Real-Life PMCF: Practical Lessons, Notified Body Insights, and 2026 Priorities

Good morning and welcome. Or good afternoon, depending on where you are in the world. If you want to go ahead and drop into the chat where you're dialing in from, that'd be awesome. It's good to have everybody here today. Welcome to our speakers, Daniela, Mateo, Diane. I'm so glad to have you here, for today's webinar on, PMCF, and, what's changed, what's new, and, strategies and practical, tactics for strategies for 2026. So, we're very lucky to have today's speakers. I'm gonna, begin with a few housekeeping announcements before we, dive into everything, let's see. This webinar will be recorded, and the recording will be made available

with the slides, everyone who's registered should receive an email 24 hours following the event. If you don't receive it, feel free to reach out to me, by whatever channel you choose, to… let me know that you didn't get it. Also, in a few days, it will be available on our YouTube channel. There is a MedTech Leading Voice YouTube channel with all of our webinars on it. I recommend that you subscribe to that. You can find it on YouTube, but simply by searching Medjek Leading Voice. Let's see… What else? If you would like a certificate of attendance from today's call, please put your name and your name as you would like it to appear, along with your email address, into the chat. Use the chat to communicate with me, use the Q&A function to communicate with the speakers. I cannot copy your

questions from the chat channel into the Q&A channel. So, again, please use the Q&A channel to ask questions. Of today's speakers. I have, I think that's it for the announcements. A short statement that I would like to read This presentation is intended for educational purposes only and does not replace legal text of the legislation, standards, or guidance documents. These are personal opinions of the speakers and interpretations as subject matter experts, appreciating that the regulatory landscape is ever-changing, as is our understanding of it. And then I want to, frame this conversation with the recent publication of the MDC… of the MDCG

2025-10, where proactive is defined, but more importantly, gives us important insights into the current thinking and direction of market data. Proactive emphasizes the manufacturer's obligation to actively seek out available information and not merely wait for it to arrive through such as complaints. Manufacturers should play an active role in role during the post-market phase, systematically and actively gathering information from post-market experiences… experience with their devices. This involves deliberately planning the gathering or generating information from a variety of sources. These may include, but are not limited to.

Customer surveys, gathering of clinical experience, user feedback, screening of scientific literature, and other clinical data sources. Meta-analysis of published clinical data evalu… clinical data, evaluation of suitable registers, and post-market studies. So, with that in mind, I would like to introduce today's speakers, Daniela Carrere and Karar and Matteo Mosso are, with the CRO, Donovan Life Science, based in Rome, Italy, with a global presence, and Diane Lager. from… DNV, I'd like to welcome the three of you. I'd like to note that, Diane recently went under… underwent a small operation, which has left her voice capabilities somewhat damaged. So, if her voice gives out during the call, she's prepared remarks for me to read. So, if I am speaking as Diane, you'll know why.

Why? Oof. What am I leaving out, gang? Anything, or are we ready to rock and roll? Let's do this. Thank you so much.

Okay. I'll start… Sorry. Ugh. As usual, I need to share…

While Dalian is pulling that up, thank you, Sean, for being my pitch hitter, if I need you.

Sorry, I cannot find my screen. It's… it's this last min… oh, there you go, there you go, sorry. Last minute. issues that I'm having.

We were having too much fun chatting before the call, the three of us were progressing, and .

That's true.

And someone got distracted, that's okay. Thank you so much.

Okay. We're nearly there. Okay… So… and, Okay, maybe I can make it.

You got it.

Share… No. That's a live session, you can record that. Oh. Okay. It seems that we're done. Okay, welcome everybody, sorry for the last minute. issues. So, real life. We wanted to keep this terminology, so we're trying to be, to stay with, real life examples that we've been having, since MDR came into effect. It's nearly 5 years ago, so we have something to share.

