviewed by OHT8 in a second marketing submission specific to the use of the, … The other part of the device. Which would be reviewed by OHT3. So that's the actual wording that they used. … And so, side note, I think I shared with you guys, a few meetings ago, I had worked with a client earlier this year where they make, … It's, like, basically vibrators. And FDA came after them and said. You are advertising these for vaginal use, and the product code only doesn't… it doesn't specify female, and the only devices that have been cleared for this are for males.

And so, vibrators are illegally marketed if they don't have a 510K, because they're intended for males and not females. Don't have to comment on that one. Just wild.

Yeah, that was my facial exp- wow.

Hold on. On that call, I told them, I said, as a woman, not even as a regulatory professional, as a woman, I am offended that this is where you are spending your time and resources. Of all the things that are wrong in women's health, of all the things that are hurting women, this is the very limited thing that you are focused on right now, and there were 6 FDA people on that call. And I'm like. Anyway….

I think we need a stand-up comedy, because it just… it reiterates something I've been talking about with my team, is that FDA individuals are human. We've put them on… and I don't even want to call it a pedestal, we've elevated them to be all-knowing, and the reality is they're not all-knowing. You have to go the extra mile to be confident in what you do know, so you can challenge appropriately. But they are human. We have to keep in mind, they're human, they're human, they're human.

At least they're fallible. So we've also gotten some, weird feedback about clinical trials, related to the OUS market. And so, they said, you propose to conduct, the clinical trial sites.

Outside of the U.S.

Comparison of the user population and clinical practices at your study sites to the U.S. provider population and standard practice. They need to do that. And then, in general, we recommend that marketing submissions that provide clinical data from OUS studies include at least 3 sites. at least 75% U.S. enrollment to account for differences in population and clinical practice across different sites for each clinical investigation conducted with OUS data. that 75% of your data has to come from the U.S, and that for that 25 other percent, you have to do this in-depth comparison about the population clinical practice.

Ronald? Comment?

I've heard this a lot. Just saying, I've heard this one a lot.

Well, ….

Human factors, where it's 100%.

I work with clients, and they say, yeah, we're going to do a study in the US, and I say, are you sure? Because The whole healthcare system is so different that you may end up with completely uncomparable patient population. it really happens, and so I'm very careful in those locations. Having said that, we just had a very good study done in Africa. And the data is used in, in, … and the Notified body had no problem with accepting that for Europe, because we, we, we managed to, to… extrapolate those data. That worked very well. But it's… yeah, but what I see in Europe is that you can just justify… if you can justify it, it's fine, but you must be able to justify it.

And having something like, okay, you must have 75% blah. Doesn't make sense, because you cannot compare that. It's… it's… yeah. No, I don't think it's practical. But, I mean, yeah, you have to live with that in the US over there. My own problems, but, ….

Okay, so that's… that's the bulk of my slides. So, what are some of the other regulatory challenges that you guys have seen where you're like, wait, what? Like… Well, we're doing that now.

So back to the clinical trials thing, getting ready to do a webinar with another… with a CRO from Europe on, clinical, trial expectations country by country. Like, where you should do your clinical trials in Europe and why. Gonna be pretty interesting, excited about that. … Side note. Sorry to interrupt.

No? So I'm just looking for….

That makes me really curious, Sean. Because there's so much… You can say about where to do a study, … There's so many angles to look at.

Yep. Well, you know, this is based on experience, so… You know….

I doubt it.

… Interesting. … On the same call. where this idea originated from, we were also talking about, some positive things that you've seen come out of the FDA recently. … you know, I think it was Tiana that said, you know, I really have to give her credit. And I think you sort of covered it in the Review Times. To a certain extent. Where have you or anybody else on the call seen, the FDA excel? Or, you know, their performance.

moving back a little bit more towards, you know, kind of being policy and objective, whereas before, I felt like we were just talking to a bunch of people right out of graduate programs that never worked in industry, and so every pre-submission was like an academic debate. instead of an objective assessment that was based on both FDA policy and, you know, a real predicate comparison. They were often hypothetical land. You know, most of the time, so….

And drink?

andre DiMino: Thanks, Michelle. You always provide great information, appreciate that. We've got a situation we're working on right now where a client has brought us a product. andre DiMino: That was previously in a pre-sub in 2023, but that company stopped all their effort on it, and now they're picking it up. And the question is, should we be referring, or can you refer back to the original pre-sub response from the FDA, and just continue on, even though that was from a different company?

You… I think you'd have to, explain the relationship there. And then, yes, I would think that you should be able to do that, but even explain that in the cover letter.

andre DiMino: Great, okay, thanks.

So, I'm gonna take a minute and ask Jose a question. I, I am negotiating a contract with a new customer who is insisting on using all of their agreements, and no offense, Ronald, they're from Europe, so the things that they're asking for are, like. in, like, a service agreement. I'm like, that… what are you talking about? Like, I can't agree to that. But one of them is, like, they asked me for a copy of my cybersecurity policy. for… for how I handle my client data, and I'm like. we don't have one. I said, but if you want to send us yours, I'll take a look at it and see what we… what we may already be doing that complies. And so, like, what are your

A, have you heard of anybody being asked for a cybersecurity policy like that, like a service provider like us?

Oh, yes.

