The compliance dates for the European Medical Device Regulation (MDR) and In Vitro Diagnostics Regulation (IVDR) are quickly approaching. In response, Treximo is hosting a webinar series that will cover a wide variety of pertinent subjects applicable to the device / diagnostics industries as they relate to the change in the regulatory framework in Europe.
The second webinar in our series was held on November 10, 2020 and focused on “Understanding the Concept of a CEP in the Context of MDR.” Treximo’s Senior Vice President of Device and Diagnostics Ronald Freeze hosted the session.
A full transcript of the webinar is below along with a link to download the slides.
During the training, Treximo covered how to prepare for a compliant clinical evaluation plan for the MDR, including:
- required elements laid out in the MDR
- expectations for legacy devices
- standard expectations not detailed in either the MDR or the Medical Device Coordination Group (MDCG) guidance documents
Part three of the series will be Tuesday, January 26th and feature a panel discussion and Q&A session with TÜV SÜD, one of the world’s leading Notified Bodies providing conformity assessment services related to medical devices. Register here.
Welcome everyone to today’s webinar. This is the second in a multi session webinars series that we’re doing here at Treximo. And today we’re going to talk about Understanding the Concept of a CEP in the Context of MDR. And so first, we’ll give a little bit of overview of our training.
So, our training is monthly sessions. Again, as we mentioned, this is the second. The idea is that this is for both MDR and IVDR. That being said, we knew that MDR was the more pressing topic and the enforcement date is coming very quickly next year. And so these first two sessions have been more focused on MDR, but next month we’re going to take a much different approach and go much more broad. But we hope you enjoy the deep dive today into an MDR topic. The idea we want to make sure is that we do continue to focus on topics that are most pressing for you and your concerns.
And so we mention later on this slide as well, is that we do use a survey to hear from you about what you’re looking for. Regarding content, again, this is a very technical presentation today, you’ll feel that, we do a lot of medical writing support in our organization, we support a lot of clients. And so that’s what the topic is today. And what you’re going to be hearing from in the series are different experts sharing their knowledge with you on those different topics. We do want to look at this as being very cross functional. We know MDR and IVDR cuts across all groups. And so we’ll continue to transition across different functions as we go through this series. And finally, we want to keep it very interesting. So what you will hear, is case studies, you’ll hear real examples that we see as we deal with different clients. And hopefully, from those case studies and real examples, you can take something away to help you more efficiently manage this transition to MDR and IVDR.
And finally, on the mechanics. Today, we are going to use the Q&A section of the technology. So, if you have questions, you can pose those questions in the Q&A session as we will be monitoring that. Today is a very full session with a lot of details. So we might not have time to address the questions at the end. If we don’t, we will make sure to get back to you on your questions. And at the end, you’ll see how to contact us here at Treximo to address any questions that you have. After the session, there will be the slides available as well as the recording, and then we’ll just reemphasize one more time, we want to make sure to maximize the value for you. And so you will be seeing a survey after this where you can ask us any questions you have. You can pose recommendations for future sessions and let us know if there’s anything with this training that you would like to see improved as we go forward.
We already have session three planned for December eighth Tuesday. As I mentioned before, it is going to be more broad, it’s going to focus on both IVDR and MDR. We are finalizing the details for our guest speakers, and we hope by the end of this week to be able to communicate the details about what that session will cover. And we hope you find that exciting and we hope you find this session exciting, too.
Today’s focus is to hopefully help you prepare for a compliant clinical evaluation plan to the medical device regulations. So, to start with, we’ll take a step through the required elements laid out in the MDR specifically Annex 14 part A1 (A), then we’ll talk briefly about the additional expectations for legacy devices, as highlighted in the MDCG guidance document 2020-6 Appendix A. And lastly, we’ll cover what some might consider to be standard expectations, not detailed in either the MDR or the MDCG guidance documents, but those originally highlighted in MEDDEV 2.7.1 rev 4, they keep giving some folks some problems by way of feedback from the notified bodies.