So, the agenda, basically, the webinar is in two phases. On the first part, I will talk about the regulatory references, and what guidance documents tell us, and then present a few case studies with the feedback from the notified bodies, and I will ask a couple of questions to Diane for her… Precious feedback. And then, in the second part, Matteo Master is going to… go more on the practical side of things, and talk about the methodology, the operations, and how the MCF needs are identified, and above all, how data can be found, and pros and cons of each, approach.

So… That's a kind of excuse slide, because we wanted to do something less traditional, but and that's a praise for the regulators. How can we do without regulations? And why do we start again? Because… There's a lot here. I have to say a lot more on this part, rather than compared to when you have to plan, for example, preclinical studies for C marking. You're a bit lost, to some extent. It's less easy. In this case, there's a lot. to share, and it's what notified bodies will, rely on. So I think that these two reasons combined explain why

We opted for starting in a kind of traditional way. So, there's a list. I'm not going through all the details, these are the key documents. It's a non-exhaustive list, you may find others. So the MDR Annex 14, Part B, it's where everything starts. what is the PMCF scope, and the content of the plan, and then the description of the general and specific methods, and how PMCF fits into the Continuous assessment of performance and safety of each device in the post-market phase. I want to say something on the second, document listed here. It's a guidance on how to write a PMCF plan template.

Which contains, a lot more details on, how the plan should be, completed. It's a lot of details. It's good reading quite early, I think, because it poses you in the mindset of what are the questions that at some point in the development of the device, if you reach the market, you'll need to answer. So, which are the methods that you're gonna use? Why do you think they're appropriate? where you're gonna find the data, how much data you're gonna find, how many patients, how many subjects, which methodology, etc. So, a lot of details, very useful.

And then, I put here a list of other guidance documents from the clinical evaluation to the, what is called PMCF studies. It's a document from 2021. Of the IMDRF. Then there's quite an old document, MedDev, from 2012. It's post-market follow-up studies, but that's, a document that I found quoted by a notified body in a presentation that they held. A few months ago. And that also contains useful input. And for manufacturers of medical device, of software, medical devices, I'm sure you cannot miss the last… the last document, which is… which provides some input tailored to

The study to the product, to this product category. What is that this guidance document tell us, and why I think it's important? Well, the scope, and the scope should always be, kept in mind. PMCF plan. Now. You face the need to submit a complete plan, and then I have a question for Diane as soon as this slide is up. as soon as I finish, talking on this. PMCF plan should be submitted at the time of the CE marking, but it's quite likely, at least this is our recent experience, that you won't receive feedback from notified bodies, and this, sometimes, it's going to take months.

Even if it's negative, and the data and the plan is considered insufficient. This tells us that you should do everything you can to do it right from the beginning. So, this exercise Should be done quite right from the start, otherwise you'll be waiting for feedback that may come who knows when. So, the guidance documents address how, and the reasoning, and the logic process that would bring to the identification of what is the need, what you need. And that's the same things that notified bodies we look at.

I've put here some distinction between the PMCF general methods and the specific methods. Well, the general methods are expected as kind of default. type of activity that takes place for all medical devices. So, it's complaints, vigilance, feedback, spontaneous feedback from users. Review of the scientific literature on the device, or similar, etc. there is nothing… there is no debate, I think, on this part. The main debate exists on the second part, on the specific methods, because they're… Specific for the device, and…

they are the most expensive compared to the other methodology, and they traditionally include these categories. Surveys, real-world evidence data, registries, PMCF studies. This is taken from one of the guidance documents that I've mentioned. This will be discussed in the second part of the presentation. So, here comes a question for… Diane… If you can share what your experience is in terms of how are PMCF plans usually set up. Are they detailed enough? Do they allow an adequate assessment from a notified body perspective?