I personally haven't seen, but I know we have one. And I mean, I think in your QMS, I think it's recommendable to have you know, both how are you implementing cybersecurity for the products that you develop, but also what's your internal cybersecurity policy? Because, you know, if you're not training your team, for example, like, keep in mind, like, I always tell people, like, the number one source of real vulnerabilities in cybersecurity is humans. Right? You know, you can have all the encryption in the world, but if you don't train your team on how to properly see when somebody's attempting a phishing attack, then you've got tremendous vulnerabilities. So you want to have,

you know, an SOP for your company, and how you treat data, and how you, you know, what controls you have in place, and that sort of thing. Absolutely. It's kind of funny, like, we never get asked, maybe just because people assume that we, you know, employ best practices, but I recommend to companies to have that in place, yeah.

I just had a funny thing.

This is….

Like, what if I asked AI to write me an internal cybersecurity policy?

I, I think….

The bigger the company, the more likely you're going to get asked this. We've been asked by BD, Medtronic. J&J, every one of them, when this got busy, they came to us and said, yep, here it is. Your big guys, your little guys typically don't, but it's, it's… it's….

I've had major….

companies.

I've had mediums do it, too, because I have a design house I work for that, is… I'd say not even medium asked for one, and I've had multiple Fortune 500s ask for ones, too.

If you're doing business with a Fortune 500, they're gonna ask.

Yeah, and I have multiple Fortune 500s I work for.

That's been our experience. If you place capital equipment, medical devices, in hospitals. They're going to be asking, and it's got any kind of connectivity, they're going to be asking you about information security. And that's a little bit where I see the whole noise about cybersecurity is missing things completely. You can have a device without any connectivity at all. And you should have information security for that device.

Will somebody share theirs with Michelle?

Yes, that was exactly….

I was gonna say, can you share it with me, too? Because I'm going to be writing mine.

Yeah, but mine's such a small company, we have a lot of balance. we… we don't feel like….

It's a lot of outs.

I don't lie.

Okay, we need to help.

Exactly, exactly. Because I'm not designed anything.

We don't design squat.

Yeah, I hope y'all take.

The risk that they worry about is… it's funny, because, like, just to go full circle to what we were talking about at the outset, … you know, AI, for example, right, and all your large language models, you can set them up to protect your data, or you can set them up so that your data trains the models, right? And so, if you have a policy in place then you have a means of showing that your team follows proper procedures to protect your data that way. But, you know, you don't have to be a design house. If people are sending you their proprietary information, and you don't have proper controls in place and training for your team, then there's a decent chance that their proprietary information is going to get leaked to the outside world. And that's what they're looking to protect against. So it's not about design, it's about, like, they're sharing sensitive information with you.

What assurances do they have? That you're gonna properly manage it.

Right. A cure is intellectual property, right? So you create a SOP for someone. That is intellectual property. And then someone in your company says, oh, that's good, I'll take that with me. So they stick a USB stick in the drive. People don't lock down their USBs enough. They should just be locking down, so people can't take data. They'll still put it across on OneDrive or something, but….

I share this… I share this story from, like, 10 years ago, where an accountant Clicked a link that automatically transferred $40,000 out of the business account. an accountant, and it was like… okay, and it was a phishing email. It was, like, something they shouldn't have clicked it, and thank God they had insurance, because they got the money back, but it's like, it's that simple. Having that policy, don't open an email that you don't recognize. It can be just that basic.

Absolutely, yeah. I heard a CISO of a major hospital network saying that 90% of their cybersecurity incidences are phishing, right? So they're just phishing attacks to doctors who aren't properly trained on how to sniff out sensitive email sequences. And by the way, that's an area where, you know, AI's having huge impact on cybersecurity, because one of the primary means of attacking people is through human, you know, human engineering, like, knowing how to get people to give you information that they shouldn't be giving you, and AI is excellent at writing persuasive copy and stuff, you know? Those emails that you used to get that you could immediately sniff out because they were so poorly written are now very well written and structured, and they can pull personal information off.

They look like financial stuff now.

Sure.

Like, all the bank stuff. We bring some fundraising, just as we talked about financial. Good timing, Sean.

So….

I'm like, Siri, I just, I just listen. So this is a reminder that, Michelle and Renee are going to be doing this. Michelle, if you want to say anything about it, go right ahead. And, also, we're gonna be… … sorry, dude, dude. Just a reminder of next week's speaker, Ari Hankin, is gonna be here, presenting. the solution and case studies, practical strategies for regulatory. The following week is going to be, Marcus Emney from Huden, and the week after that is going to be, someone from Clarimed is going to be speaking on human factors.

How many of you, feel like you are experts in human factors? Anyone? Usability testing? Okay, I expect to see you all there. And that's… You still… Don't know what's happening on September 19th. Working on it. Anyway, I want to thank all of you, especially Michelle, for showing up. Even though Tiana was supposed to be here, and … Doing a great job. Thanks for leading us in a very interactive discussion. I enjoyed the back and forth. I'd love to hear your feedback on today's presentation. Feel free to send it to me, you know, through any channel that you choose.

And I will wish everyone a happy weekend. Cheers! Ball.

Happy Friday!

Yeah.

Bye.

Have a great one.

weekend. Bye.

Thanks. Bye, man. Thanks, Sean.

Thank you all.

Nice job, Michelle.