So the required CEP elements are related to the GSPRs, the General Safety and Performance Requirements. The intended purpose, intended target groups along with the indications and contraindications, intended clinical benefits, methods used for examination of clinical safety, residual risks and side effects, acceptability of benefit risk ratio for all of your indications and intended use based specifically on state of the art for that device.
And if applicable, addressing health benefit risk issues relating to pharmaceuticals, non-viable animal or human tissue are going to be addressed, followed lastly by the clinical development plan.
So, the first required element then is to demonstrate conformance with the general safety and performance requirements, also known as the GSPRs. It’s important to document which GSPRs you’re claiming conformance to not only in your CER, but also in your CEP.
So article 61 of the MDR notes that the clinical evaluation must conform to GSPRs 1 and 8. Number 1, like all things MDR tends to be a bit long and wordy, hence we’ve bolded the key points here for you. Your device needs to be according to its intended use, be safe and effective, and have an acceptable risk benefit profile when compared to state of the art.
And number 1 really goes back to risk. So, all residual risks should be acceptable when compared to the clinical benefits afforded to the patient. So just like the MDD notified bodies may be expecting to see more than just these two.
Depending on your device, you may also need to demonstrate compliance with GSPRs, 2, 5, 6, and 22. These are the top-level ones here. Some of this will depend on your device type, especially at the early stage when notified bodies are still kind of working their way through their first MDR reviews. And of course, the disclaimer here is, please read through all of the GSPRs yourself and make sure all the relevant ones to your device are included. This list isn’t meant to be a complete list, it’s definitely not like the MDD with the essential requirements where it’s a one size fits all, where we’re all only complying to 1, 3 and 6. There’s a little bit more work into it this time.
Now, we’re not going to read each of these to you as you will have access to the slides. But 2 and 5 are really focused on risk. And being in that risk and clinical are so closely intertwined. This isn’t a huge leap to understand why reviewers may ask to see compliance here. Now, GSPR number 6 is essentially expecting you to ensure that your device will continue to demonstrate strong performance over its entire lifetime. This of course, wasn’t added to article 61 as the requirement because some devices are only transiently used, so wouldn’t always apply. But if your device is a pacemaker, you can be assured that compliance to GSPR 6 will be expected evidence likely by clinical data. And then GSPR 22 relates to ensuring safety for end users that are lay persons. So for example, if you have an infusion pump that’s intended to be used by the patients or families themselves, what has been done to ensure operation of that device can be done safely?
So we’ve bundled the next two elements, both intended purpose and intended target groups, indications and contraindications together as they really work hand in hand. Per MDCG guidance document 2020-6, it actually refers you back to MEDDEV 2.7/1 rev 4 which notes how important the device description is, and inputs into the clinical evaluation plan that they are in line with the devices “quote, label, instructions for use, promotional or sales material, or statements.” In other words, these pieces lay the groundwork for the evaluation. And from a reviewer standpoint, this was really my starting point, is the IFU.
So, it’s helpful to understand what the difference is between intended purpose and the indication statement. MDCG guidance document 2020-6, did us a huge favor in walking through these, as most people use these terms interchangeably.
An indication is the clinical condition that’s to be diagnosed, prevented, monitored, treated, alleviated, compensated for, replaced, modified, or controlled by the medical device. In other words, what is the problem that needs to be fixed? While the intended purpose or intended use and both terms are used synonymously in the MDR. That really refers to the effect of the device on the patient.
So if it helps you to remember, think of the indication as the cause and the intended purpose or intended use as the effect. So here’s just a quick simple to help this hit home a little bit more. And that is, the indication for a stent might be to treat chronic total occlusions, while the intended purpose might be to improve coronary luminal diameters in patients. So hope that makes sense.
This should explain why you can actually have devices that have an intended purpose, but not an indication. So a good example of that would be like those tinted contact lenses that their only purpose is to give you baby blue eyes, but they don’t actually correct any vision issue. They don’t actually have an indication, but they have an intended purpose. And the reason we bring this up is that this leads us to the most important point in then discussion today.
Yes, this actually probably should be the biggest takeaway of the day.
