Who Decides What Evidence Is Enough? The Reality of Medical Device Regulation

00:00:00: small groups of people coming together can change the world.

00:00:03: In fact, it's only thing that tends to change.

00:00:23: Have a wonderful career with many different facets to it from, you know medicine to law.

00:00:29: To being regulatory advisor...to being professor and an overall great person.

00:00:35: I really look forward to our conversation today touching about a lot of topics clinical trial regulations especially with the focus on medical devices And what's happening?

00:00:45: You have a lot insights to share because you talk With many different sides From students academics regulators.

00:00:52: I really look forward to our conversation today.

00:00:54: So let's pick up there, tell us and especially also the audience a little bit about who is Tom Maven?

00:01:01: What's your background?

00:01:02: How did you get into clinical research?

00:01:04: And what's your general perspective informed by

00:01:08: most?".

00:01:13: first became involved in research eleven years ago when I moved from being a hospital doctor, i was a junior doctor on the wards and started work in the irish regulatory body called HPRA or equivalent to beefarm here.

00:01:28: It was very interesting time because medical devices were still regulated under directive system.

00:01:33: so i remember The very first clinical study i asked for.

00:01:37: review involved bronchoscope and liquid nitrogen with the hypothesis that if you freeze the inside of the lung, it'll grow back beautifully.

00:01:48: And I guess from seeing some very interesting technologies and some planned protocols... You know?

00:01:55: I was a medical officer responsible for reviewing the clinical aspects.

00:01:59: I essentially caught a bug for getting into figuring out how do we best frame research How to make sure protects patients and appropriately translates technology.

00:02:11: and it's a topic I guess eleven years later, i'm still looking at.

00:02:14: So that is kind of how I started At the time working as junior doctor and planning to take year out.

00:02:23: so life was not planned.

00:02:27: but because my background in law well as medicine once joined HPRA found very interesting there were an opportunity sets of perspectives.

00:02:40: when looking at a technology, you have to make sure it's compliant with the relevant legislation and rules that apply.

00:02:48: You also want it to be good quality research so it draws into a sense on both those kind of domains.

00:02:53: And how come we started out in devices?

00:02:57: As a doctor was there already a tie or connection to that?

00:03:00: did you get experience with devices because typically be more exposed to the drug side, even though we have much more use of medical devices.

00:03:08: But it feels like they're flying a little bit on the radar maybe?

00:03:11: So what was that transition?

00:03:14: opportunistic because there's a position available?

00:03:16: or how come you don't get into devices?

00:03:19: You know I guess in later years working at the hospital i actually worked with medical gerontology and medicine for the elderly.

00:03:25: so its not very medical technology focused specialism.

00:03:31: So it probably wasn't from my clinical day-to-day work.

00:03:35: And, you know because I arrived at the agency with a fairly open mind that thought well try this and see if its interesting or not go off to do something else.

00:03:46: but once i got in started looking for files think about technologies under evidence.

00:03:51: found fascinating was nascent field arguably still is so when I would have joined our regulator.

00:03:59: Uh, i would've been the only medical officer looking after medical devices and then I'll have a team of four medical officers in senior medical officer responsible for area.

00:04:08: so it was small but growing field with very interesting.

00:04:11: you know Very new technologies um And It was an opportunity to work across.

00:04:19: You Know number of different life cycle stages essentially.

00:04:22: You'd see the very early things going into clinical studies, we also would have an opportunity to look at vigilance or incident reporting and trying how do authorities appropriately assess safety in the post market.

00:04:34: So you got to see technologies right through their life course, so I found it fascinating but i didn't come from one of the more traditional fields like orthopedics or cardiology where your hands on and then planting on a daily basis um But still had a good appreciation for general medicine and ensuring the best overall outcome per patients.

00:04:53: And I think medical gerontology is very nice when that regard because it really is focused around what is best for that individual patient.

00:05:02: And many of the choices you had to make was, looking at the frailty of the patients or their goals for care?

00:05:09: What would be the appropriate thing?

00:05:11: You could take every patient down and open her chest and replace your heart valve but from any patients probably wouldn't have been a better overall decision.

00:05:19: so trying to figure out what's the best thing for an individual patient was a nice thing I learned working with gerontology.

00:05:27: Do you think from your, let's say background as you joined the competent authority was the medical side more informative over how you approached working for a regulator or the lawside?

00:05:41: Because both are clearly prominent in your position there.

00:05:45: So

00:05:46: where did

00:05:46: you check balance

00:05:48: with Medical devices?

00:05:50: something I hear even these days From current assessors is that When you work with medical devices, your far more reliant on the legislation.

00:06:00: Uh when you talk to a medicines assessor they're often far more interested in data or methods.

00:06:06: but in the medical device world it's often trying to find out if whatever their methods are that have been prepared do they meet an often reasonably generic legal standard.

00:06:15: so we haven't much greater focus than the Medical Device World interpreting and understanding the applicability of the law, then guidance that sits under the standards.

00:06:27: It was one where I didn't think i would need a lot quite so much but it turned out to become part of my job within about a month starting as a regulator in March.

00:06:41: by April or April I went to Brussels because the MDR is still being negotiated Again, I guess the background in law was helpful because i got to join some of the Irish delegation for European Council meetings where they were having their end negotiations with the regulations.

00:06:56: So that's a really interesting experience and when I was leaving law going into medicine it did come back but turned out very useful.

00:07:07: Yeah yeah...I can certainly imagine at It's interpretation of text written, which is very essential to law.

00:07:15: Which you have to apply on a day-to-day basis how much of that especially in your early days was still let's say you negotiating with yourself as You are sort of Confronted with the interpretation off third party, which has the person approaching the regulator versus yeah How much like and how much did they evolve into?

00:07:40: You having a general, like deep understanding of regulations and therefore sort of being sure in yourself.

00:07:47: And projecting an outward and setting boundaries how did that evolve for you?

00:07:52: I think it took about two years from me to just understand How the actual architecture Of The medical device system fits together.

00:08:01: or remember when i went For my interview To become a regulator um...I didn't really know what notified bodies were Or What they Did government body did.

00:08:12: So, you know medical devices has a complex governance ecosystem which we could talk about for many hours.

00:08:21: but understanding how all of those cogs work in order to deliver things out the other side of the system that was I guess helped by the legal background.

00:08:31: But it did take about two years to properly understand How do different parts fit together and and the entire framework.

00:08:39: And maybe also for the audience, can you briefly sort of from your perspective touch upon the interplay between these players?

00:08:46: So you mentioned competent authorities like we have in Germany, the Mayfarm for example... ...and the no default bodies which are non-governmental before profit companies that at end of day issue C certifications.

00:09:01: how do they enter play?

00:09:02: where is the differentiation which is the responsibility of which and who informs whom to which extent.

00:09:10: Yeah, so essentially if we think through how a device translates from prototype true-to-a-marketed product or true trials two marketer products into the postmarket So national competent authorities have the role of looking after pre-market clinical studies And with the MD order that came they had responsibility for some Post Market Studies as well.

00:09:32: So that's one of their main jobs, is looking after non-CE mark devices going into clinical studies.

00:09:37: Then as you say for market access it goes from the twenty seven competent authorities in EU member states to about fifty five notified bodies.

00:09:47: and these are private organisations.

00:09:52: some operate on a profit basis many grew out national standards bodies we had.

00:10:00: They look after the market access and when you get to the post-market there's a bit of a dual responsibility.

00:10:05: So if something happens, let say it is an incident here in Germany You'd report that to the German authority or you could report it to the manufacturer And then periodically um...you reported notified body or sometimes you also send individual incident reports to notify bodies.

00:10:21: So it's clean in the pre-market for studies, and its cleaned from market access.

00:10:26: but there is some overlap within post-market.

00:10:32: Was that a struggle for you also as you joined?

00:10:34: I mean, you said it took your two years.

00:10:36: There's the reason why it took you to here because... So

00:10:39: those things are reasonably clear.

00:10:41: but understanding the governance framework how does guidance get created or how do standards work?

00:10:48: these both have their own ecosystems too them.

00:10:50: so For the medical device and IVD regulations we've over hundred guidance documents now.

00:10:57: So the way that's prepared, it is governed by the Medical Device Coordination Group.

00:11:03: That group meets periodically and chaired by the European Commission.

00:11:11: They would have about maybe twenty or slightly more with revisions happening... ...about twenty policy officers in the unit of Brussels.

00:11:19: Their job was to bring coordination amongst all these different actors.

00:11:22: So you have the Medical Device Coordination Group.

00:11:24: You also have a Notified Body Coordinated Group, that was new thing with regulations.

00:11:30: Then another group called the Component Authorities Medical Device Network which is more informal collaborative way that authorities come together Which would be somewhat similar to the Heads of Medicine's Agency Group for Medicines.

00:11:43: And then you have lots of other policy actors in the wider ecosystem.

00:11:47: So right now it's very active because there is a proposal to revise that two sets of laws, so then you've got the Commission and Parliament and Council all very active... ...and under different groups on an international level.

00:11:58: but the organogram starts to explode out as we start to

00:12:02: do this.

00:12:04: I think he painted a vivid picture where takes someone even from inside for years or just unravels Who to talk, who should give guidance or direction here and there.

00:12:13: And one of the things that you mentioned is we have a tremendous body of guidelines documents as well in your role sort.

00:12:25: So let's say from an outside perspective, guidance documents especially everything that is published by the MDCG are sort of seen as law for us.

00:12:35: It not formally form a regulatory point-of-view.

00:12:39: being in EU regulation or embedded one and not even directive it very clear adherence to these guidelines anticipated unexpected I'm assuming this same perspective is shared by competent authorities and it's helpful for you to have that additional guidance to refer back.

00:12:59: Yeah, so under the directive framework we might've had something in a region of twenty MedDev guidance documents.

00:13:06: So things expanded a huge amount And were very different from the medicines world.

00:13:11: In that if want try deal with topic like AI in health product regulation.

00:13:18: A group like European Medicines Agency could write concept paper you know, here's the things happening and how we need to think about it or a reflection paper.

00:13:26: And then build towards a guidance which would be the auditable standard than in the medical device world.

00:13:31: as you say everything has generic Medical Device Coordination Group nature to us.

00:13:36: so within those MDCG documents you'll see questions and answers.

00:13:41: documents are different kinds of thing.

00:13:43: We only have one type formatted document that is published sometimes like I remember when we published the old MedDev document for clinical evaluation, very important guidance.

00:13:59: It's from twenty sixteen but a lot of it is still applied today.

00:14:03: We had a huge amount debate at that time about how does this apply?

00:14:07: Does this apply now immediately or would they supply in maybe five years' time?

00:14:14: so there was an interesting thing where could become auditable and you could have findings possibly raised against it.

00:14:24: But at the same time, people after incorporated into their quality system and enter next clinical evaluation update on that kind of thing.

00:14:31: so what does exist in a somewhat gray zone?

00:14:34: How these things go out and get implemented under timing?

00:14:39: The other challenge I think with the MDCG documents is that they're often themselves as result of debate and compromise.

00:14:49: So in a similar way to when you're preparing legislation, it becomes an negotiation amongst the three legislative bodies commencing council and parliament.

00:14:58: Even when you are making guidance that often become a negotiations.

00:15:01: say we'd like that wording or prefer that wording.

00:15:04: sometimes its almost left deliberately vague but can be really challenging thing for end user because their left was requirement is hard.

00:15:13: Do I need to do a study like this or maybe not?

00:15:16: To a study and write a justification.

00:15:19: So, you know it would be better for the system if we could have likely fewer guidance documents that were more focused at the applicability was clear up front.

00:15:32: but i think the mindset drives a lot of the MDCG guidance documents is to further interpret.

00:15:39: But I think a different change of mentality where we focus perhaps more on discrete technologies or discreet methodological challenges, that would be nice way to focus some of these guidance and maybe adopt the framings our medicine colleagues have.

00:15:56: We're in bad need for things like concept paper for things like AI, medical devices.

00:16:02: So we need to try and better figure out how to build guidance that works for the users in everyone asked to implement.

00:16:07: it's such a polycentric system... ...that we really have to make sure that the thing coming from top are as clear they can be.

00:16:17: Adding a layer of complexity especially I think for the medical device world.

00:16:20: on top next to the MDCG one of the institutes publishes very relevant guidance is ISO.

00:16:29: So not only is the ISO-Forty-One for five standards, according to which we should conduct clinical investigations.

00:16:35: Which of course are relevant for how even gather evidence from our devices and you have numerous guidance also on quality management systems... ...how to manufacture

00:16:46: etc.,

00:16:47: And then for indication specific guidance.

00:16:49: in addition How does that interface with level off guidance or requirements set forth by regulators in the way that you have observed it.

00:17:01: What's the interplay with the ISO group?

00:17:03: Yeah, they're quite distinct and ISO standards would have a different form of application in that the application of standard is always voluntary.

00:17:13: You know this was the original conception when I first wrote to directives That you could apply as standard voluntarily or not.