Yeah, thank you for the question. I'm going to say the answer is really a mixed bag. Excuse me. I found higher risk devices. Tend to submit more detailed, specific. PMCF activities. They follow MDCG 2027, as you mentioned. You know, I will tell you, every single clinical assessor I know at Notified Bodies use that… guidance… guidance. to assess PMCF plans. I don't know anyone who doesn't use it, maybe just by memory now. I know that in our greater team, within decision makers, panels, whatever you want to call them at your notified body, they are specifically looking for, the components of that guidance document as well. You know, Section C really

tells you, you know, what to do. Give a rationale for the appropriateness. Choose general and specific, You know, the part that's often missed, funnily enough, is the timelines. I find that sentence that says it should be at least quarterly. To annual, for some reason. It just gets missed. And I don't know… I don't know why, but that's one of the most common issues, I think that there might be a misconception that that is risk-based, you know, risk-based, and that if you have a Class 2A, you don't need to do it quarterly or yearly, but I'll tell you that sentence is more about the requirement for continuous

PMCF activities. It's not meant to be. risk-based. It is really focused on… on the continuous. So, I'm going to say lower risk classes, especially legacy devices, have the most challenges with specific methods. I think because, you know, they can't rely on literature, and it's challenging anyways to come up with specific plans to get the, you know, good data. So that's reflected often in their plans. You know, you alluded to the… the general things, and I think there's still confusion that You know, PMCF is part of PMS plan, so anything that's already listed in Annex 3, you know, a PMS plan should have. The things that you mentioned, spontaneous, you know, complaints, feedback, literature, those are general.

You know, you have to go beyond that, and create specific activities, like, like the things that you, listed already. So those are the common gaps. Alright. Timeliness and specific for lower class.

Thank you, thank you very much. Very useful. All right. Let's go on. So… key considerations that's taken from the… it's my list, and how I see that from the list of guidance documents that I've mentioned. So. What is the… what are the considerations? I don't think that there is an easy algorithm, or maybe AI now would do that, and you put all these considerations here, and it gives you the perfect PM… PMCF. I'm not sure we are… already there. But the device risk, it's the risk classification that takes into account the material design, invasiveness, length of interaction with the body, whether it's an implant or not.

etc. And that's, in a way, embedded in the device risk classification. But then there's a risk regarding the intended use. So, what is the anatomical location or the target population? Are there vulnerable subjects that may receive treatment with a device? Children, for example? How severe is the disease that is being treated or diagnosed. So… Clearly, devices which are new, that use new technologies as compared to devices, legacy devices, for example, that have been on the market for a while, they are kind of opposite.

type of, approach. In one case, you have a lot of data hidden somewhere, and Matteo is going to talk about that. In the other case, there are no data, so you will be forced to plant something prospectively. How many data you have behind you, how many data are part of your technical file, or you can trace from previous clinical investigations? And if, in case that you did, in fact, clinical investigations, either But legacy devices, they may be PMCF according to the directive, or classical pre-market clinical investigations. How can… can you generalize these results to the entire population, to the cases that you're going to meet once the device goes on the market?

Are there unanswered questions? As an example, long-term safety or performance, residual risks identified, are there uncertainties? This word, uncertainties, it's… I found it on the, on one guidance, and it's repeated in more than one, actually, and I found it very crucial. It's very intuitive. So, what is the uncertainty that we need, somehow, to clarify? by collecting data in the post-market phase. And then, clearly, what type of device are we talking about? Are we talking of a device that we're gonna sell

100 this year? Or… 30,000 this year. It's… it all changes, because the availability of data As time progresses, it's another element to… consider when planning for specific PMCF methods. So… It's… a lot is around planning, and which data we're from, and that comes to the… the hot question that is the cost. Something I wish to mention here specifically is

rework. I mean, cost for rework, which usually you think of cost as the first thing that I… that I do. It's cost for my next activity, but saving at the beginning may not always be the best strategy. I mean, there are risks involved. So the cost of rework, sometimes it's underestimated. Okay. Here we are. Again, any, any aspect that you think are, that, based on your experience, are underestimated? Diane?

Yeah.

Based on your experience.

So, I'm gonna say… I'm gonna say, generally, I should be able to read your clinical evaluation and your risk management. And clearly be able to draw a line over to what your PMCF activities are, the specific ones. The general ones, as you said, everybody has to do them. It should be easy for me to… to see why you, planned for that specific question and methodology. So, for example. You alluded to, you know, maybe specific severity of diseases, or conditions, or ages, and that you didn't have… there were gaps in your data. So you may be… or variants, you didn't, you know, you have a larger size that's not used very often, you don't sell very many of them. So I want to see that you identify that gap.