One of the most important and common issues that we’ve seen goes back to the lack of focus and update of the indication statement. Many are old and were written so broad many years ago that they’re almost impossible to support with clinical data, because of the fact that they were made years ago, and quite honestly, they just haven’t kept up with the changing clinical requirements.
So now you have a device on the market with an indication statement, far too broad in context of the supporting clinical data. And no company is going to voluntarily tighten their indication statement. But things in the past few years have changed significantly. This was made loud and clear, with MEDDEV 2.7/1 rev 4 on clinical evaluation. This push for clinical data has and will continue to be a hot topic for European regulators. And thus you will see the notified bodies following suit and pushing all the manufacturers on this as well. If you went through an early renewal of one of your devices in the past couple of years, you were likely pushed on this as well. So expect the same or even more under the MDR.
You’ll need to clearly define what the target indication is and who specifically the target populations are and who the end user is. Is it a healthcare professional or is it a lay person? So, let’s walk through an example as a reviewer would.
So we have an ablation device here and the indication statement reads surgical ablation device XYZ is intended to ablate cardiac tissue during cardiac surgery using radio frequent energy. Okay. So now we understand the ablation device is to be used in surgical versus interventional ablation. That’s a really good distinction. But it’s not clear if this can also be used during minimally invasive procedures. Okay, so now that’s on your radar, you’ll take a note and look for that clinical data to support it later. But then your next thought is, the indications seem to be missing the medical condition to be treated. So you’ve now looked at the eye view, and you can’t get this clarified, which would make you think, why would we want to essentially burn cardiac tissue during an open heart procedure? Well, it’s probably used to treat atrial fibrillation, right? That’s the only rational decision here.
But really, it’s funny is that you can’t find anything that will support that. This means that you also don’t know what types of… If it is AF, what types should be treated? Is it paroxysmal? Is it persistent? Is it permanent? And without that clarification, you will now need to assume that there will be clinical data on all three types. So, add that to your list. And what about the target population? Okay, so that seems to be missing. So, you will then assume that it can be used in all specialty groups: peds, women who are pregnant or breastfeeding, those who have active endocarditis, and those who are considered to be at high surgical risk. So do you see the rabbit hole that this can actually lead someone down? Basically, by keeping a broad indication, you’re leaving yourself open to proving that you have data on all relevant conditions, and in all populations niche or not?
Now I realize that tightening your indication statement might be hard sell within your organization, but remember, you also have to talk about the clinical benefits afforded to the patients for having this procedure done. So, if you’re looking at this example, if the indication statement isn’t to treat atrial fibrillation, what kind of clinical benefit would come from a heart that now looks like it’s been sitting on the grill for a few minutes?
One way of doing this from a very high-level view is to simply go back to using a table format to assess what is sufficient clinical data. So, here’s a quick example using a transcatheter valve. Parse out all the indications and the target populations that are to be covered. This will force you as the clinical evaluator to put all the pieces together, and then make it clear to the reviewer that all the bases were covered.
And they’ll likely spot check this for accuracy and then move on. We don’t have to cover it today, but know that MDCG guidance document 2020-6 Appendix III, is a source of what is considered sufficient clinical data according to the device classification. So, we hope this example gives you just a high-level overview of exactly how you can parse out what your clinical data needs to look like, according to the indication and the target population.
The next required element is to provide what the intended clinical benefits are afforded to the patient. Notice its intended clinical benefits and not clinical benefits. Because in some cases, there might not actually be data, such as when you’re putting a device on the market for the very first time, and you actually have no clinical data. Now, this isn’t going to happen very often, but it would be most applicable to lower class devices, or to devices using the equivalency strategy. The key here is to understand how clinical benefit has been defined.