00:17:20: if you don't apply the standard your supposed to apply something at least equivalent To what's in the standard.

00:17:26: I guess a lot of my work tends to focus on clinical evidence and testing in humans.

00:17:30: So as part for research project, um... In the last couple years we went and mapped out all the ISO documents with.

00:17:38: you know there's something like six to seven hundred documents in the ISO universe And they have their own hierarchy.

00:17:43: so it is an ISO standard.

00:17:45: You could have technical specification A technical report or guidance document and probably other things Like uh..in United Kingdom For example publicly accessible standard, which could be made with BSI but by an independent drafting group.

00:18:01: So there's a whole universe to these kinds of ISO documents.

00:18:06: I think that they serve very important function for many areas and things like manufacturing control toxicology or biocompatibility and things where they are detailed and quite rigorously applied in the sector.

00:18:20: But my kind clinical interest when you look across two hundred or so medical device standards that we would have, only a handful of them would have technology specific clinical requirements.

00:18:34: Like really it's only a hand full out at the two-hundreds.

00:18:36: So you know We don't have clear clinical evidence rules for things like heart valves or coronary stents.

00:18:43: It from the MDCG universe?

00:18:49: We've been very slow in the medical device world because there are legislative instruments.

00:18:57: So another type of thing you can write is a common specification and these were prepared for devices like non-corrective contact lenses or liposuction machines, Because the way MDR was written said that they had to have those common specifications to market product.

00:19:13: but we haven't developed them any other general medical device yet.

00:19:17: but it would be a really good way to try and clarify, harmonize things like clinical evidence requirements when that's important for things such as heart valve, coronary stent or continuous glucose meters.

00:19:31: But we haven't managed to get there yet in the system.

00:19:35: So how is the need of clinical evidence derived?

00:19:41: That is one key question!

00:19:45: you know, I got to experience as a regulator the shift from the directive to medical device regulation.

00:19:52: And one of things that was most challenging for those applying rules is lots of legal requirements changed when we went directive to regulations.

00:20:04: there wasn't single definition left untouched.

00:20:06: every definition change clinical data equivalence all these kind.

00:20:10: key thing.

00:20:12: But did the evidence standard improve significantly?

00:20:16: There's no evidence that shows it has gone up or down, to be honest.

00:20:20: We did some analysis in an EU funded project called CoreMD Coordination of Research and Evidence for High Risk Devices And we did systematic reviews across cardiology orthopedics and diabetes In general.

00:20:34: what we found is that amount high quality pre-market clinical studies was quite low.

00:20:40: So, you know even for things like high-risk devices and cardiology only a very small fraction had a randomized controlled trial either at all And even less of them had one before CE marking.

00:20:52: so This is the real challenge.

00:20:54: If you're a product developer what?

00:20:56: You have to do as figure out The language in conceptions of MDR which is usually based around your intended purpose.

00:21:04: What does it that you say this is going to do And then you work backwards to a research hypothesis for how can we justify that.

00:21:11: But, That's very different way of clinical trial methodologists usually think.

00:21:17: So it does take some time For people working on new product development To figure out the MDR conceptions Of how evidence should be framed and what is exactly study I'm going do?

00:21:28: That was real challenge.

00:21:30: The system has been designed in such ways.

00:21:33: It allows freedom to developers.

00:21:36: we'll do an observational study or a non-oferiority study, maybe experimental studies with control group.

00:21:43: But that sometimes feels challenging for the developers.

00:21:46: even though they have freedom They're often not quite sure.

00:21:49: well is this sufficient?

00:21:51: Will this satisfy the notified body?

00:21:53: will it help us on our journey towards reimbursement and someone paying for a product?

00:21:58: And I think this is one of the key challenges in Europe when we compare ourselves to United States, and I hear this all the time from working at university.

00:22:06: You can go and meet with the US FDA and say here's our protocol outline thinking study like this that were going to put it in here year-in-year and we're hoping for abroad indication.

00:22:17: And you FTA would say That's fine.

00:22:19: or but if he are here her hand here and give your broad indication.

00:22:24: So you can get that clinical surety at an early stage, which is vital for new product developers because it's strengthened their funding pitch.

00:22:31: And the challenge in Europe was we haven't developed sufficiently or pre-market scientific advice.

00:22:40: I hope that's something that will grow.

00:22:45: We mapped this last year and found it.

00:22:47: about two-thirds of authorities when we surveyed, got results from nineteen out of twenty seven told us you can have a pre submission dialogue or talk to the regulator but weren't able find any case where you could have detailed scientific discussion.

00:23:02: agree on an end point for study design aspect so currently left unaddressed one are most vital areas support.

00:23:09: you need to translate your technology with greater confidence.

00:23:13: So I think it's a real opportunity for further development in our system.

00:23:19: Why do you think we touch upon bunch of topics?

00:23:27: Let us leave evidence behind for now and let's briefly touch upon.

00:23:30: you mentioned sort of the opportunity to exchange with regulators in good opinions.

00:23:35: I think one of the elements we touched on earlier is, We have such a system where your trial is competent authority Your approval goes into notified body And then postmarkets are little bit of pick-and-choose or lets say multiple players involved it not so clearly distinguished.

00:23:52: Now In Europe You can have something.

00:23:56: Well, let's say that you can have a framed exchange with the notified bodies.

00:24:05: Yeah sure it said You Can!

00:24:07: It took a long time to really implement and there is lot of hesitancy... ...it feels also in conversations with notified bodies at times in attendance of conferences.. ..and if we discuss them they are not so sure where their line between advising too much staying to their role as controlling body that has been assigned to them.

00:24:32: What is the regulator from your perspective?

00:24:34: The regulatory intention of what the notified body should or shouldn't do and why, like...what would be best avenue to bridge this conversation vacuum which at moment we really struggle with in Europe?

00:24:50: Yeah yeah no thats a great question and fundamental issue.

00:24:53: I think So, structured dialogue is a wonderful term.

00:25:00: It was introduced in twenty-twenty two and it's defined as a dialog to improve the efficiency and predictability of the conformity assessment process.

00:25:11: Sounds nice!

00:25:11: A wonderful definition.

00:25:12: I'm not sure if we're any more clear on that now than i've given you the definition but your exactly right at the parameters.

00:25:19: what can discusses to be minimal acceptability or clinical evidence set is unclear.

00:25:24: I think in practice, some notified bodies will give you a nod or wink or some form of steer.

00:25:29: um there is hesitancy?

00:25:31: because uh You know?

00:25:33: There has been problems in the past where Notified Bodies would do things like offer to certify our product just at client meeting as very early stage and they'd say well we're going to claim equivalence today.

00:25:46: should that be okay?

00:25:47: absolutely fine Just sign up for us then We'll get this done a month or something.

00:25:52: So the challenge with structured dialogue is that there's a hesitancy because of the innate nature of the client-service relationship, that a notified body has at a manufacturer.

00:26:05: so this natural hesitancy of trusting that kind of process?

00:26:10: I think if you want to make this whole early advice work in complicated EU ecosystem people sometimes talk of a single source-of truth, you know.

00:26:25: Could we at least make sure that the advice being given is consistent around this system?

00:26:30: I think things would empower it as a way documenting and sharing outcomes in these structured dialogues.

00:26:37: if happens in silos its going to be very difficult orchestrate but our systems kind drawing us into silos currently.

00:26:44: so what could happen is you can have structure dialogue might be minute meeting but if that's only checked every five years by a designating authority who are mostly auditors and the knowledge is not shared back anywhere centrally, we're going to be able to grow collectively get more refined in describing these things.

00:27:03: So I think figuring out some way to have single source can at least monitor or oversee the kind of advice happening with this system would very useful thing But i'm not sure on current implementation table.

00:27:17: One of the other challenges we have to remember though is that The medical device regulation Could be described as being embraced by subjectivism.

00:27:26: So if you say I Have clinical data providing sufficient clinical evidence, right?

00:27:31: And if a notified body accepts that then it's compliant.

00:27:34: Mm-hmm.

00:27:35: But it means that we have a system where product A may go through high quality clinical study, a randomized controlled trial and go for CE marking.

00:27:43: And product B may claim equivalents that have no clinical evidence in Go For CE Marking if notified bodies accept that it becomes compliant which is just very different and conformity oriented way of thinking about devices compared to you know our friends in the medicines world where safety and efficacy as much clearer standard with more rigorous set up methods and study designs needed get you there.

00:28:08: So I think there's two challenges.

00:28:09: It is the polycentric system where if these advice things happen in silos, recorded in silos we can't learn so those silos would have to be broken but we'd also have to introduce appropriate methods that people are not just giving advice from top of their head.

00:28:24: That sounds good, do it.

00:28:26: We should have a more methodological framework and I was involved in some work against that CoreMD project.

00:28:32: but we did try to produce recommended study designs specifically for high-risk you know therapeutic implants.

00:28:39: so there are the workings of this starting material.

00:28:42: get us into another methodological place in the advice we give.

00:28:46: But i think theres alot implementation work needed in the system.

00:28:51: I Think one big challenge is but technically one that should be manageable to overcome because it's the same for other regions in Europe.

00:29:01: We have such a vast difference between individual devices, they all need to be regulated by one piece of regulatory... guidance and framework.

00:29:15: And in all of that, we need to address also the high-risk ones which should have a more structured approach to how evidence is generated?

00:29:23: Yeah!

00:29:24: Probably generally half but will come to this topic I'm sure today.

00:29:32: But majority devices on markets fall into lower risk groups sometimes or largely not requiring any evidence, which there is ample reasons.

00:29:45: Not to if it's like a syringe or the status cope.

00:29:50: you shouldn't have to run study at some adherence to technical standards as enough and then you come into this gray area.

00:29:58: so in that polar opposites of an implantable long-term device versus a stethoscope It's clear that somewhere between those are border where You should start having evidence don't put on the market.

00:30:09: Now, even within individual devices and you mentioned this earlier a key differentiator is what do you claim.

00:30:18: So when we need evidence for it's what to have on your label?

00:30:21: And that means if the device is technically the same or should only be part of something else.

00:30:31: so given complexity there are opportunities structured enough guidance to determine the level of evidence, or do you think that topic cannot be approached so generously?

00:30:47: Because it feels... The letter is the current let's say default with which we approach.

00:30:52: We need to have device-by-device decisions unfortunately by a group of well fifty five different companies and in there yeah thousands of people working So off course will have differences.

00:31:03: Yeah

00:31:05: exactly there will always be intra-assessor differences.

00:31:11: You know, even in a refined system like the medicines world people have disagreements over.

00:31:14: is that method okay?

00:31:15: Is it acceptable?

00:31:16: and that kind of thing?

00:31:21: I think when you consider broad array technologies exactly as we said i tend to think about three ways very much.

00:31:29: as described high risk therapeutic implants should subject to rigorous pre market clinical studies I think if you talk to anyone and ask them, If they're going get a heart valve or neural stent would like that tested in others before

00:31:47: it's

00:31:47: implanted as new product?

00:31:49: People will almost unanimously say yes.

00:31:54: But we have system where applied the same generic requirement to wheelchair and stethoscope has their heart valve.

00:32:01: They all must clinical data providing sufficient clinical evidence.

00:32:05: so i think your right for certain high-risk therapeutic areas, especially in growing fields where you might have two or three products available but another twenty and a pipeline.

00:32:16: I think they're important opportunities to try get minimal clinical evidence requirements clearer because that is were we are taking the greatest risk of patients with quite significant uncertainty.

00:32:30: So you can run into significant public health challenges and I saw lots of those in the directive era from the post-market with devices.

00:32:39: When you get into the middle zone, so here we might be talking about things like let's say an insufflator something that distends your abdomen when you're having surgery does not need a clinical study?

00:32:52: You could have lots of debates around it or getting to debate on what claims they make but its legitimately difficult question where conceptual approach alone wouldn't solve it.

00:33:04: You would need a somewhat of a conceptual approach, but someone has to make the crude policy decision at the end of yes or no clinical study or not?

00:33:11: Or small clinical study?

00:33:14: and we like you're saying We don't have a central place to get that sense within our system currently.

00:33:20: And for the low-risk devices they can proceed to market generally without clinical studies.

00:33:25: But there are some exceptions.

00:33:28: You know, I remember...I saw a product once.

00:33:30: It was class one device but it was marketed for people with acute airway obstruction.

00:33:35: There were number of products that the numbers still are available.

00:33:38: But there is a plunger one or those ones look like a shovel.

00:33:41: you'd kind do a Heimlich abdominal truss to maneuver on yourself.

00:33:47: And they were often widely marketed as Class One.

00:33:53: You know, a huge risk attached them.

00:33:55: Right?

00:33:56: For the plunger one for example if you have an acute airway obstruction The first thing to do is stand up and bend forward So your not going to lie flat on the ground And get a plunger.

00:34:07: These were marketed with no clinical study Or they might've been trialled in a non-clinical or cadaver model But had a huge public health consequence If it didn't work.