You probably gave a good justification in your clinical assessment of why it's still safe and performs as expected, but now you need to go and plan and show me, okay, so the example of variants. You know that you sell them to, you know, this space. So you're going to have a targeted, activity to go to that… those users, and ask them about that larger variant, and make sure that it's being, it performs as intended, and there isn't any other safety, issues with maybe, like, a larger guide catheter, for example. Or that the age, that there isn't anything specific with comorbidities that are affecting the benefit risk. So I… I really wanted to see, an easy transition from your clinical assessment to your risk over to your PMCF, even if it's

We know all the known risks, and we're going to just… You know. continuously assess this. But it should be easy, I shouldn't have to be shaking my head, reading it again, and trying to make that trace.

Thank you very much. Thank you. Interesting. Yeah, to make it easy for the reviewer, that's something that's another thing which I think sometimes it's underestimated. Okay, so it was just very briefly the proposed MDR revision that came out in December. The only change that in case that the proposal is, approved as such, that's going to affect PMCF is the elimination of the PMCF evaluation report. Therefore, the results should be fed into the CR and other technical documents directly. Certainly a step that's simplified for manufacturers, but not

it's not gonna change the overall, efforts required by this, by, the regulation. There are other, I think, positive changes for clinical data collection, but they are referred primarily to the pre-market phase. There's some tweaks about the well-established technologies, equivalents, etc, but they… based on our assessment, they will not impact PMCF requirements. I just wanted to briefly touch on this. And then we go on the case studies. A few case studies. So, Case studies of late clinical gap. So we had a Class 2A device, for transient neuromodulation in urology, and the device had been MDRC marked.

Based on equivalence. The manufacturer had a kind of naive approach, and they did not plan any specific method, and they did provide a rationale, because in case that you're not planning specific methods, you have to Justify that, and they did provide it. And the reason was that the equivalent device had been on the market for several years. Feedback by Notified Body was that There was one program, in the device, and that affected pulse and frequency of the, Stimulation in the new device.

which was not present in the equivalent device. So, there was a very… quite late, it was too late, identification by the notified body. of a clinical gap. So, what was the outcome? The PMCF plan had to be Entirely changed, how there was a prospective, clinical study plan on the, to, collect data on the safety and, performance about the new program. With all consequences that are listed here, more cost, late study start. etc. Need to add additional sites while the study was already running, because enrollment was very slow, etc.

Not a perfect story. Case Study 2. Implantable Class III cardiovascular devices. There were two models, one with material A, another with material B, In this case, there were indeed specific PMCF methods planned, and they included The purchase, let's say, retrospective data which were purchased by… commercial database. I mean, there are companies that offer this service. And this data were coming from more than 100 US hospitals, from electronic health records, insurance claims, etc.

So, there was a lot of retrospective data, but there were… Important missing data, and the data were considered insufficient. data and individual implants types were missing. Not all of them, but many of them were missing, so it was impossible to come to identify data that referred to one material or the other. and other data and outcome parameters were missing. So, the outcome was that, practically, There was a new study. That included both retrospective data analysis and prospective.

data. It was sponsored by the manufacturer, they had to plan this. And select clinical sites carefully to ensure that these sites did collect the data that they needed. In terms of cost for eWork, that's quite clear to everybody, and the commercial database was… was not useless, but certainly didn't serve the purpose for which it was. Purchased. Three. Class 2A device in musculoskeletal indication, 5 main treatment programs, treatment modalities. Again, another… device that was CMARC based on equivalence. In this case, the manufacturers did

Plan and execute, partly execute. specific PMCF methods, they've included data. So, at the time that the notified body saw the plan, they saw the plan and even saw the data, because in the meantime, they had Performed two retrospective surveys with data from more than 400 subjects. And they plan a prospective study. The feedback was… The survey was nice. As long as it contained data on all treatment modalities, which was not the case. So, there were, missing data on 3 out of 5 treatment modalities.