So per the MDR, a clinical benefit refers to the positive impact of a device on the health of an individual expressed in terms of a meaningful, measurable, patient relevant clinical outcome, including outcomes related to diagnosis or positive impact on patient management, or public health. MDCG guidance document 2020-6 took it a step further, by adding that the benefit can be either direct or indirect. So, a good example here is standard guidewire. In and of it by itself, it doesn’t actually provide any benefit to the patient, but when it helps the healthcare professional safely place a stent, it’s indirectly adding clinical benefit. And for the most part, manufacturers are pretty good at this section. It’s usually clear what the benefits are, but the requirements of needing it to be measurable is new. Which brings me to the third reference source for this area. And that goes back to MEDDEV 2.7/1 rev 4 Appendix A7.2, sections B and C. Section B talks about the nature, extent probability, and duration of benefits, while Section C talks about quantification, which can actually tie back to the benefits being measurable.
So, if you’re having trouble determining clinical benefit, you can always start with that document.
Section C specifically focuses on clinical study endpoints but also talks about the types of benefits meaning that there may be more than one, as well as a change in condition compared to before the device was being used. All the while clarifying if there is any variation and benefit across a population. And then what’s the possibility of a patient experiencing more than one benefit? Because we know that our devices do that, that’s for sure. And how long is that benefit expected to last?
A large benefit experienced by a small proportion of patients may raise different considerations than does a small benefit experienced by a large proportion of patients. So, if you have a niche patient population, talk about that, explain that, make it very clear to the reviewer that there is a large benefit here, or explain what the benefit or benefits are.
And it should be stated that not all benefits need to come directly from clinical data. And I know that that’s been a myth out there for quite some time. You might also be able to support some of this through basic bench testing certain products standard. So if there’s an ISO standard that’s related to your device, or even common specifications. Now I know that we’re really waiting for more information on common specifications. But know that that is a source that you’ll be able to use in the future too.
The next required element, CEP element is to explain the evaluation methods that you’re planning to use for the overall clinical evaluation as part of your CER.
So, many manufacturers have already been doing this specific to the evaluation of their lit reviews, but the change here is to choose a tool for your overall clinical evaluation. That said, and we’re pointing back to MDCG guidance document 2020-6 again, but it points you to several tools that you can use.
Now, we’ve only chosen here to highlight what we have seen as the most commonly used tool, likely because of its ease of use, even for brand new CEP and CER writers, and its testament over time, with references in several clinical evaluation guidance documents from the Global Harmonization Task Force years ago, MEDDEV 2.7/1, back in rev 3. They removed it in rev 4. And then the most updated International Medical Device Regulators Forum document of clinical valuation that just came out about a year ago. The choice of valuation tool is really up to you. The only requirement is that you use a tool that’s been verified and validated. And now there’s been also a misunderstanding and pushback here because the wording used for this element seems to focus on clinical safety, residual risk and side effects.
That means that’s leaving everybody wondering why this tool shouldn’t be limited to just the vigilance data. And that makes sense knowing that complaint data in general can give great insight into new or increase in residual risks, the specific patients and trends that the devices being used and in what they’re seeing. But the problem is, is it can’t provide you confidence intervals. So in other words, you may not be seeing the whole picture here, as your numerator might be wrong. And we all know this, not all complaints are reported back to the manufacturer.
So, now you need to address how you plan to assess the risk benefit ratio of the device based on state of the art. The key here is to have a full understanding of state of the art for your device. This is in general, an area that manufacturers haven’t really excelled at. It’s important to keep in mind that state of the art in context of a clinical evaluation, does not also necessarily imply the most technologically advanced solution.
And we think that’s a big hang up, people think that that device always needs to be the best and most tech savvy device and that isn’t always the case.
Now you need to address how you plan to assess the risk benefit ratio of the device based on state of the art. And the key here is to have a full understanding of the state of the art for your device. In general, manufacturers haven’t excelled at this, it’s important to keep in mind that although we don’t have time to go into great detail here, that at minimum, the intended purpose of the device and its prevalence, and what kind of alternative therapies they are highlighted in this section.
Now, this isn’t meant to be written like a thesis statement. So if your device for example is a total knee replacement system, you don’t need to start with talking about aspirin for pain years beforehand. You should be starting at the point along the patient’s care continuum, where your device or therapy would be considered, such as when pain meds or physical therapy have failed that patient.