00:34:18: Sometimes there's rare exceptions but that three ways of thinking about medical devices is the right one.

00:34:25: This is exactly.

00:34:26: how do United States figured this out in nineteen sixty nine?

00:34:30: So back in the nineteen sixties, there were increasing medical devices making their way to market under U S but with no rigorous regulation and place.

00:34:40: that was rules from nineteen thirty eight.

00:34:42: what they didn't have much.

00:34:46: What the US government did at that time is they asked some independent clinician scientists appointed by The National Academies to do a report to Government.

00:34:54: That was called A Cooper Report and a guy called Theodore Cooper wrote it, And thats why we have three risk classes now.

00:35:00: because he said if you're class III You are treated quite like medicine.

00:35:04: Class II should comply with standards in Class I General Controls Register and Repair Documents but no pre-market evidence.

00:35:12: So i guess... a way that helped the US system to be set up in much more proportionate ways than technologies advanced.

00:35:21: And if you think about the arc of last hundred years, it was only from late nineteen fifties when the highest risk for class threes started coming into widespread clinical use.

00:35:33: In a short couple of years between nineteen fifty-eight and early nineteen sixties we had the first fully implantable heart valve The first pacemaker and the first hip replacement implant that was successful, The Charnley Implant which is still available today.

00:35:48: And that really changed the requirements of regulation because we went from very low risk scopes on equipment to implants with therapeutic function.

00:35:57: I think in Europe We have try catch up with simple delineation To make things clearer for everyone.

00:36:03: But our challenge in Europe Is having same generic requirement and proportionality as implied.

00:36:09: That can make thing's own clear.

00:36:13: And do you think sort of like a collective growth off decisions?

00:36:16: You talked about the decision making behind closed doors being one other problems.

00:36:21: Also there again, as I think it's a little bit more transparent if you go and look at how devices are registered on the market... ...you can get the five-ten K report for example.

00:36:30: that makes quite clear what level evidence was presented.. ..how was it argued?

00:36:34: where an agreement therefore made is their.

00:36:40: Is there any discussion or idea from a regulatory point of view to just say, okay sure we might have high range different devices and we cannot pre-specify everything but once make determination.

00:36:55: And simply this determination is widely applicable so that everyone who falls under the category has more clearer ideas of the requirements that need to be fulfilled and we can sort-of make better standard without needing to, let's say, sort or predict what type of devices and what types questions will receive in future.

00:37:23: What kind of claims would you really see from the future?

00:37:25: We could have it semi reactive but just accessible like here is your database with decisions made.

00:37:33: I

00:37:37: think there's a couple of interesting points within that, Andreas.

00:37:43: The first is you mentioned the five-ten K summary or the authorization for high risk class three device in United States and You can get it and see the evidence submitted an evidence accepted by the regulator.

00:37:57: When we think about middle risk devices That's important because a lot things under five ten k you claim substantial equivalence.

00:38:03: you are don't have any of your own clinical evidence But it comes down to the nature of that demonstration of substantial equivalence and your preclinical testing.

00:38:13: So, to help with the middle risk devices I think there's a lot good we could do by validating appropriate non-clinic methods And as a world of researchers working on this from in silico things To other ways evaluating devices.

00:38:30: That is one thing where transparency is important because If you're introducing a product in the US, you can search and see has there been something authorized or cleared?

00:38:40: And the basis on which it was done.

00:38:42: You could essentially do the same thing then if your quite similar product In Europe.

00:38:47: when you try to do that It's much more difficult.

00:38:51: The reason for that is we have some rules of transparency.

00:38:55: So high risk devices Class trees are certain class two products that administer or remove medicines.

00:39:03: They can have a summary of safety and clinical performance, but there's no the Udemy database that is supposed to be hosted as not functioning completely for that yet.

00:39:14: I know they're doing some work in the background.

00:39:16: But um i had a master student look at this last year And we were only able find about eighty Of those summaries available by going To companies asking directly or looking on Udmy.

00:39:29: So we have transparency written into the law, but it's very difficult to access.

00:39:35: Um so that makes it quite difficult to do a similar thing for Europe.

00:39:39: what you could do in the United States um...so uh..you know if we think of the medicines world then and compared out to devices You get a full European public assessment report And within that you can actually see often quite detailed findings.

00:39:58: regulator accepted this trial data, but didn't like this data because of this methodological reason.

00:40:03: And that gives you an even greater richness of knowledge about where the water marks are regulators.

00:40:10: and if we want to really empower translational developers thinking of clinical evidence in Europe You have to be able to show them how similar things I've gotten through the system.

00:40:21: So over last year trying to promote the idea that we should make The Clinical Evaluation Report publicly available.

00:40:29: For two reasons, really I guess.

00:40:31: first is a principled reason in which every research participant has put their hand up and enrolled on study maybe for their own good but betterment of others.

00:40:42: so when thats principle it generates all evidence within your clinical evaluation report.

00:40:50: But the other thing is that it's a real boost for new product developers, because they can see where do watermarks lie.

00:40:56: And I would also help regulators then to see if watermarks are wildly different in one place or another compared.

00:41:05: We didn't win that battle.

00:41:07: The clinical evaluation report will not be public and with the revision proposals to medical device regulation we'll have less of those public summaries, the SSCPs-the ones for patients are proposed to be removed entirely and summaries of safety in clinical performance for things that meet the new definition of well established technologies won't need a public summary as it would have less publicly available information.

00:41:33: That kind of information is really important to product developers.

00:41:36: And when I talked in you, Product Developers it's one of the first things they do was check the five ten K clearance or checked us FDA data sources because that helps them a lot and setting overall boundaries

00:41:47: exactly.

00:41:48: yeah It's One Of The First Steps.

00:41:49: You go ahead and see.

00:41:50: what has someone done?

00:41:51: Let I could use as inspirations or as guidance To give certain that took uncertainty.

00:41:57: Yeah Because at the beginning you just don't know.

00:42:01: Sure, I think it's clear enough vast majority of cases on where you categorize yourself in terms of a risk level that already gives an indication how much needs to be done.

00:42:11: And then you start looking at other devices and that space?

00:42:14: Yeah maybe one other point on that.

00:42:18: when you go the US and read their summary safety & effectiveness data for high-risk device its laid out very clearly.

00:42:26: so someone who is new to regulation can the preclinical evidence, the clinical study data.

00:42:34: When we looked at those summaries with the master's student what we found is that because of the way that summaries of safety and clinical performance are set up in Europe you can only extract it directly out of the clinical evaluation which means and that's a very difficult thing for your average clinician or certainly if you're patient to understand.

00:42:56: It is just framed quite differently, right?

00:42:59: And it can lead to documents there could be as short of two pages or as long as two hundred pages where you might find the summary for a coronary stent will have maybe twenty pages dedicated telling about what atherosclerosis is.

00:43:13: That's not what you need.

00:43:14: What do we need here are preclinical evidence.

00:43:18: so I think As well access these documents.

00:43:22: We have some important work to do, I think yet.

00:43:24: To make sure that what is provided Is readable and understandable to a clinician or patient And i don't think we're quite there yet.

00:43:33: So hopefully it's something That could be improved in the future.

00:43:38: Let's maybe move one step earlier still In The journey and let's talk A little bit more about evidence?

00:43:43: We touched upon the topic early on.

00:43:45: Then we jumped into how we even can get guidance And understand the framework under which we should gather It.

00:43:52: And I think this is one of the key elements also where we see a huge difference between drugs and devices.

00:44:00: When you start development for drug, let's say generic societies or biosimilars... ...where maybe bridging studies are good enough there was clear expectation to go in and basically have two but mostly three studies before your product on market.

00:44:21: what purpose each of the stages serve.

00:44:24: It's very different for devices, and you also said that we find limited evidence or randomized control trials which is highest level of evidence that can generate from a scientific point-of view on device side.

00:44:39: One thing sometimes makes it difficult to say these are devices best comparator for you to identify.

00:44:52: this seems a bit easier with drugs one.

00:44:54: You have in the lot of cases an opportunity.

00:44:57: simply say we use a placebo and so what's your take from regulatory perspective on, just to frame it, we're talking high-risk devices.

00:45:11: We are not talking randomized control triads for status quo.

00:45:14: I'm sure that's an agreement and this is what you were talking about.

00:45:19: but the higher risk device can have a high impact where they don't really have standards.

00:45:26: How do you see the approach of evidence generation in the concept of medical devices?

00:45:32: And i think we should touch upon the topic non-functional or no treatment groups, how would you suggest to medical device manufacturers out there?

00:45:45: To start thinking about the level of evidence that they can employ?

00:45:49: not from a scientific view.

00:45:50: Not necessary for my logistics because always faster and less complicated than better.

00:45:55: but if you wanted it generate good evidence.

00:45:57: sometimes there is real logistical or ethical hurdles.

00:46:01: what will you lead into the field?

00:46:04: I

00:46:05: think for practical purposes, this often comes down to the individual developer and where they see their product being first sold or perhaps first reimbursed.

00:46:17: And that then makes them work backwards say how do we hit these targets appropriately?

00:46:24: So you know, often to get your CE mark You're typically dealing with an observational study.

00:46:31: sometimes you have an experimental study with a controlled group, but it's typically an observational study.

00:46:36: There are other types of ways you could set up your studies—you know?

00:46:42: You could perhaps do non-inferiority study and when you look at these things purely methodologically the appropriate methods are often quite clear.

00:46:52: so if you're introducing technology to already busy field non-inferiority, but with some additional benefit like a faster surgery or you know, faster recovery time.

00:47:04: Or something like that and otherwise you look for superiority um.

00:47:08: But these are kind of ways of doing comparative studies.

00:47:12: when you have just an observational study And you just want to get your CE mark Um methods can be quite variable in my experience.

00:47:22: So depending on if you'd just want To get your ce mark That's usually the kind of route You go.

00:47:28: We have to remember, as you say safety and efficacy in the medicines world is very different from safety performance or sufficient clinical data.

00:47:37: In our worlds.

00:47:38: but then when we go towards reimbursement this standard relative clinical effectiveness which is a jumping much stronger than efficacy even because it's safe and has efficacy on real-world population.

00:47:52: so there an evidence chasm.

00:47:55: we have to figure out a better way, firstly avoid research waste by doing observational studies and then having another study for reimbursement purposes or something like that.

00:48:06: And when does it matter most?

00:48:09: For the high-risk devices with minimal set of clinical design requirements you should have an experimental study in some scenarios if its new or first in class.

00:48:22: I think In our system, we're kind of in middle ground.

00:48:30: There are guidance documents now and further legislative change coming for things like breakthrough devices where if you read the guidance that was produced at the end last year or look at proposals into law it implies that these certain devices can have less than average but don't quite know what the average is.

00:48:53: so it's quite difficult to try and figure out what is the appropriate study design for a breakthrough technology where there might be significant unmet need, unlimited alternatives but still at high risk in unproven technology.

00:49:08: There's tension between wanting to have early access to promising technologies versus having clinical surety that will actually do them some good.

00:49:18: So these are essentially policy questions where we could all have personal opinions, but it's going to be policy.

00:49:24: I think that will drive us.

00:49:26: But the methods that were going to apply... ...I think a lot of work has been done to figure this out and like our previous topic, transparency is a huge part to play in trying to help bring that sense of shape or consistency to it.

00:49:40: if you don't know what being accepted into system as strategies how would further grow and develop?

00:49:47: So I think it is a real challenge and there's a risk that could be persistent for years to come.

00:49:53: Yeah, so probably have an answer about what the design you should have but essentially because this system is subjective its up-to-you You have freedom to select it But comes with the tyranny of uncertainty will actually work or answer questions.

00:50:13: When we think about those studies with comparators and sham interventions, you know this will be done perhaps for things like renal denervation let's say.

00:50:23: Where you might want to have your intervention group where you apply energy through your renal renovation device maybe a control group were you punctured the leg but not applied the energy.

00:50:36: these do raise legitimate ethical challenges because I guess, of some experience that I sit on the National Research Ethics Committee for Medical Devices in Ireland.

00:50:46: So I get to see studies that pass through and there occasionally are studies with sham controls in them?

00:50:53: When you're looking at those study designs... You're interested in first potential patient going into this study so they might have a fifty-fifty or some chance intervention or sham control.

00:51:05: It's making sure their aware about it.

00:51:07: And then secondly, if you end up in the sham group it's about ensuring that The kind of intervention that you need to do to blind us To the investigator or the person who'll see them after the intervention.

00:51:21: You need to sufficiently blind this.

00:51:22: but you have to think very carefully About doing something like an invasive procedure just to fulfill a sham requirement.

00:51:31: So is quite difficult.

00:51:34: But I think sometimes when there was no other way to answer That question we do sometimes have to apply some degree of risk-to-the-sham group.

00:51:42: But the patient, their prospective research participant has to be fully informed about that and it should often in a scenario where there is no other way you could generate this necessary knowledge to help with the question they have.

00:51:57: So its essentially an ethical question The difference between being a regulator or thinking about ethics for these devices.