And I found this particularly interesting because the notified body… I mean, there were reasons in this case to believe that this device could be used off-label, with a certain, easiness. So they wanted to know how much off-label use, and there was no specific methodology to check off-label use. outcome? the PMCF study, which had been planned, but it had to be redesigned, they had to do a powered study. Powered, I refer to the statistical power. And particular attention to off-label use, startup was delayed, etc. This is still ongoing.

As we speak, my last case study, it's a positive one, so it's Class III catheter, transient use. We all know it's well known and widely used. No residual risk, no major residual risk was identified. Even in similar device, the method proposed was… a specific method was a Level 4 survey. Data were collected on 70 patients. The plan and the results was accepted, survey on time, on budget. And then we have a case study that Diane kindly shared with us. Let me remove the… this.

Thank you, that's good, no worries. I just want to comment about a couple of your case studies, if I may, for some time. What was interesting for me was… Almost every sort of negative one where they had to rework, there was always a gap. And so that's… that's exactly what I'm talking about in my previous question was, I need to see that you have… In your clinical assessment and your risk management that you've clearly identified. any gaps in age, in variance, in modes, in any of it, because the expectation is that you will have clinical data to support the whole spectrum. We know it's never perfect.

So then you're going to provide a justification of why, you know, you believe is still clinically going to perform as intended and safe, and you're going to target that and get that data in your PMCF. That is the expectation.

Alright, would you like to add anything on this particular case study? Oh.

I think she's frozen.

Okay.

So, let's, run the polls.

Let's move on.

Never. Yeah, I'm gonna run the poll question now, and, keep moving while she logs back in.

Okay, oh, there she is.

You're muted.

Sorry about that, I obviously dropped. I don't know where… what you heard last, and there's a black square in my… slide there. Sorry, I guess I dropped, I dropped the link. So, I just wanted to make sure you heard that I said that, you know, the expectation is any gaps.

Yeah, yeah, yeah, we did, we did hear that, yeah.

Okay. So here's an example from me, and I think is similar to your Class 3, that even though it's a Class 3 device, like my example here, it, you know, delivers into the cardiac space, it's a guide catheter that delivers other devices. So, high-risk anatomy. But, you know, just because it's Class 3, sometimes a simpler, survey can work, because, as you mentioned in your other studies. The residual risks are known. Similar devices have the same safety profile, and you build that story. And you justify why a simpler survey such as this one that looks at the important parts for a user, you know, the torque, the valve seal, you know, fluoroscopy.

That you can use something like this. It's easy to scale up and, and send, add questions for similar devices to compare, or send to various, other sites, or repeat the same survey to get trending data. from, and make it even more powerful. So just because it's a Class 3, I think it doesn't… have to always be, you know, have sophisticated approaches. It really depends on the state of the art, your control of the risks, and the story that you're telling us to, to show that this data is sufficient. I accepted this, study for a… or, sorry, this survey, for a Class 3, device with no… no problems.

Right, thank you. So, I think that we are at the end of Part 1. Correct, Mattel? Because my screen is now shared, so I'll stop sharing, and you.

Sure.

Yours? Okay.

Thank you, Daniela. So, while I'll start sharing my screen, I just want to briefly introduce myself. So, I am Matteo Mosso, Director of Clinical Affairs here at Dano Wildlife Science, and I have been in the clinical environment for over 25 years, with, the last, almost 13 years, dealing with medical devices. So I hope you can see my screen. Yep. Okay, good.

Yeah, yeah.