You’ll also need to do a full assessment on similar and benchmark devices and if available, what type of professional guidelines or consensus statements say about this type of device. So what type of guidelines are relevant, and how are you ensuring that those will be included in your evaluation? I think one of the easiest way to do this, is actually through a lit search. And we’ll actually cover that in a little while.
The second step starts with documenting the risks and benefits associated with your device. Here’s where it is expected that you’ll be using qualitative or quantitative data, or sometimes both to make a sound analysis. This is best done with steady endpoints or acceptance criteria. And then there are other things to consider here as well, such as the duration of the benefit, the extent of the benefit, and even if it’s for a very niche patient population, if it is just call it out and talk about that.
And for risks, it should include the severity of risk, the number and overall rates. And then for more details, just based on time alone for today, if you have questions on risk, we would refer you back to MEDDEV 2.7/1 rev 4, and then again, it’s Appendix A7.2.
So, the next required element notes that manufacturers must include an indication how benefit risk issues related to specific components such as the use of a pharmaceutical, non-viable animal or human tissues are to be addressed. So basically, when your device contains a drug in non-living animal or human tissue, and that also includes their derivatives too, how is the benefit risk in using one or more of these components going to be considered as part of the clinical evaluation. Now, for the purpose of the discussion, we’re going to focus on animal tissue, only because it’s the most common and then, again, due to limited time here too.
We think it’s always easier to understand the background on why some of these elements are part of the regulation, it really helps to give some context when you hear the backstory.
So, with this slide back in the 1990s, you could read about mad cow disease everywhere. It was termed mad cow’s because of the neurological impact it would have on a cow’s behavior, but it’s also referred to as bovine spongiform encephalopathy, also known as BSE, which is a larger part of a group that’s called transmissible spongiform encephalopathy or TSE. Some of the cows that acquired the disease ended up in meatpacking plants with horrible consequences. BSE can’t be eliminated their typical sterilization process, it’s a prion disease, and prions are extremely hard to kill. And of course, if a person consumed that tainted meat, the person would now acquire what’s called Creutzfeldt-Jakob disease. And 177 people in the UK alone died from the disease. So, this was a hot topic for many years and is still a real concern across all of Europe.
So, to start at the beginning of this one, you need to answer the question, what is an animal? How do you determine if part of your device should be reconsidered and looked at if it has an animal component to it, such as a collagen impregnated graft?
Most people don’t realize that collagen is coming from…a portion of that is coming from animals. Well, the MDR doesn’t actually provide you a definition, but ISO 22442-1 does. And we realize that although the standard hasn’t been harmonized with the MDR yet, it was just updated actually just a couple months ago, here in September of 2020. So it will be considered state of the art in the eyes of the notified body, and it defines any animal as any vertebrate, invertebrate, explicitly excluding humans, seeks to reduce or eliminate transmission of agents from animal tissues or their derivatives, including bacteria, molds, yeast, parasites, viruses, transmissible spongiform encephalopathy agents, like the mad cow disease we were just discussing, and unclassified pathogenic entities to humans.
So, some common examples would be collagen, which is the one that we just mentioned. Cartilage, allograft tendons, demineralized bone, heart valves and pericardium.
Now, if this scenario doesn’t apply to you, that’s actually great. This will save you some time. But that said, please don’t simply skip this required element just because it doesn’t apply to your device, you will still need to address it. So, there are several ways of doing this, but we’d advise you to keep it short and simple. So one way is to create a section within the CEP and then simply note that this element does not apply to your device. Your device contains no animal tissue, so you can just say, this device doesn’t contain the use of a pharmaceutical, non-viable animal or human tissue. You can also do this by adding in a checkbox item, and we’ve seen that as well. So, if the answer is no, you can just simply check the box and move on.
Now, if your device does include animal tissue, then there are a few things to consider in the documentation of how this information will be considered as part of your clinical evaluation. So, the first one is, what part did risk management play in all of this? And are there some source documents that you can reference? The next is justification. Why did you choose bovine pericardium over a porcine pericardium? For example. Why aren’t you just using a synthetic? So, know that these questions are going to be asked of you, and make sure that in advance that you are looking at how you’re going to use that information to evaluate your device within your CER.