00:52:05: When you're a regulator, your worried about its conformance with the law.

00:52:09: And when working in an ethics committee You are interested into value of that research and what are the obligations or risks that perspective patient might have?

00:52:18: So it's kind just two different ways to thinking about this.

00:52:24: When I started working at the Ethics Committee a bit like a regulator, I thought we'll be citing bits of the Declaration of Helsinki and looking at rules.

00:52:32: But really it's more about values.

00:52:34: so you don't have to know The Declaration Of Helsinki off by heart You just think about the consequences for patient.

00:52:40: with that kind mentality This might very valuable research.

00:52:43: but is what were asking if patients are too much here or Is there better way?

00:52:49: So its different perspective

00:52:54: especially I think when we're talking about sham devices.

00:52:57: and why even if you are well-intended manufacturer, not saying that they are driven primarily by speed to market which there always will be pressure for.

00:53:08: That's clearly acknowledged i guess at the end of day.

00:53:18: perspective, sort of like I don't know it's a well-researched group.

00:53:21: We know what happens if you don't treat the patients.

00:53:24: or can we have like no treatment groups versus our group?

00:53:27: Is that really necessary to move with the

00:53:31: chandeliers?".

00:53:31: And we've had studies ourselves where it was sort of... Like as he said kind of like If You Need To Puncture The Patients Or Something That In and Of Itself Can Instill Boone Healing Activities Cell Regeneration sort of damage to the tissue could incite some level of regeneration, which might have an impact on outcome data.

00:53:57: And then you're like yeah it's difficult we need to go a sham.

00:54:02: so let say treatment without active device group because at least procedure needs to be conducted perspective of the claims that you are making is, classified because of, let's say intermittent stay in the body for twenty-eight plus days but maybe not fully long term multiple years already you are a higher risk glass.

00:54:59: If all those considerations mean that we need more evidence and at this point I have no expectations on side effects then it becomes hard as manufacturer to optimize away gathering comparator data because you can just say no control, I have data from literature or whatever.

00:55:19: Why do I need to subject half of my potential treatment group in an effective treatment when i don't get that additional information over the control group?

00:55:31: Here due to the nature of medical devices and focus on performance as you said to more subjectively discuss what's the body evidence we need.

00:55:43: It's natural that we will have a tendency of wanting to one move fast, but also if there is an opportunity really argue for it or try to argue.

00:55:53: To seek out An opportunity to present low evidence.

00:55:58: and even in the general conversation I can say We would always recommend push above-the-edge Suggest less than you think You may need because If you don't get pushed back promised too much already from the get-go.

00:56:12: Like, for manufacturers that's not a good strategy!

00:56:15: The strategy should be find something that regulators don't agree to because it means you have hit the edge and... That is the fine line that we should write.

00:56:23: so I'm not sure if this is annoying or if your regulator will understand.

00:56:27: but thats perspective of person on other side.

00:56:29: But i dont think its naturally meant in way minimize evidence.

00:56:38: When I finished work as a regulator and then went to academia, started supporting some very early stage startups or product developers.

00:56:47: You start understanding them much more vivid way.

00:56:53: things like the challenge of time delay.

00:56:56: so you know it's a time-delay for three months in regulation but somethings take us two years!

00:57:03: Two months is nothing.

00:57:06: You know, for a company where your overhead costs you might be spending fifty or hundred thousand the month and it takes another three months.

00:57:13: That's just an enormous challenge For the company.

00:57:15: in front of really early stage guys with early-stage funding It can be a survival type.

00:57:20: that challenge exactly enough.

00:57:21: things take too long.

00:57:24: So I think there are certainly prerogatives especially what this startup world?

00:57:30: Where they simply have to work within their funding envelope they have to answer a board who will say, oh you said it takes six months.

00:57:37: Why can't we do that in three?

00:57:39: So these kind of pragmatic things certainly percolate their way down and I don't know the extent which regulators get frustrated or not but one thing after finishing work as a regulator is we spend very little time focused on methods of study design when you're looking at a medical device study.

00:58:02: We got to engage with regulators last year and, um...we reported it in the paper in January but.. ..we've gotta put the regulatory system up on the examination table and have a look at it And I was really interesting!

00:58:13: We did our workshop again.

00:58:15: there's quite few authorities attended But their clear message went into terms of their responsibility.

00:58:22: Their job is to see, Is this product a fit one?

00:58:25: To be placed in human.

00:58:27: So their focus as often on determining was the extent of preclinical testing enough to justify that there's no more useful knowledge That you need.

00:58:35: together You can now justify exposing it to humans.

00:58:39: so that kind of way they see It.

00:58:40: so its much more a product protection than methods question and like we're saying earlier that means that you'll have applied methods generated evidence And then it's assessed by notified body.

00:58:52: So methods find themselves in something of a blind spot on the system.

00:58:57: But you know, I did get a deeper understanding In last years and academia seeing startups about pressures that they have to survive And implement things under the funding envelope They have.

00:59:09: See yeah i'm not sure.

00:59:11: does it cause frustration?

00:59:13: Not sure if there's enormous focus on methods amongst regulators.

00:59:16: when thinking of study You've tested it enough to justify its exposure.

00:59:22: But I think, you know... ...to help with that whole process.

00:59:26: because setting up a study and having the run-it for three or six or twelve months or whatever these are huge investments And they're often pivotal to company's development.

00:59:38: In my simple way of looking at new product developers.. ..and I interacted them as an academic so i haven't done this myself.

00:59:46: But what I see across all of them is that there's kind of four key strategies you need to translate.

00:59:53: You need your clinical strategy, your regulatory strategy, Your technology and business strategy And if you map those out There's a nice framework called Gates.

01:00:04: They have some tools for project management For these things.

01:00:07: If we look at examples It's not linear.

01:00:10: We've done this now and go into the next bit Often.

01:00:14: find Your prototype, your first version of the prototype test it and you have to go back in redesign.

01:00:19: There's a lot of things where you might do your First Animal Study No Good Back Redesign Next Animal Study.

01:00:26: So It is not linear process even though You Have These Strategies.

01:00:31: This Is The Thing That In Our European Regulatory System These Can All Be Incorporated.

01:00:37: But I Think We Still Have Work To Do.

01:00:40: Just Try And Make A Clearer For People Especially In The Clinical Domain Water the minimally acceptable standards and always go above it, but you shouldn't go too far below It all right.

01:00:50: But we don't have a home for these kind of things.

01:00:52: You know that's not really an ISO Standard or we haven't done that kind of thing through MDC G documents.

01:00:58: And you could do us true those common specifications?

01:01:02: But to justify writing one of them you have to have an absent or insufficient standard Or an identified public health concern.

01:01:10: so We haven't really got a forum yet that can make the decision, those criteria are met to write a common specification.

01:01:17: So it's something where we have to make this clearer because in my experience now when I sit down with new product developer and there might be hardware device or AI enabled one... ...we talk through the pathway into Europe and the pathway of the US And what i hear every time is oh will go with us then Simply because its clear for the bits around Clear you get an answer.

01:01:38: That is the vital thing that we have to try and address in Europe, I think.

01:01:42: To really support innovation?

01:01:45: You mentioned observation of data a little bit In terms of study designs when you talked about evidence.

01:01:53: One of things since you brought up AI before everyone... The only thing anyone ever talked about was AI.

01:02:01: A lot of talk around reword data, reword evidences which are also more of an observational nature exclusively of the observation in nature.

01:02:11: So can you tell us a little bit about where?

01:02:16: real world data, first and what can you define it for us?

01:02:18: How are we use as a regulator?

01:02:20: see real-world data revert evidence And What do think is its most appropriate use In The clinical Evidence which again Is A broad answer to give For a spectrum Of devices that We have.

01:02:34: but I would love To hear your perspective.

01:02:36: yeah Well, there is a formal definition not within our regulation or EU guidance but the International Medical Device Regulators Forum would have a definition.

01:02:46: So in general real world data that which has collected routinely through electronic health records are defined registries.

01:02:54: so it's means of passively collecting data and then gives you an opportunity to go on analyze later.

01:03:06: now For medical devices, this has greatest relevance for post-market monitoring.

01:03:12: Because if you think of the orthopedic world You could do a clinical study and you can run that to six months or year And you might enroll three or five hundred patients.

01:03:22: But If put same device with CE mark into high quality registry you have thousands Of outcomes and are collected every three months.

01:03:31: It's updated then get an annual report.

01:03:34: it allows you to see the granularity of let's say, two parts of a knee replacement implant.

01:03:42: Sometimes we might use different products or that can be used in orthopedics and allow us if there are any camouflaged poor product performers from these combinations.

01:03:55: so maybe overall is okay but using this way It allows you to see that very quickly and a good quality data.

01:04:04: So for post-market monitoring, it's really excellent.

01:04:07: I think there is great opportunity beyond post market monitoring.

01:04:12: when you have very good registries You can actually embed the trial in your registry And run into standard almost as well as GCP.

01:04:21: There are few examples like this but not many.

01:04:25: When we think of cost involved running pre-market clinical study versus the cost for registry, you could generate high quality data for affection of the cost.

01:04:36: But that's not quite empowered yet.

01:04:37: now what are the barriers to real world data from medical devices in Europe?

01:04:42: I think The first thing is that uh You know again comparing two medicines and then a lot today.

01:04:47: but For medicines you can have a mandatory post authorization safety study or voluntary one right.

01:04:54: um And these have the you know defined rules clear approaches and defined data sources, you know the Darwin network.

01:05:04: And that various something like a hundred eighty hospital electronic health record networks I can link to so that allows.

01:05:11: let's say during COVID we had some vaccines.

01:05:22: So you could run a real-world study and actually examine that safety signal very quickly using electronic health record data if it's of sufficient quality.

01:05:30: But in device land There's no one who asses the regulator for post market.

01:05:35: Study, the US can do at they call us a five two to study so No one asked for them?

01:05:40: So it's really only if the manufacturer wants to or denotify body encourages them too.

01:05:45: um It's an area may grow organically, but could grow a lot faster and better if we had a greater push from the regulatory side to define when should this study be required?

01:06:02: When should notified bodies set it as condition.

01:06:06: This is something that's talked about a lot these days because with changes in legislation, notified bodies will not be able to reassess advice within five years which would delete them.

01:06:20: That's likely to incentivize the notified bodies start asking for post-market evidence and setting conditions on certificates.

01:06:27: But if we let fifty five notify bodies do this in their own way, We'll have many different approaches.

01:06:33: So I think to harness The true benefit that you can get from analyzing real world data?

01:06:38: We need to set some simple parameters For the regulatory system when You ask it why To try make them more consistent.

01:06:46: And then a lot of work on the quality of data.

01:06:52: A couple things I would say there, firstly when you look to the registry world lots of registries were started by clinicians or clinical communities as a result of safety problem.

01:07:03: so they really grew after metal and metal with orthopedic implants growing cardiology for different reasons.

01:07:10: but in many these registries it might just be that you've had an implantable aortic valve right, product X from manufacturer Y or version whatever and some do.

01:07:22: So we have a lot of work to do.

01:07:23: understand what is the granularity of device information but in these registries?

01:07:27: And then we have to figure out how to help those who might use that real world data.

01:07:32: so there's a huge amount of work needed to figureout the feasibility of doing a real-world study either with registry data or electronic health record data.

01:07:41: where you Are these coded in this way?

01:07:45: Were they populated, and the way we'd have confidence then.

01:07:49: So that feasibility check is one where I think it's valued a lot more on the medicine side for real world data.

01:07:56: They put effort into looking at the feasibility of running a real-world study And we have nothing to do something like that from medical devices in Europe.

01:08:05: so unless we actually help manufacturers or developers To understand the watermarks for assessing feasibility going to be as confident, go and actually pay to access.

01:08:16: And then analyze that data.

01:08:18: so there's a number of quite fundamental things we have to do to empower real world data in Europe.

01:08:23: um To try and catch up with our medicines friends.

01:08:28: Yeah That's probably very true I think from all perspective.

01:08:34: A lot of the work happens In post market space Which is is probably minimally interventional studies, and their primary focus is marketing driven.

01:08:50: And I think this one of the differences between drugs and devices as well for a device that customers are doctors in lot cases because it's sort like an end user need to be happy with them.

01:09:02: if you get into hands they're more natural use it prescribe or whatnot in the post-market space having, let's say reword evidence like studies where you basically follow the patient through but a limited number and then go to sites that are not using your device yet.

01:09:21: You can use it simultaneously for marketing purpose.

01:09:24: so this is something we see more of.

01:09:29: And it allows for the data standardization that you also sort of like called forward.

01:09:32: more because this is sort of, like structural beta assessment.

01:09:35: It's not really strongly regulated to use as a device has already reimbursed there Because its on-the market if they devises reimbursed.

01:09:44: That makes everything little bit easier for the manufacturer.

01:09:47: so it sort of incentivizes a little bit data collection at a slightly larger scale under more real-world conditions, not really stringently like not really heavily restricted.