Thank you. So, I will be, focusing on, the various specific methods for collecting PMCF, data. In my presentation, let me see if this is working… Oops. Okay. All right. So, first of all, it is important to be aware that whichever is the specific method that you are going to, use, of course, this must be somehow documented and justified. And as we will see, of course, depending on the various types of devices that are on the market.

there is, always… there is not one unique solution that can, fit the purpose, for, for addressing your needs in terms of PMCF, data to be collected. As already, Daniela has showed this in her first part of the presentation, there are many aspects that contribute to the decision on which type of PMCF method I should use to collect the data that I need. And of course, we need also to take into consideration that, you don't need to use only one method. It can be more than one, so… and we will see, following, in the following part of the presentation, which are those specific methods that are…

To be used to collect PMCF data. So, as Daniela already, showed, we see… we have four main, type of source of, of data for PMCF. And, of course, the most known are PMCF studies, but, aside of that. There are other different metals to collect the data. So, we have seen real-world evidence, and we will see in the presentation that the definition relative to real-world evidence actually is so broad that it, it, includes, also, somehow registries, surveys and, and studies. So there is no

We can say that, to some extent, there is no real distinction in certain circumstances. And there are, as said, registries and surveys, and we would see pros and cons of all these methods, and when it's most appropriate to use one rather than the other. First of all, I wish also to highlight that all those methods are relative to the highest level of quality and strong evidence, as reported in the MDCG 2020-6 Appendix 3, table, we… we will be relying on the, absolutely the first four,

categories, and also we will include, surveys which are normally, known as Level 8, Level 8 surveys. So, talking about studies, if we look at the NXI of the ISO standard 14155, we will be falling under the let's say, the area that is relative to the post-market. So, in this area, there are, of course, different types of studies that we can perform. Those that can be, interventional or observational. Or, those can be either, registry, and, and we will see further on that,

The type of studies that we can undertake can be not only prospective, but also retrospective. So, when it is more appropriate, let's say, to start a clinical study to collect PMCF data. So, normally, let's say that it is a good approach when we are talking about high-risk devices. or when we have a high level of innovation. This, this is because, Either we… we need to, provide… solid performance and safety data, which can be obtained almost uniquely, by performing a clinical study.

Alternatively, it is also useful when we have very specific defined data gaps that we need to address, because, of course, a clinical study has a limited amount of data that can be collected. We do not have a wide, broad of objectives. Normally, a clinical study has very defined, objectives to be satisfied. And of course, also when, for example, the CE marking has been obtained on equivalence, so there are no, clinical data, collected, so far. The advantages of… of using, of course, or performing clinical studies is… is that… that the high… they provide the highest quality, and

They also, are subject to none or very low bias due to the several, let's say, procedures that can be implemented, such as blinding or, randomization. As already said, the fact that they will allow to address very specific gaps. The downside of running a PMCF, let's say, study is, of course, the cost, which we all know are pretty high. The time required to set up, plan, prepare. receive the approval, conduct and analyze the data, so the length of the time required to obtain the results. And of course, the fact that having only limited number of endpoints or objectives, of course, there is a limited… the results are of limited use for generalization.

Moving forward to real-world evidence, I have, put here the definition from the FDA guideline, and as we can see, the definition is so broad that, of course, it does not exclude also some of the other metals that we will see moving forward. So… when it's useful to, let's say, rely on real-world evidence. First of all, when we are speaking about devices that have been on the market for a very long time, so we consider, for example, well-established… established technologies.

Or if we want to confirm the long-term safety, or we need to verify if the device has late adverse events, or if some rare adverse events occur. And as well as so, trying to, let's say, address the fact that if we want to verify whether, the real-world performance is, let's say, consistent to what has been, approved. The advantages is that we can rely on very broad coverage, so this is in terms of number of sites, regions, countries, different type of users, and the fact that we are involving unselected populations, which means that this includes also

The extreme age ranges, which are normally excluded from clinical studies, or if patients have certain comorbidities that also may exclude them from the participation to a normal study. The other advantage is, of course, the fact that we can access a big number of patients. And, let's say compared to the… to the… to the clinical studies, in general, let's say that accessing this type of, I mean, using this type of, of approach, has somehow reduced cost. Compared to clinical studies, and as Daniela had mentioned previously, also we can rely on third-party providers who have commercial database that can satisfy your requirements.