And then where are the tissues being sourced? So going back to mad cow disease, there’s a whole rating system based on where the cows are sourced, called the geographical BSE risk, or GBR for short. So if you’re choosing to source cattle from Australia, or any other place other than Australia, or US, which have the lowest rating of one, you’ll need to consider what type of impact that will have on your overall risk benefit of the device.
And then what type of inactivation or elimination studies have you done to make sure that there’s no living, that the tissue and nothing is getting into that tissue that you’re working with? And then what does the current literature say about the use of this animal being used for medical condition in the patient population that you’re trying to target? Is there anything from relevant professional guidelines, or consensus statements to support your decision in using that type of animal tissue? And then of course lastly, are you aligned with that new updated ISO guidance document 22442? And when it comes to what you need to document, our same advice would apply here and that is to keep it simple, and reference the source documents that you plan on reviewing as part of the clinical evaluation. The concept here is that the CER is really where you’re going to provide all of your justifications and rationales for all of these areas. But the intent of the CEP is to be transparent about what will be considered, and what are the relevant source documents that will be reviewed and included in the overall clinical evaluation.
So then, the last element is the clinical development plan.
So, before we talk about CDP, it’s kind of good to have a solid understanding of the differences between a clinical development plan and a clinical development strategy. Because both of course, are mentioned in the MDR. And that just leads to some confusion. So just quickly in general, a clinical strategy considers all of your product and marketing goals and translates them into some type of actionable plan using the right trial design, the right endpoints, the right sites and investigators, the right patients, and then the right data points being collected. So in the MDR, the clinical development strategy refers to the voluntary pre-CE mark consultation process with an expert panel where they provide an opinion on your proposed strategy and trial. This extra step is only needed for class three devices, or class two devices that are intended to administer or remove medicinal substances. Now that differs from the clinical development plan, which as its name implies, is essentially a plan within a plan.
So much like a matryoshka doll. That said, the CDP can live within the CEP, or can be a standalone document, whatever the manufacturer chooses, and there are benefits to both ways. So say that you’re developing a novel, a brand new device. Well, then likely your CDP might exist before you ever start working on a clinical evaluation plan because likely you’re looking in the initiation process, you’re looking at what you’re going to do from a clinical standpoint. And so you’re already going to have a document that’s already in the works. But if you have a legacy device, it might be easier to have it reside within the CEP so that you don’t need to update and sign off on more documents than are absolutely necessary. And to be perfectly clear, the MDR is not prescriptive about this, so do what’s best for your company. The other difference between a clinical development strategy and clinical development plan, is that the CDP is required for all classification of devices.
And as it notes there on the left side of the screen, currently there is no MDCG guidance document available on either of the clinical development plan or the strategy, but knowing that there’s an expert panel that’s being pulled together, we’ll actually likely see guidance on that in the future.
So now we know the difference, where exactly should we go here in the CDP? Well, as usual, it really depends on your device. We know you hear that quite a bit, but it really does. And the clinical story that you’re trying to tell about your device. So say for instance, that it’s a brand new, novel device, then maybe you have less information on what has been done in the past, which at a minimum would likely include some type of a pilot study. But the roadmap for where you’re heading might be far more robust with post market studies to support new indications steadily planned over the next several years. Or maybe your legacy device has been on the market for like 15 years, it’s well characterized, and there’s really no plans to change the good thing you’ve already got going.
Well, then you’ll at least need to take a retrospective review of the places you’ve already been. So, maybe a first unmanned study wasn’t needed, but maybe a feasibility study was done, followed by the completion of both a pilot and pivotal study. And maybe right now you’re just in coasting mode with one open registry. Okay. Well, then just describe that. But maybe your device’s somewhere in the middle, you’ve accomplished all the pre and post market investigations with success, but now feel it’s time to expand some of your indications. This is then the place where you need to document both where you’ve been and where you plan on going in the future.