01:09:59: And can already give you a lot of indications on what preclinical data because the pressure off time money investment and pull to market cannot really.

01:10:09: It's probably a little bit too distinct from the level of real-world evidence that you had and it is again very sort like pointed.

01:10:16: You take, if you sell hundreds thousand units of your product then say let us assess one hundred or two hundred to three hundred of them in a specific study with people who haven't used this before.

01:10:29: That's your population.

01:10:32: Wide scale assessment in a more general European data space and we have those European level pushes, at least for regulators to have access to that data.

01:10:44: whether manufacturers will have the same.

01:10:46: We would have to see could enable definitely more knowledge for devices?

01:10:54: Would you say if this is some reality structured follow-up eventually, I'm sure it's going to take a little bit more time but eventually we have simply a more interconnected health space.

01:11:09: We know more about the devices.

01:11:12: naturally would you say that could also be an incentive for regulators?

01:11:17: To say... ...we gain so much insight so quickly!

01:11:25: With having that as a let's say post market committer because any sort of like signal detection would be very easy for us.

01:11:32: So we can lower the burden off pre-market evidence if we come to that space or

01:11:37: Yeah, well I think thus.

01:11:39: That is a desire that Regulators are policy makers.

01:11:43: what have?

01:11:44: That way allow devices come to market with some uncertainty.

01:11:48: but that's okay in the regulatory world If we have reasonably good way of monitoring at the most markets.

01:11:55: So you see this in, you know, The Breakthrough Guidance, the Orphan Device Guidance.

01:12:00: it is mentioned there.

01:12:02: I think It would be really great for the system if we were able to better implement that.

01:12:09: and You Know In addition to the kinds of challenges We talked about earlier And you know i think one Of the challenges in our regulatory System Is That?

01:12:23: wording and apply that as a procedure.

01:12:25: We're not yet thinking about the methods, or the evidence.

01:12:31: And to best utilize The potential of things like you know European health data space and Registries I'm better utilizing.

01:12:39: You know things coming from electronic health records As well as connected devices.

01:12:42: on all these things there's enormous potential.

01:12:46: But if we just look at it in terms of so are you calling us?

01:12:50: proactive post-market surveillance or a post market clinical follow up activity, or a Post Market Clinical Follow Up investigation.

01:12:58: That's the very surface level way of looking at it.

01:13:01: so I think we have to find a way to get more into methods data and evidence but its currently underappreciated in this system now.

01:13:11: So you know when we think about how did real-world data or big data that is possible for medical devices, we have to start forming interest groups linked to regulators.

01:13:28: That can actually help them with that.

01:13:31: there are some promising research projects.

01:13:34: I was involved in one innovative health initiative project.

01:13:37: so it's a public private partnership called Greg as guidelines For Real World Evidence Generation.

01:13:43: It's very interesting projects.

01:13:45: brings both medicines and device people together.

01:13:47: So there's really nice cross-learning that can happen in that kind of place, but I think one Of the nice things about that project is that it's looking to develop use cases and then share experience based on those Use Cases where we can go in and think about feasibility of a device study Sort of how they would think a feasibility for medicine study.

01:14:06: And then think of you know What is the best way to consider the methods?

01:14:10: Then now does a lot more information in the medicines world.

01:14:13: You can see that well, In this medicine's authorization they liked but didn't like the external comparator arm.

01:14:19: for some reason you Can find things out from Regulators reasoning.

01:14:23: we don't have that in Europe right so We're gonna Have to get people around a table and start talking this out.

01:14:29: But I think if The focus shifts More towards methods And away From the procedures because That's A very superficial level way of thinking Of us that Would Really Help Empower Us?

01:14:41: With all of these kind of technical terms, if you think in a very simple way.

01:14:46: You know pregnant women have cardiotucography done as like an ECG of their tummy right now those machines capture great data but sometimes they're coded and slightly different ways And there's lots of different providers.

01:15:02: If we had something simple Like a simple way of coding the same data on a way to pull it with just a few little bits of data from electronic health records start to look for possible early signals of fetal deterioration.

01:15:17: So there's enormous potential in these things, but that can only come true if we start actually thinking off the data rather than the procedure is applied to accompany.

01:15:28: and so you're just yet to get it at next level as hopefully I will comment on stage.

01:15:32: yeah

01:15:34: If he were lacking behind a little bit.

01:15:35: they are especially internationally speaking I mean, even the

01:15:40: U.S.,

01:15:41: if you just look at registration data they have already structured the drug site quite well.

01:15:46: Even there their device that is not as far You know.

01:15:48: we'll see this then to make it very accessible for them.

01:15:53: To really assess what level of data do you have This one step further If you go through ten different hospitals For example in Germany you will find ten different ways That data's coded and has records And sure everyone says like A level of, you know fire standard for how the health data could potentially be accessed.

01:16:14: But it's still a big struggle and we are not sort of like mandating strongly enough one system.

01:16:20: I have friends working in IT and hospitals and university clinics And they also say like yeah The first thing that says let's develop our own standards You know?

01:16:30: That is exactly wrong thing!

01:16:34: We have, I guess a tendency to on the one hand think that we are sometimes over-regulating here in Europe and it's complicated.

01:16:42: To define standard as widely applicable... ...we've noticed this in our entire discussion today.

01:16:47: with medical device in of itself The diversity is there And the diversity is they're in the health data space but somehow.. ..We need to put together and you need start at certain point.

01:16:58: If something has been established let us build upon it.

01:17:03: We have great European initiatives, so they've already started structuring large sets of diverse data.

01:17:11: A good approach would be to simply say we're just going to use this as a standard now and stop the conversations.

01:17:17: Sometimes I feel there's more appropriate reaction meaningfully moving forward rather than continuing conversation for ten years.

01:17:31: It is better to move forward with the inelegant, not perfect solution because it's better than doing nothing.

01:17:37: Absolutely!

01:17:38: When we think internationally about real-world data I do have certain strengths in Europe for certain clinical areas like orthopedics For example.

01:17:51: my colleagues who work on orthopedic and are involved in registries will tell us that You know, if you look at the Dutch Artiplasty Registry that would have over ninety-eight percent completeness.

01:18:02: Whereas for equivalent orthopedic registries in United States it may be sixty per cent completeness.

01:18:08: So we do have lots of really high quality data sources.

01:18:12: I don't think its'the Quality Of The Data That's The Challenge.

01:18:16: But It Is Applicability Or The Incentives And The Regulatory System That Seems To Be Where Its Just Unclear.

01:18:25: And for a company that just means, well why would I invest in analysing or accessing data when they don't really know if i need it.

01:18:34: So

01:18:35: regulatory

01:18:35: side has more work to do there and on different data standards outside of my scope of competence.

01:18:43: but from having worked with some methodologists you research projects.

01:18:50: They can do really excellent work on trying to standardize different data sources.

01:18:55: You know, they all mop the data or you try to apply a common data model.

01:19:00: that's huge amount of work in practice but it can allow you to compare disparately coded electronic health record data.

01:19:08: so It is possible.

01:19:09: But your right would be much nicer if we all had one standard not The world we find.

01:19:15: yeah

01:19:15: Yeah very true alright let maybe move to, you briefly mentioned already here and there breakthrough devices often diseases.

01:19:25: I know it's a really special topic And its one that since we are on the topic of regulations in The Impact then how can we sort like stay innovative?

01:19:35: What type of evidence do We need at all ties together and It significantly all of the topics we discussed today and disproportionately impacts often devices and also pediatric devices.

01:19:48: because at the end of the day, uh...the level of evidence needed can sometimes be even like or comes.

01:19:58: The level of evidences required if it stays the same for these types of devices is so much harder to together because there's few of those patients and its harder that information.

01:20:12: overall The return on invest is harder to argue for because there are fewer patients that will use the device over benefit from its use.

01:20:23: So what we have seen, and a lot of industry driven but general outcry over loss off legacy devices removed form medical device directive to medical device regulation Because a lot sometimes in off-label use, but also just on the market with less evidence under medical device directive were reclassified or had to meet higher evidence requirements.

01:20:53: So a lot of conversations around lack especially often devices and pediatric devices where there is simply no business case that keeps them with the end of the transition period.

01:21:08: That means that a bunch of devices were sort-of like pulled off the market, it depends some could stay on stock and so on... And then just retrospective view.

01:21:18: now looking into the future.

01:21:20: these there are still.

01:21:22: rare diseases would strongly benefit from available medical devices.

01:21:26: There is lot of pediatric diseases pediatric associated indications that would really benefit from devices, but we don't really at least not to my feeling have a very strong incentive structure for those economically less favorable indications.

01:21:48: To provide devices and also meaningful but sufficient evidence.

01:21:54: what's your take on balancing especially evidence versus risk, versus unfavorable market condition?

01:22:06: but the unmet need that we find because of exactly that.

01:22:11: How do we solve that conundrum?

01:22:13: Yeah no I think you lay it out perfectly.

01:22:15: um We had certainly quite some problems over the last years when we saw the MDR being implemented and this phenomena of disappearing devices.

01:22:29: I was working as a regulator and chairing the clinical working group back in October of twenty-twenty one when a number of cardiology Practitioners came to us via The European Society of Cardiology, and they said you know We've seen about twenty products.

01:22:46: You know clips coils stents of certain sizes that we use That's the companies have told this are disappearing And i still remember to call because oh, this might be a problem that we're... This may be the crest of a wave where getting hit with.

01:23:04: And there's very little things you could do as a regulator to turn that tide around.

01:23:09: and As you describe it.

01:23:11: but challenges were for general medical device manufacturers.

01:23:16: they trimmed their portfolios so They know internally if something is running into marginal or potentially loss making territory.

01:23:26: What that meant for the clinical community is, we used to have three different Rashkin balloons available in Europe before twenty-one.

01:23:39: Two of them were withdrawn from market and there was one remaining... ...and then they couldn't get through their CE certification.

01:23:46: so this left us in Europe with no Rashkin Balloons.

01:23:49: And a Rashkin balloon is something that's used newborn babies when they have congenital heart disease and you need to open the atria in their heart.

01:23:59: If you cannot get into the neonatal ICU, put this under ultrasound guidance... You must bring that child sometimes for emergency surgery which as my colleagues in pediatric cardiology would say is no disrespect of the night surgeons but it's not necessarily team lead is going to be in doing that emergency surgery.

01:24:24: And this was all because of products becoming unavailable, all the sudden.

01:24:29: So it's a very real problem which did result in guidance document and now some legislative proposals.

01:24:36: And when you talk about the incentives, it's very different to the medicines world where there can be exclusivity periods or things that are valuable to companies offered.

01:24:49: When you engage in pediatric investigation plans and proceed with marketing and that kind of thing into device world we don't have a carrot In a similar way.

01:25:02: So what the guidance and proposed changes to law would say is that you could become designated as an orphan device.

01:25:09: We have a definition now, we transplanted it in some sense from the US.

01:25:14: so there has be a device not expected used for more than twelve thousand patients per year in EU where there's lack of alternatives and unexpected clinical benefit.

01:25:25: If you get designated as an Orphan Device then can open out two new things with the revision proposal.

01:25:31: You get a fifty percent fee reduction and you also get advice and support, um...you know when we looked at those rashkin balloons uh..we wrote a paper on it and tried to explain to regulators what we think needs to be done because clinicians were not happy because that just meant they cared for their patients.

01:25:51: so was disrupted but when The cost in Europe for your initial CE mark was something like a hundred and thirty thousand, I think.

01:26:02: For this company and his product.

01:26:04: when you looked at the equivalent costs of United States that went through to five ten K process so light touch process but with a small business fee it's something like five thousand or three thousand euro or Something Like That.

01:26:15: So It Was A Huge Difference And You Could See Why Companies Would Say Well We're Deprioritizing The EU Market Now.

01:26:23: As A Result I think the fifty percent fee reduction will be a help.

01:26:28: It probably won't resolve things because, you know it's we still may be more expensive in some cases.

01:26:36: One of the other changes that's proposed is an interesting one.

01:26:40: You mentioned the transition timeline as well.

01:26:42: They're proposing now that if your designated as an orphan device and you could have grandfathering So if you are legacy so CE marked under directive And an orphaned device Your grandfather into new system.

01:26:55: And I think it's very interesting that grandfathering is only being applied to orphan devices.

01:27:01: We could have had a much clearer path over the last years if we'd figured out an approach for grandfathers in more widespread way, you know this has been done in different ways as legislation has improved when i think of US experience so The requirements for medicine just to show safety, but not safety and efficacy.

01:27:27: That led us to thalidomide And after thalidamide we said medicines need safety and efficiency.

01:27:32: But this landed as a requirement to the FDA where they were already busy with their day-to-day case work.

01:27:37: and suddenly They had to go back into a retrospective review of something like five thousand medicines approved from nineteen thirty eight to nineteen sixty two.

01:27:45: To see that.

01:27:46: does that work?

01:27:48: So Nothing happened under the FDA commissioner.