On the downside, of course, There is the fact that the data collected are variable in quality and reliability, and of course, not having controlled studies, there is always the possibility of introducing bias. or confounding factors. And also, if we speak of novel high-risk devices, in certain circumstances, this has limited, applicability. If we want to see, more or less, an example of various real-world data sources, we speak about electronic health records, or patient claims, or patient reported outcomes.

real-world data include, as I said, registries, a commercial database, or different type of studies. And in certain occasions, depending on the device type, you also may have the possibility that the device automatically records and collects parts of the data that you may need. So this is also another option to keep in mind. So… Okay, I hope you see… The slide entirely, sorry. So, for… for registries, Also here, if we see the definition taken from the ISO standard, of course, it is pretty vague and gives the opportunity several different interpretations. In fact, it's not very uncommon to have, and we have witnessed this, that in certain cases, the term registry

It's often used as referring to an observational study, which, if we want to be Specific, actually, it is not totally correct, because for registries, we intend a different type of, of… approach, and I will provide examples of, registries, just further on. So… Similarly to what we have seen for real-world evidence, of course, the… the… when is it, use… I mean, when is… when is it to be used for collecting data? So, as similar as I said, when the device has an extensive clinical use, or if we are interested in having information on the long-term follow-up, which,

normally is not captured by regular clinical studies, where follow-up, you know, are 1, 2, 3 years. It's difficult that the clinical study extends to over 10 years, which registries normally do. Or, as said, also to monitor delayed or rare adverse events. The same as for real-world evidence, I can see that… I can say that pros and cons are more or less very, very similar. So here, just to show an example of what we consider a true registry that normally are systems that are held at national or international level, and they, involve mainly implantable devices, so we can see that there are, registries for

processes, for joint processes, or, for implantable cardiovascular devices. As also Daniela mentioned, there are cases… I mean, there are also other kinds of regi… type of registries for breast implants or other type of surgeries. So those are more or less The typical registry that we are referring to. The last method is relative to the surveys, so… Of course, there is a lot of discussion about surveys,

And we will see in the next slide some key considerations. So, when are they mainly appropriate? So, if you want to collect usability or user experience information, for example. Or if you want to identify use errors or off-label use, this is also another way to proceed. And, given the level of evidence that the surveys are able to provide, they can be considered as a complement also to other PMCF methods that we have seen. So they can be used together with other methods.

Despite the fact that they are considered to be simple, fast, and quite inexpensive. We will see in the next slide that actually, let's say that implementing a survey does require a certain amount of activities. The other advantage is that in certain occasions, and depending on the type of survey, you have… you may not need ethical approval, and so you have reduced regulatory burden. The downside of the surveys is that if they are not well designed, of course, they lack of systematics, and they are, of course.

susceptible of bias. If the question asks, for example, what have been the bleedings that you have experienced in the last month, it is possible that the user of the device May not, recall correctly all the information, or he may be biased to, provide, I mean, better results than what happened in reality. Also, this is due to the fact that surveys are not controlled, so there is no monitoring on the data, and this is another problem. So, let's say that in After all, they have a limited ability to provide data regarding safety and performance.

so, yeah, additional consideration regarding surveys. So, regarding the fact that we can have Level 4 surveys which have a high degree of evidence compared to level 8. So Level 4 can be both retrospective or prospective, while on the other hand, Level 8 surveys are exclusively retrospective. there is a wide range of surveys, and they… they can be, really useful or totally useless. So, we had experience where, And Daniela has discussed about some examples of surveys that

did not satisfy notified bodies. And of course, as I was anticipating previously, preparing a survey is not always all that simple, so because you need to describe a survey protocol, which shows and reports how… what are the number of cases that you are going to analyze? What are the questions relative to the endpoints that you are addressing? What are the acceptance criteria, the timelines that you require, and also how long is the questionnaire? Because, of course, if the questionnaire, the survey is too long. There is the possibility that the person who completes it loses the interest and provides only partial information, so…

must be, it must address the data that you need to capture, but not to be too long. Also, how is it designed? What is the programming? How is it going to be implemented and distributed to the end user? You need to use a validated software, so it cannot be just a simple program, but the data needs to be collected through software that are validated. And of course, not to forget that you should avoid collecting personal information of the patient, so you need to be in compliance with the requirements of GDPR, and if it's in the US, also the HIPAA.