And then, the last scenario that comes to mind is when a device is seeking approval via the equivalency route, which we haven’t really talked too much about already. In this case, you may not have any retrospective data at all on your device, but you will have solid PMCF plans because that will be an expectation.
And with any plan study, those “stops” along the way in your product roadmap, will have milestones. And with each milestone, you should have a description of the potential acceptance criteria. And we highlighted this, because according to MDCG guidance document 2020-13, your notified body will not only be reading the CEP, but they will also be documenting their review in the clinical evaluation assessment report. And one thing that has been missing from manufacturers, has been acceptance criteria. So, we would advise you to keep an eye on that one. The idea with all of this with the CDP, is really to provide transparency. So, to kind of cement this idea, let’s take a look at a quick example here.
So, here’s a tabular view of what we were just talking about. We’re not proposing that everyone use this method in your CDP. This actually doesn’t even give you enough detail, but what it does do, is it shows you how to approach it from a 10,000 foot view.
Using a transcatheter valve as our example, let’s safely assume that it was first CE marked for use in patients whose surgical risk level was considered either inoperable or higher, severe risk. Once the device was out on the market for long enough to support ongoing safety and performance in the short term, you decided to expand its indication to patients considered to be at intermediate risk, and then later you move that to lower risk. But then to further expand your indications, studies then began to look at patients who had a bicuspid valve. So, this one’s a little different, and that it doesn’t have anything to do with the indication, but it is its own niche target population, much in the same way we think about devices used for pediatric use. So, don’t always think of expanded indications, think also of expanded patient populations.
And just as mentioned earlier, this is why having a solid indication statement and a clear target population is so very crucial. This is the type of approach that’s really now expected, where you build on your indications over time, all the while being transparent about what you’ve done in the past and what your future clinical plans are in the future.
So, yay, we’ve come to the end of the required elements. So, let’s take a minute to cover legacy devices. And we mentioned this at the very beginning, but there are additional requirements for these devices, which most of us are dealing with. And those are detailed in MDCG 2020-6 Appendix 2.
So, we’ve already covered many of the requirements in the legacy devices, because they overlap with all of the elements from Annex 14. So let’s focus on all of the extras for these devices. So, the first addition is to describe how we plan to address our strategy and identify the clinical data for the device.
The reason that this one was added, was to highlight the fact that they changed the definition of clinical data to now include data from post market surveillance, which of course you wouldn’t have this if you were having a new device. That’s the reason it wasn’t originally included in Annex 14. So now you need to report on PMS activities data, in addition to what we have previously been reporting on, and that’s normal studies and or peer reviewed literature on the device or the equivalents device.
So, let’s say that you were doing a survey for example, what is your plan to assess the data that’s coming from that specific survey? Here’s where we would suggest you looking back at what you’ve decided your safety and performance objectives are going to be, and then align your PMS activities to ensure that you’re trying to capture that same information. So, it’s just a two birds with one stone kind of approach.
And when you’re looking at your post market data, be sure to address both proactive data, like post market clinical study data, and your reactive data, like the complaint data that you’ll be getting. So let’s quickly look at an example here using a guidewire.
A safety related objective might be… Guidewire A, B, C, will have less than X percentage device related vessel injuries like dissection, or perforation. And a performance related objective might be related to the technical access of actually placing that. And that is evidenced by greater or equal to X percentage. So now if these are your objectives, your CEP should explain how your PMS data, like your complaints, or your survey, is going to collect the right data and how that will be evaluated in your CER. The next is evidence on the equivalent device. And we all understand this very well, but do you know the required level of evidence? And this is where we would point you in reference to the MDCG 2020-6, Appendix C.
So for example, it talks about the fact that if you have a class three device, and it’s an implantable legacy device, which is not a well-established technologies, and we talked about that previously, that those are what we consider to be WET devices, well established technologies, that they should have sufficient clinical data that are rated at a minimum of a level four. And then also noting that leveling is also detailed in that appendix. So, use that leveling to help support your decisions on what you’re going to use.