01:27:51: that was there when the amendments were passed in nineteen sixty-two, but a new FDA Commissioner called James Goddard came into post and he made it his first priority.

01:28:02: And you just reached out to the academies again.

01:28:04: I got a bunch of scientists informed a variety of committees on.

01:28:08: they actually got that retrospective review with nearly five thousand medicines done quite quickly.

01:28:14: That's how you appropriately tidy up the sector when you want to improve an evidence standard.

01:28:18: With medical devices, they had grandfathering that couldn't go back.

01:28:22: They didn't even have an inventory of all devices.

01:28:25: You know by nineteen seventy six?

01:28:26: They hadn't idea but not a full list.

01:28:28: So they needed to use grandfathering and with The Medical Device Regulation in Europe we didn't do a retrospective review or grandfathering.

01:28:36: We just kind of baked new requirements into the cake and expected everyone to comply.

01:28:41: So I think that just volume of new requirements being applied, where no one was quite sure.

01:28:48: Do we need to repeat testing or do something new?

01:28:52: We're not sure and maybe you would have needed more in the post market.

01:28:56: but all this evidence uncertainty is a main driver for companies who are going to invest on certain pathways anymore.

01:29:07: with these types We're at the particularly blunt end because, like you say to market dynamics and a number of alternative devices.

01:29:17: But also as you mentioned in fields like pediatric cardiology nearly two-thirds are the devices or adult device used for children.

01:29:26: Pediatric Cardiology would not survive without off label use.

01:29:30: So yeah it's been quite challenging topic And one of things that I've been thinking about more recently.

01:29:39: With the changes to the MDR we'll have orphan devices and breakthrough devices, but there still remain a lot of devices that would fall over the orphan threshold at twelve thousand.

01:29:49: That's didn't need to be protected.

01:29:50: so examples of this will be things like accessories for dialysis for children.

01:29:57: they'll go over the thresholds.

01:29:59: it is quite niche area with vulnerable market dynamics.

01:30:03: What we have suggested to policymakers through Biomed Alliance, which is a group of clinical associations Is that they include the third definition of pediatric devices?

01:30:13: And We find a way to designate and protect them even if They fall over the orphan threshold.

01:30:18: So we've proposed it.

01:30:20: when will see what Will happen?

01:30:21: but something like That I think would help protect other vulnerable products for the future years.

01:30:27: Do you generally See a concern with off-label use especially for these groups because at the end of day it feels unintentional.

01:30:37: It sort of feels like you're moving

01:30:41: requirements

01:30:42: away from the manufacturer and say, not our problem we'll deal with that We will outsource to another party one of the other fields, and we'll come to speak to it I'm sure soon in sort like next section.

01:30:57: We see that also with artificial intelligence new algorithm developed not for clinical use And user can decide if they want to use or not.

01:31:04: It's up to them.

01:31:05: so i-it feels very unintentional when at end of day its sort of caters through idea.

01:31:13: as a doctor you have best interest for your patients and so you're willing to take that risk upon yourself, but not the intention of this system.

01:31:24: So what's your perspective on off-label use which definitely has its place?

01:31:31: But as a replacement

01:31:41: The way the medical device regulation is set up now, it just has a tiny bit of wording on off-label use.

01:31:47: It's in the annex on post market monitoring and says that manufacturers shall monitor for systematic off label use and analyze their devices to verify if they intend purposes correct.

01:31:58: so somewhat fake.

01:32:01: you're right.

01:32:02: when product used as off label its responsibility shifts to declination.

01:32:07: something goes wrong.

01:32:09: The manufacturer can say, well we didn't market it for that.

01:32:11: So falls under professional liability which is a challenge for those clinicians if there's legal case or poor outcome.

01:32:20: I think that It's essential to protect necessary off-label use and simply just how whole fields would stop If you were to prohibit us.

01:32:28: And does draw an interesting tension because Regulators are product regulators.

01:32:34: they have no role in the practice of medicine.

01:32:37: And FDA said clearly, you know we are not responsible for the practice of medicine.

01:32:41: and then that opens out interesting questions around.

01:32:45: You know what's sometimes called in the US?

01:32:46: The right to try or people who have well-meaning and a potential technology But one that isn't approved in the appropriate regulatory way.

01:32:56: So when there is no alternative And at the off label use of it devices into best interest of the patient.

01:33:04: I think there's a duty too have that informed consent discussion with the patient or family.

01:33:10: You know, in the UK is one of a few regulatory bodies that has a policy on off-label use and they would say there should be documented into clinical chart which is nice idea but it's very much within the practice of medicine right?

01:33:25: But I think that regulators over responsibility if there is a device regardless of its label.

01:33:35: If something is going to disappear and if regulators can find out early enough, you can issue a derogation.

01:33:40: So let's say the product is essential but it's gonna lose its CE mark.

01:33:44: Regulators can step in place of notified bodies and give a derugation.

01:33:48: This is how rashkin balloons got into a number countries when they were needed But It's heavily bureaucratic currently so every country has different process.

01:33:57: In Ireland You have a manufacturer document And clinician documents for each patient case.

01:34:03: So companies can only support that for so long if there's many cases involved.

01:34:08: We need to have an appropriate way of giving EU-wide derogation, which is possible.

01:34:12: but it was underexploited and I think this has been one derogations so far under maybe a second one in the next while.

01:34:19: But i don't think it could be regulators alone who make decision.

01:34:23: they need support from clinical communities say yeah no This actually essential.

01:34:26: If we dont get at this bad things will happen.

01:34:30: We just need to figure out the appropriate way to have a forum where that can work.

01:34:34: No, we don't really have that yet.

01:34:36: There is the medical devices shortages steering group That was bought in post COVID as part of you know extended mandate at the EMA To Have A Group Looking At That and there Is Now A Requirement on Manufacturers to Give Six Months Notice In Certain Scenarios Two Regulators.

01:34:54: What's The End Effect Of That?

01:34:56: To clinical practice, we don't know yet.

01:34:58: So I think we need to be a bit more proactive about it.

01:35:00: And if you do get that six-month signal and If it looks like it's a really important problem Get clinicians to advise if it's essential or not.

01:35:08: That should be detected and derogated.

01:35:10: ii think We have to make that work A little better in the system

01:35:14: right?

01:35:15: It sounds Like It's sort of fancy initiatives and you can say that we have done something to address the problem because, y'know.

01:35:23: The problems are widely discussed so well.

01:35:25: but this group has this pathway here at end-of-the day practical implications is zero or close to zero.

01:35:36: So that doesn't make it any easier.

01:35:38: I'll

01:35:39: bring into Brussels.

01:35:42: Welcome to Brussels.

01:35:44: So maybe let's switch to AI, because that is of course a very relevant topic and it hugely impacts especially the medical device space.

01:35:56: Within I mean matter-of-years.

01:36:01: lot has changed in terms how we can diagnose in various different ways through medical devices, more data of our bodies accessible to us.

01:36:22: We can analyze that data better or more easily.

01:36:29: let's say we can help break down diagnostic processes on for a lay person and we have huge strides with image assessment.

01:36:47: Let's say white space in which artificial intelligence engages with our health and fraction of witches is designated.

01:36:53: It sometimes feels like a bad joke, you know And I don't know.

01:36:56: Like a large language model Wellness product and the medically wise walking through bars type situation.

01:37:02: You know it's.

01:37:04: it feels exactly like that.

01:37:05: Yeah How do you differ?

01:37:08: Like what your take on how we should effectively like how are we and how should we differentiate the space around, especially AI and Avid medical devices.

01:37:22: What is a medical device in that space?

01:37:24: Where should we draw our line?

01:37:25: where do we draw a line at the moment And what needs to change for consumer protection but also ensure there's continued opportunity for innovation within Europe.

01:37:36: We're not just left with a bunch of devices from China.

01:37:39: Let's

01:37:41: see,

01:37:46: yeah.

01:37:48: When we think of large language models in healthcare I wrote a paper with Steve Gilbert from Tresden here in Germany very simple paper that was published on Nature Medicine.

01:38:00: you know large-language models require approval as medical devices when they use them medically.

01:38:06: the way our regulatory system is set up it's all predicated around the intended purpose.

01:38:12: Now, if you have something that... If you ask it am I having a heart attack?

01:38:16: It'll give you advice or something along those lines.

01:38:20: That seems to be fulfilling the medical purpose.

01:38:23: You know diagnostic predictive whatever you'd like to call under definition of a medical device but they're not appropriately regulated as such.

01:38:31: and if were say let's take chat GPT bring true MDR.

01:38:37: there's lots of challenges in seeking to validate a model that has a black box nature, which is not capable of having the same output with the same input every time.

01:38:48: So... That leads often into discussions around guardrails and real-world validation.

01:38:57: Large language models are one thing but we have so many other uses artificial intelligence like you say an imaging that is now rolled out and implemented in hospitals, In many cases.

01:39:08: And I think one of the basic points i would make Is That if we Think Of Technology & Us Civilization Both of us change in a symbiotic way.

01:39:20: So before clocks people met and interacted in wildly different ways Like I'll see you tomorrow Whereas Now We've Clocks ,I can See You at Exactly One O'Clock.

01:39:32: A similar change to what happened when clocks arrived will have to happen, how AI has now arrived into radiology suites and hospitals.

01:39:40: What it would mean is that changes the way in which we can provide decisions or diagnosis for patients.

01:39:52: It also changes the way in which radiologists will interact with their images, and people talk a lot about de-skilling if you become overly reliant on recommendation.

01:40:00: And there's lots of legal questions around... A lot of these algorithms would not give you defined thing.

01:40:07: it is just advice or prediction but still left to do further interpretation because that's illegal liability.

01:40:14: wording worked into that.

01:40:16: So I think we have work across variety.

01:40:19: different communities understand what's valuable and understand what needs to change in how we think about the technologies.

01:40:27: So I think that, you know for things like mammography screening where let's say some countries... Like France from what i understand You need have three radiologists signing off on every mammography to ensure quality.

01:40:38: so an AI algorithm could go in and drive efficiency there if We have data of sufficient quality.

01:40:44: That's convincing to say.

01:40:48: So there's a huge amount of, it just means we have to think about these things quite differently.

01:40:55: And when you're thinking in the regulatory sense about them.

01:40:58: and I think if a mammography screening tool You can build into some of these models an offender button.

01:41:04: so human and algorithm disagree.

01:41:07: That should be flagged on.

01:41:08: that should go back to company for analysis send.

01:41:11: someone should decide which is right?

01:41:15: radiologists or the algorithm.

01:41:16: that has come to a different decision here.

01:41:19: So we need to figure out.

01:41:20: firstly, clinical evidence is often quite poor as it's a number of you know analyses that have been done.

01:41:28: You can go through the list I think nearly fourteen hundred AI devices on the US web page.

01:41:32: Or try and find ones in Europe.

01:41:34: but you know clinical validation In terms of perspective study is quite rare.

01:41:41: so lot these things are prepared on training data, in some form of tuning data and rolled out.

01:41:48: And then the regulatory world.

01:41:51: how do you put a?

01:41:52: what's your pathway for clinical evaluation?

01:41:54: For that I think it is likely to be variable amongst notified bodies.

01:41:59: Some will be happier say clinical data in classical sense not deemed appropriate.

01:42:04: so go via article sixty one ten no perspective clinical study.

01:42:08: maybe other notify body have different perspectives.

01:42:13: But yeah, it's one where I think the basis of regulation is to ensure that product is safe for people using.

01:42:21: So again just like with the implants we have to get in and figure out what are those critical decisions or recommendations?

01:42:28: Where we should have rigorous perspective clinical evidence.

01:42:31: so you know That would be things Like if You Have Automatic Hemorrhage Detection For A CT Scanner In The Context Of Stroke.

01:42:40: Vis AI Would Be a Company That's exemplary of that.

01:42:45: You know, there are scenarios where you really do need good quality evidence to justify implementing this safely and for lots of other things it is going be much more case dependent.

01:42:57: sometimes you can get quite excellent data from things like imaging data banks.

01:43:03: so if your doing an imaging device in a data bank then things labeled effectively you can get quite good evidence from it.

01:43:10: And for those kind of examples or areas we need to figure out, is that an appropriate validation for its context-of use?

01:43:21: But I think our challenge in Europe currently is that we're figuring out the overlevel architecture.

01:43:26: so over the past year there's been a huge amount of debates about does the whole AI Act apply as well as the MDR or IVDOR and was decided last year.

01:43:35: this will probably just be the MDOR but that is not how it's prepared in the AI Act.

01:43:40: So there has been legislative changes and they're looking to, you know bring these high-risk devices regulated under other product legislation solely under that legislation.

01:43:51: But part of the way that will be implemented Is that the MDR regulators need to come up with guidance That will incorporate important elements from the AI act so things like explainability or transparency which are Not really apparent on a medical device world in a whole scale way.

01:44:08: So I think we're starting to get clarity on the overall legal approach, so it'll be MD or with AI act requirements within us.