We have seen, and we have experienced, of course, that poor planning has caused to either discard all the results or to require the necessity to repeat the survey from scratch. And here, Diane, I have a question for you. So, how about… how often do you see surveys that, they are not really proportionate? And, are there devices types where this becomes an issue from your, reviewer… as a reviewer standpoint?

Yeah, I think you covered a lot of it, where the surveys are more, preference data and more of a marketing tool than they are assessing, clinical performance and safety. So these, These surveys are… can be part of… of, the clinical data package, so obviously a device that's easier to use, and, users like it, doesn't have as many steps, so they're… they could be safer, because, you know, humans tend to… try to find a way around all the multiple steps or whatever, so they usually like devices that are simpler, so that could be safer, but, but they're really only, they're not standalone for clinical data. They need other, data as well.

Okay, thank you, thank you so much, Diane. So, I see that time is running short. So, I will need to rush for the next part of the presentation. I will try to be as quickly as possible. So… what are the key considerations for planning, about PMCF collection? So, the advice is to start as soon as possible, and to start with performing a gap analysis. In order to compare, what are the existing data, towards, the required evidence that you need. try to prioritize the data gaps that, of course, are relative to safety and performance, which are the most important. And,

For sure, you need to justify and decide on what is the most appropriate method to collect the data that you need. In terms of implementation, so where do I find the data that I need? Starting from list of customers, or sites where the device is used, or it is sold. try to access what type of data are collected as a routine or as a standard of care to see whether any of those are useful for you to respond to your needs. Verify if there is an interest by physicians or healthcare professionals in collaborating and sharing the data that they collect.

as previously said, verify and assess if your device already collects partially of the data, and last but not least, also involve the marketing team or the sales team. Of course, this needs to… These people need to be educated and trained, because we have, very often seen, sites suggested, or, healthcare professionals suggested by marketing who had no idea of, of what a clinical study is, or, how to, respond to, to the needs, from. from… the requirement to collect, to collect data. So it is very important that, manufacturers, have,

Joint meeting between, the regulators and the clinical and the, marketing or, sales team. So, as a concluding remarks, so, use the guidance document wisely. Try to define, what are your needs for PMCF data collection that are specific to your device. Select the most appropriate method and justify this method and document it, and early planning is crucial.

And I'm so…

Yeah, thank you, and this is it.

On that note, I'd like to thank the speakers, I'd like to thank everybody for their time today. Daniela or Diane, do you have any closing remarks as we sign off? I would like to encourage you to contact Donwell Life Science. if you'd like to discuss your PMCF needs, Daniela and Matteo are available to speak to you, so… Yeah, it's a free consultation. I encourage you to take full advantage of it. Thank you from me, Diane. Any closing remarks?

Yeah, thank you, everyone, for your time and attention to this. I'm just gonna close with saying I think post-market activities are going to become more important with the targeted MDR changes. You're gonna see, you already see it with the FDA, so stay tuned.

Yeah, thank you, it's interesting, and the audience and the attendance shows how this, topic is hot. So… Yes, stay tuned, and we'll meet somewhere, probably, maybe.

Oh, and if any of you missed part one, which was pre-market clinical studies with Mateo and Daniela, please just write pre-market, in the chat, and I'll make sure that you get a copy of that webinar. And we actually did a course with that, if you want that, it's free. So, yeah, great. I see a lot of people writing pre-market.

Yeah.

Thank you all, sorry we went a little bit over, and it was a great session. Oh, as far as the QA is concerned, I will circulate that to the speakers, and we will try to respond to you directly. Might also write a recap article, which includes some of these, so look for that in the next week. But overall, fantastic, wonderful, thank you, and thanks, everyone, for your time.

Thank you all.

Yay.

Bye-bye.

Thank you.

Bye, thank you.

Bye.