01:44:16: there's going to be some challenges based onto timelines because will have to designate notified bodies do that kind of task and that hasn't been figured out yet.

01:44:26: once you got these big chunky things figured then i think its really important when good quality evidence is needed, but I don't think any regulator around the world has gotten a completely clear steer on that yet.

01:44:40: And to make that better for everyone... ...I think we have more work to do.

01:44:45: thinking about risk classification especially AI medical devices.

01:44:51: There's a proposal to change what's called Rule Eleven which are software classification rules and they've been challenging Especially here in Germany.

01:45:00: there different interpretations, is it possible to have a class one software medical device or does that always need to be class two A or above?

01:45:08: So we need the first better classify these.

01:45:12: And once we figure that bit out then figuring out the kinds of evidence generation based on a risk classification will be easier.

01:45:19: but when you look across US where things are uniformly Class One because they had the old directive type rules and in Europe what things could come out with quite different classifications all over the place.

01:45:34: So I think we have higher level challenges in the regulatory world to get the basic blocks in first and hopefully then that would lead us to a better approach to clarifying evidence for devices, yeah there's lot of implementation work

01:45:51: yet right?

01:45:54: And specifically to AI let say The radiology example that you brought, we have seen AIs outperform doctors.

01:46:06: It's very interesting that you mentioned the case in France where actually to ensure quality need three opinions if We go through trials and how to generate the evidence?

01:46:15: You would also say let's say two people maybe three as a third person has a tiebreaker.

01:46:20: humans rate an image And this we use is a gold standard for AI compared because then single investigator is unlikely to be more accurate than an AI looking at the image.

01:46:32: And for that reason, you can't even establish ground truth anymore.

01:46:35: so in this space I think we need to understand that even fundamentally For the care of our patients it's advantageous To use these tools.

01:46:47: yeah So we need make sure yes We do provide ample evidence but also ease of integration To allow them to get the market.

01:46:56: because for day-to-day routine, especially given strain on healthcare budgets.

01:47:03: Extremely

01:47:03: high workload and therefore ease of cases going undetected or care not being provided.

01:47:11: well it's good if we have these tools.

01:47:14: so yeah I think you pointed out very adequately that yes We need clear guidance And it's urgent.

01:47:27: Yeah, and you know You touch on an interesting point in that Like we're saying earlier these technologies will change the way We interact and think about things in some ways.

01:47:39: one of The most challenging things in my experience is working On changes to health service delivery.

01:47:46: That can be a real challenge and It's quite different across different countries or member states.

01:47:51: but When you're looking to implement a new digital health technology, getting your clinical validation is one thing.

01:47:57: But understanding well could this actually work?

01:48:00: You know let's say you have a decentralized care-at-home model.

01:48:04: uh we want to shift patients home and monitor them there and we've got this great evidence that this can be done in a safe way.

01:48:10: but it might actually be the change an organizational nature of that service within the hospital.

01:48:18: So I think the

01:48:18: risky.

01:48:19: Exactly, yeah exactly.

01:48:22: so that's just another challenge you get when trying to implement these digital health technologies and as you were saying Andres budgets are constrained and hospitals might buy some AI technology for them.

01:48:40: but if we talk about radiologists A lot of these AI algorithms are distinct, and they have their own workflow.

01:48:48: So you've got to go into your packs or imaging system.

01:48:51: then in a work flow if there's five different workflows it becomes a real pain.

01:48:56: so There is lots possible refinements that can be done.

01:49:01: just make the user experience better as number of tools could increase.

01:49:06: Yes!

01:49:07: Lots practical things will help for better integration.

01:49:12: It is certainly changing medicine.

01:49:17: You know, it's changing the way we think about decisions and especially for patients The most common person they consult or person resource to consult his doctor Google Or maybe Dr Chatbot before an actual Doctor.

01:49:32: So this is a hugely different Way in that patients think about their diagnosis sort of condition symptoms.

01:49:41: So yeah, it's just a huge change for all of us as well.

01:49:47: It is very dynamic and we have just ventured into this space.

01:49:53: Just

01:49:53: getting started?

01:49:54: Yeah so let's move to the future in what will be changed because I think clear line through everything that we touched upon And I think this is especially true for Europe.

01:50:14: We've just sort of touched upon it in other countries a bit better, the level of clarity over requirements and how to get there... ...and different stakeholders that we have involved.

01:50:28: You played on various different tables.

01:50:31: you started out as a healthcare professional applying the outcome of what our innovation system applies.

01:50:41: You moved over to the regulator side, you have discussed on a European level your professor and therefore teach let's say in new group of people to venture into this space and look out for how we can change the future of our own health care here in Europe.

01:50:58: so you have very broad perspective.

01:51:03: And now we are at a time where you mentioned it initially, also.

01:51:07: We're talking about MDR two point oh and we sort of like want to discuss some ways off changing our legislation.

01:51:15: What's the core problem that should be addressed in the near future?

01:51:19: How is it being addressed?

01:51:21: so what's the let's say was your take on the status of this conversation right balance to be struck between innovation, risk for patient adequate level of evidence and accessibility off health care products.

01:51:43: Yeah.

01:51:47: Predicting the future is never easy we cannot be sure but we have some directions on where things are going with revision to legislation.

01:51:56: what these changes to the medical device in IVD rules proposed as support is competitiveness an innovation.

01:52:05: So this is very clearly the mandate of what's called a simplification package for these two regulations.

01:52:12: That runs into some tension from my clinical perspective when we have, you know.

01:52:19: to give one very simple example if you want to market in your heart valve today MDR says you need A clinical investigation.

01:52:27: doesn't say what kind?

01:52:28: You need.

01:52:30: with these revisions that won't need as long.

01:52:32: there is something similar available to which you can claim equivalence too.

01:52:37: That has been challenging with MDR, because we need access the primary data?

01:52:42: Yeah and regulators ended up describing different levels of access.

01:52:46: so in article sixty one five it's a part that says you needed contract to claim equivalent when your class three... Which was completely

01:52:54: unrealistic.

01:52:55: You know what happened except its own.

01:52:57: Exactly!

01:52:58: So it was an effective prohibition by the legislators for that equivalence route in all but very limited circumstances.

01:53:06: And back in twenty twelve, The European Commission said that you know equivalents as a mean to avoid pre-market clinical studies or high risk device should not be allowed and still are quite clear after the crisis of metal on metal hip implants and breast implant crisis uh... That should be the approach.

01:53:25: But it's now flipped one eighty demonstrative of a system where we're still reasonably well gripped by policy-based evidence rather than evidence based policy.

01:53:40: When a heart valve needs clinical study today and the heart valve won't need one after these revisions, that's an enormous change.

01:53:45: And in a lot of advocacy work I've been doing over last months it has been very simple just to explain to policymakers all under the previous system.

01:54:00: Things like heart valves, hip implants, transvaginal mesh lots and lots of things that were approved on a basis for equivalence then had major public health concerns where it may not have just been one device but whole family of devices was unsafe to be marketed.

01:54:17: I think this is one particular change as proposed.

01:54:22: when i would be concerned about Well, maybe leaving that to one side and thinking of what the future is looking like.

01:54:32: We get some sense of kind-of way.

01:54:36: it's proposed that this system will run as a result of these revisions.

01:54:40: so if we went back year ago all of the stakeholder groups were advocating for a captain of ship or clear revision of governance structure because you have medical device coordination group notified body coordination competent authority medical device network, thirteen working groups under MDCG twenty seven authorities.

01:55:01: Fifty five notified bodies.

01:55:04: this needs greater strategic direction and you know my hope for the system would be that in five or ten or fifteen years at EU Medical Device Regulators could write a strategy on publish one.

01:55:18: we have strategies from medicines and regulatory system there.

01:55:22: We were not able to form a strategy for medical devices because There's no one to form on.

01:55:28: We need to figure out these kind of things because the challenges that we saw with MDR were significant for this sector, so unless you have a strategy it is very hard for all us get behind and point at North Star.

01:55:44: What is changing?

01:55:45: I guess there will be the same general governance system.

01:55:52: They will have some greater scientific and technical support from the European Medicines Agency, so their remit is being expanded a little bit.

01:56:02: And they will have new opportunities to do.

01:56:05: things like help national authorities with some pre-market studies... ...help answer safety questions or questions raised by the regulators to the EMA.. ..and some very new ones like helping with classification disputes.

01:56:19: To get a formal classification opinion goes through something called the Helsinki procedure.

01:56:24: that has over thirty procedural steps and can take up to two years to get an answer.

01:56:30: Now, expert panel members are clinicians so they're not people well versed in annexate of the MDR or its interpretation.

01:56:37: So it's kind interesting their bought-in as a arbitrator when member states disagree.

01:56:43: but I think general is good direction.

01:56:46: we looking at more organized scientific support to this regulatory system So it'll all come down to how its implemented.

01:56:53: and do we get these channels communicating with each other, I'm working effectively.

01:56:58: That's entirely contingent on implementation rather than exactly what wording you have in the law.

01:57:05: but i think that could potentially set us on a more scientificing clear path.

01:57:10: some scenarios will have to see.

01:57:13: In general for notified bodies is going be quite difficult.

01:57:17: the proposed changes to get rid of five-year reassessments, and have lots of things considered well established technologies where you can do sampling rather than individual file assessments.

01:57:28: You know that the cumulative nature of the change as proposed for notified bodies will knock out a lot of their current income so they're likely to shrink in the coming years.

01:57:38: but what would be effective?

01:57:40: Will this reduce any longer wait times or you know, internal changes causing a more chaotic conformity assessment?

01:57:50: we'll have to see.

01:57:51: It's very difficult to know but it will be very interesting to watch because notified bodies were much bigger in number under the directives I think the highest number of notified bodies and this system was something like ninety three.

01:58:05: so there is huge amount now at fifty five.

01:58:08: But over the last years with MDR notify body has been growing every year.

01:58:13: as first-year they're getting smaller.

01:58:16: So when you think of some of the public discussion, there were lots staff removed from US FDA under their various policy approaches.

01:58:28: It was interesting to see how much industry wanted keep in-house knowledge at FDA because industry regardless of farmer or med tech they value knowing that if had an assessment previously and know what assessor thinks Better for you as a company predictability exactly.

01:58:48: Exactly So we may have Changes of that nature where your previous assessor might be gone or it might be in your team more right?

01:58:56: Or different kind of approach.

01:58:59: and I think, you know one at.

01:59:00: the interesting Hypothesis that will have to see is To what extent do all these changes actually support competitiveness an innovation?

01:59:10: I see companies at the very nascent stage, not even yet a company.

01:59:14: It's maybe a concept prototype getting into preclinical and when i think of all of the changes in mdr The majority of them tend to only come out much later life cycle stages.

01:59:26: right.

01:59:26: so are we going?

01:59:27: To actually turn People thinking who might have been considering market access in the u.s.. First the thinking of e. you first That?

01:59:36: that's an open question for me so far.

01:59:39: If you're an orphan or a breakthrough type device, You might go for the kinds of supports here.

01:59:45: But from my experience since the revision was published in December until now From talking to these early stage companies they still value The clarity of US system compared with E-One.

01:59:57: So our essential battle In Europe is that we set up These rules originally in the nineteen nineties.

02:00:04: Rather abstract terms.

02:00:06: That's the wording from Norbert Anselman, who was the principal officer in The European Commission responsible for first medical device directive.

02:00:15: So... ...the original design.

02:00:17: to write rules and abstract terms is very difficult.

02:00:21: then you know thirty years later bring clarity into a system that has developed thirty years of habits working within abstract terms justifications rationales those kinds of ways of documenting things.

02:00:34: So, I think that's going to be the challenge.

02:00:36: when in the twenty-first century a new developer needs clarity on evidence requirements and when Europe has been designed from its origins as vague on requirements how do you try and resolve it?

02:00:48: And where are most valuable places for trying to solve this?

02:00:52: If we could figure out strategy for that would really support innovation into system given the nature of European system, authorities notified bodies but also funders and people at hospital sites who run studies.

02:01:09: All support organisations that help with this will have to figure out.

02:01:16: you won't be able change everything.

02:01:17: we'll try find a few meaningful areas where clarity is most impactful trying deliver it there.

02:01:24: But I think our essential problem would still be faced after these revisions The freedom of abstract terms, but the tyranny... ...of not having confidence will work perfectly.

02:01:40: Doesn't sound like an easy challenge to solve?

02:01:43: Would you from your time on different parties...?

02:01:50: What would if you had the power say it'd be best policy decision or general structure for an act?

02:02:00: entirely abolishing the notified body concept, unifying everything across Europe.

02:02:05: Is it fine-tuning?

02:02:07: Providing more guidance?

02:02:09: where would you say is the best way?

02:02:13: or what do you see the strongest leverage to use to change this system?

02:02:20: Yeah So if I had a choice for acknowledging real politic of it.

02:02:32: You know, I think there's been a few debates over the years about abolishing notified bodies and back around twenty twelve i think was one of the strongest attempts when there were actually you know resolutions at European Parliament that we should have an agency for medical devices.

02:02:51: It probably wouldn't be feasible now given just this size notified bodies like six thousand staff in them a thousand clinical assessors.

02:02:59: No one is going to pay to replace that in a public setting somewhere.

02:03:04: So it has just grown to such an extent, that abolishing notified bodies is not politically feasible.

02:03:12: But if we were make discrete changes That would be most impactful.

02:03:18: I would recommend We take approach similar To the United States for new high risk products Like pre-market authorization rigorous clinical evidence requirements, but for a small number of therapeutic implants that are new and coming into our system.

02:03:34: You know the number of devices going through that pathway in the US is like thirty five per year.

02:03:39: we could distill out those thirty-five devices where getting it wrong would have major public health implications.

02:03:45: um and treat them rigorously.

02:03:47: That would leave you to space to treat the medical equipment on the messy middle much more liberally.

02:03:53: You could figure out by working well with standards communities how to try and make you know, standard work for those class two devices.

02:04:00: Exactly like Theodore Cooper recommended in nineteen sixty nine um And four-class one.

02:04:04: it would be much easier.

02:04:05: Um.

02:04:06: so I think that's one of the key things That we could do is treat a small number Of high risk therapeutic implants rigorously and that Would allow you To justifiably develop more focused Things from middle risk products where preclinical can have good evidence In some cases or You might need to complement it with clinical evidence, but not a full enormous study and others.

02:04:27: But beyond that you know.

02:04:29: the other change I would recommend is that we recognize what has changed in the last twenty years which Is That The Dynamic of Introduce It In Europe Do A Very Small Study CE Mark Generate Revenue And Then Go For The US?

02:04:44: That Has Changed!

02:04:46: If Regulators Want To Keep Up With Real World Strategies applied currently in this day and age, we need to recognize that early clinical clarity is one of the most important things you can do.

02:04:58: or early scientific clarity for some kinds of technologies like... Is an in silico model acceptable?

02:05:04: For discontextive use.

02:05:05: Or something like that.

02:05:07: if We could Recognize That And Apply That From A Single Source Giving Clear Advice And It Wouldn't Need a Huge Compliment Of Staff But it Would Need Some I think that would be enormously beneficial because then you will be able to say it's the new developers I meet all of time.

02:05:23: Instead, just going into FDA for a pre-sub why don't we talk about these people?

02:05:27: For a pre sub get clarity It'll be written down and recorded somewhere You can proceed.

02:05:32: That is probably one or two other most important things that could do To support product introduction in safe way.

02:05:38: And with this We need much better focus on methods Because they have been underappreciated by CE marking nature.

02:05:48: It's all about conformity, and we don't talk yet about that or methods.

02:05:52: And We need to.

02:05:54: so I guess That's one of the other undervalued things that we need To put some focus on and as you know huge amounts Of research and recommendations already available.

02:06:04: But we have to find out The governance framework in a change In mentality?

02:06:08: That would allow us to think About that rather than just reinterpreting the law All the time more These kind of approaches we currently Have.

02:06:18: Do you have confidence that we can, I don't even want to stay relevant and stay competitive but regain competitiveness.

02:06:29: And regain relevance on the global stage from a medical device let's say development an invention perspective?

02:06:37: We are confident as if we can do that given the framework of policies that we operate in and the speed at which they develop into interaction.

02:06:45: yeah i think quite fertile early innovation system in Europe.

02:06:54: Lots of new product ideas, there's a high attrition rate obviously but lots of new things happening in Europe.

02:07:01: I think where does great opportunities to improve things?

02:07:07: We mentioned early regulatory clarity.

02:07:08: But you want to do multi-site or multistake clinical investigation.

02:07:15: that can be an enormous body of work You know, to get a sign off from five member states.

02:07:20: From their authority and the ethics And data protection that could be different in many places.

02:07:24: and then contracts To make Europe more attractive for translational development.

02:07:29: We have to start to make some of these processes more combined and clearer.

02:07:35: That doesn't mean you know mixing your authority and ethics into one single committee In that Lancer for all of Europe.

02:07:41: I don't think people are ready to move in that kind Of away yet but there's lots of simple things that we can do that would help to save months on preparation work.

02:07:52: So some work I did over the last year in Galway was something we called a hypercare initiative and We took three developers, you know large multinational US small company an Irish Small Company And they had an early feasibility randomized trial and an observational study.

02:08:11: bring them through the authority and ethics on your site approvals in contract.

02:08:15: So it was up to that stage of enrollment, um... To see without changing anything into law.

02:08:20: but if we had more openness some more discussions between different people uh.. We could find process efficiencies And we found some you know quite simple ones like Your data protection impact assessment.

02:08:34: You know a couple years ago In Ireland If you have five hospitals Each one would have their own data protection impact assessment, and sometimes each hospital will go to its own local lawyer for an opinion on it.

02:08:45: And then you lose months and months a month.

02:08:48: so we made the national policy decision that just the one Data Protection Impact Assessment from the sponsor is enough.

02:08:54: It doesn't have to be repeated or things like starting your contract preparation with The Hospital site whilst you're undergoing Your Authority and Ethics Assessment can save You three Months which when you're thinking for startups and our cash burn, they can be important things.

02:09:10: But I think there's a lot of quite boring things that could be made much more efficient.

02:09:15: that would improve the attractiveness of Europe in general.

02:09:18: You know, i'm not so involved in the funding world but I think this...you know if your talking competitiveness thats essentially what you are talking about to attract funding to do these kind.

02:09:34: It wasn't necessarily any individual problem with MDR, but the sentiment that grew around it.

02:09:41: That because of MDR don't go near Europe.

02:09:44: I think that puts Judy on everyone involved in this system to find a better solutions But also explain what's happening in Europe?

02:09:52: Better Because impressions from the United States can be very general but also quite poor.

02:09:58: and You know if you think of there hasn't been huge changes for pre-market studies with the medical device regulation.

02:10:04: But I've heard lots of founders say oh MDR.

02:10:07: We're not going near Europe for a study Not much change for pre market studies, you can still come here.

02:10:12: Um, you know mdr doesn't you know cause a huge change for these kind of things?

02:10:17: So i think there's uh a lot of um Explanation that we need to do internationally as well as working on some of the boring challenges that slow up research.

02:10:31: And beyond that, I think you know... ...I did see a shift over the last years where the CE mark used to mean access or pretty much access to maybe one hundred and ten countries worldwide.

02:10:45: but when i looked at it over not formally just anecdotally there's a number of countries which went from accepting the CE Mark to CE and FDA.

02:10:56: So these competitive influences of our systems relevance are getting quietly more important.

02:11:05: I think that we have an important public relations job to do and hopefully, that will help with some of the challenges we saw at MDR but above all of that international harmonization is talked about when you go to all of this International Medical Device Regulatory meetings.

02:11:24: And there are some things happening now.

02:11:26: You know, if you look at the United Kingdom for example they're going to start accepting FDA approved or CE mark devices which would be very interesting test case to see how that works in that country.

02:11:37: and rather was a similar movement In Switzerland over the last years where they were thinking about doing that?

02:11:42: Um...and then lots of countries outside of Europe Would have kind of system anyway.

02:11:48: There is no rational reason as far heart valve into US or Heart Valve into EU should.

02:11:53: quite different clinical evidence, quite different regulatory assessments.

02:11:57: Quite different evidence publicly disseminated.

02:12:01: but I don't see very real movements to try and bring those kind of things closer together.

02:12:09: You know the kinds that you can do outside.

02:12:11: a formal political intervention is A bit like our hypercares just getting people talking To one another.

02:12:17: so in an example that regulators did like, twenty years was something called harmonization by doing.

02:12:24: FDA people would have worked with PMDA people in Japan for things like heart valves or pediatrics and things like that.

02:12:32: so there's lots of thing under the hood a big political change where you get shared experience.

02:12:38: it is interesting to see.

02:12:39: just this week MHA and USFTA will be swapping staff putting people into each other offices for awhile.

02:12:47: these help share experiences.

02:12:49: how regulators realize Oh, you're doing things very differently to how we seem to do things.

02:12:55: So we need probably better political attention on the value of international harmonization.

02:13:01: We shouldn't be competing on ease-to-market or the quickest or cheapest whatever way to get a CE mark versus at five ten K and... We should be competing under value.

02:13:14: what you can do for translational development between two ecosystems.

02:13:18: that would be nice way to compete.

02:13:20: When it's a regulatory competition, It becomes quite political and quite complicated.

02:13:24: But doesn't tend to serve anyone very well.

02:13:26: companies have double work.

02:13:28: You know patients have potentially different levels of protection.

02:13:32: So it doesn't work for any one that is currently maintained.

02:13:35: but I think if we can do some little things under the hood like sharing experience And sharing cases Maybe it's something like orphan devices or could be a nice testbed to try out some of those things.

02:13:47: So maybe after we finish with our legislative focus this year, there can be other things in the future that will be very nice.

02:13:53: but I think its longer arc for development.

02:13:57: so yeah...I guess we talked about couple of things here you know?

02:14:01: To me i think change for equivalence is one of the most publicly...in terms of public health challenges.

02:14:10: One thing I didn't mention is that, you know...I've done a lot of work with clinical and diagnostic groups through Biomed Alliance over the last year.

02:14:18: And especially for the IVD regulation in the diagnostic community.

02:14:23: being able to prepare IVDS in-house it's hugely important to them.

02:14:28: So there was some promising changes with revision to the IVG regulation To allow in house IVDS to be transferred between health institutions when there's no CE mark device available and public health need.

02:14:43: This is exactly what happened with COVID, you know?

02:14:45: There were no COVID tests in the early months of COVID.

02:14:48: they're all-in house devices but if you can only share a written protocol for how to prepare it at your hospital depending on the IVD sometimes You could do much faster If you could share reagents or other materials rather than just a written Protocol.

02:15:02: So I think this would be really helpful For rare diseases where It may almost half of the available devices in a hospital lab, but they're used less than two percent at a time.

02:15:14: Again it's rare disease challenge.

02:15:15: or for things like companion diagnostics personalized medicine you know It really has a lot of potential to help there because I think The pipeline for introduction of IVDs is very different to medical devices.

02:15:26: They can often start in a Hospital Lab and then later translate to CE mark devices.

02:15:32: So does that?

02:15:32: There's a lot we could do support innovation Or new IVD development you know, in that kind of way.

02:15:40: So it's a proposal currently um in the revision but we'll have to see if it works its way through the legislative process.

02:15:47: yeah

02:15:48: right so that I think it's very nice and sort of ends us here on a very nice note.

02:15:54: uh It seems that from this legislative change And also what you mentioned In terms of exchange of staff between FDA and MHRA or cross functional development that these are ways to really showcase there can be movement in the system and move forward.

02:16:11: Is there anything else you would want to shout out, let's say regulatory movements?

02:16:16: That give you hope things change or do they also for better?

02:16:21: despite all of this we discussed

02:16:25: today?

02:16:26: There is lots of challenges but I think certainly it gives me hope!

02:16:34: you know, small groups of people coming together can change the world.

02:16:39: In fact that's the only thing that tends to change the word and when I think about things like The Orphan Device Challenge.

02:16:46: this was small groups with people come in together realising there is a problem doing their best try make it work.

02:16:53: since then we've had some EU funded research projects where for first time How do we better establish researchers or networks?

02:17:04: And I was in one project called Decode Developing Child and Orphan Devices.

02:17:10: We got to have mentoring sessions with five product developers, seeing how getting just a small number of people who know about regulation or clinical methods or outcomes together working with these new product developers could really help them get clear on... How do we get this product?

02:17:30: through a study or introduced into the market?

02:17:32: And they were often very simple things, you know something like up.

02:17:36: A simple walker for a child with a neurodegenerative condition in it could be engineers and an academic setting who've developed us on no idea of regulation at all.

02:17:46: You can really help strengthen their translational development by telling them some of the important things that need to think about.

02:17:52: So there's lots of hope in the system as well as lots of challenges.

02:17:55: But I think its bike coming together You know, getting people working together even in small tiny little groups and trying to grow it.

02:18:03: I think that's how you try get things done

02:18:05: awesome.

02:18:06: Thank you Tom for sharing so much insights In your perspective.

02:18:10: Your perspectives Flora on on clinical research?

02:18:14: Uh i love you.

02:18:15: answer the other.

02:18:17: i i love That you are here too Despite The challenges that you've seen from within this system.

02:18:23: push For change uh to push for more science in the regulations, to sort of like flip this system a bit on its head and work from their need-for evidence backwards.

02:18:36: And how we can provide more guidance towards that.

02:18:38: really appreciated thank you for coming by.

02:18:42: it was really lovely catching up!

02:18:43: I'm sure we could talk for another hours and will... ...and i'm sure there'll be an opportunity upcoming news, changes in regulations.

02:18:58: Should MDR-II Pano come over we should certainly have you here and dig really into it again.

02:19:03: Absolutely well thank You Andres!

02:19:05: It was a really nice

02:19:09: talk.