And for this last section, we’ll focus on areas that have already been required for MEDDEV 2.7/1 rev 4 but were commonly missed and continue to cause rework for manufacturers.
So, in both writing and reviewing these, there are a few key things that were commonly missed. And the first thing we’d like to point out is accessories.
If you don’t have anything, if you don’t have any accessories, that’s obviously fine. But just don’t skip it, make sure you add a section there and say none. It then makes it clear to the reader of the document that there are no accessories, instead of making that reviewer wonder if the accessories were just forgotten about, and then now they need to raise a question and get that clarified. And the second is to understand the impact of basic UDI, per MDCG 2018-1 V3. Now, this one says that multiple devices can share the document only if they also share the basic UDI-DI, meaning that they share the same intended purpose, risk classification and design and manufacturing characteristics. The CEP should align with the CER. So, if you have multiple devices with differing indications, they can’t share the same CER. And if they can’t share the same CER, then they also can’t share the same CEP.
And then the third thought here is, this won’t be needed for initial submission so we want to preface that. But it would be to make a plan on what the device changes have been made since last assessment. This are frequently missed items for sure. And then, of course, don’t forget about claims. How are you determining what claims need to be in the CER? So what sources are you using to figure that out? Are you looking at your promo material? Are you looking at your website? Are you looking at your IFU? Anything public facing should be included. And that actually brings us to the clinical data. You need to include everything, even data not generated by the manufacturer. So do you know if there are studies on clinicaltrials.gov or on the World Health Organization website that have been studied on your device per its intended use? Under the MDD, you only needed to include data generated and held by the manufacturer, but now all data needs to be considered. So what’s your plan to ensure that you’re capturing all data.
And again, it goes back to, the CEP needs to talk… You don’t need to talk about what data is available, you need to talk about what your plan is to ensure that you’re seeing all of the relevant data. And then don’t forget to document the frequency of updates that will be made along with the document number, version and date. So, per MDCG 2020-13, the reviewer at the notified body will need to document all those elements in their clinical evaluation assessment report. So don’t forget them. Just make sure that they’re easy to find for the reviewer. And then lastly, there are usually gaps seen in the literature search protocol so we’d like to spend just a couple minutes on there.
So, the search strategy should be detailed in your CEP. And that should be explaining in advance how you’re actually planning on doing your lit search. This can include, but is not limited to the number and types of literature searches that you’re doing.
So depending on the type of the device and the state of the art for that area, you may choose to do one search for the subject device and another complete search for state of the art, or you could combine them into one. There is no wrong way, as long as what you’re doing gives you thorough data. And that’s the most important piece here. You also want to explain which databases will be sufficient. And we say databases plural, because many notified bodies will not accept just one database as being sufficient. So you need to find more than one. And then be clear about the timeframe of the search. There should be no gaps in literature from one CER to the next. So get some help on these next two if needed, but make sure that your keywords that you’re using are broad enough to cover all of the indications and all the target population groups that you’re looking at. And along with that, is using the name of the device in the search along with the model number. We rarely would ever see that happen.
And it is something that’s actually called out in MEDDEV 2.7/1 rev 4. And then speaking of MEDDEV, if you need help setting this up correctly by using one of the systematic search methods like we most commonly see, than I would actually point you to MEDDEV 2.7/1 rev 4, specifically Appendix A5, which was one of the sections that’s actually still in play with the MDR requirements.
So now part of the plan is to also discuss what you plan on doing with the output of the lit search, such as what types of things will qualify an article to be included or excluded. So for example, all articles need to be from peer reviewed journals. We all understand that. Typically the results come back and the abstracts are read through very quickly to pull out all the non-relevant articles. This diagram is just a quick example of what might be displayed to kind of demonstrate the process that you’re looking at. You might also choose to use some other type of diagram like a Visio or even write it out. And that’s fine.
The takeaway here is to document what your process will look like and what you’re planning on doing. The example here is actually pretty standard. The only thing that we would note here that we highlighted in red is that yes, it is acceptable to add in articles that were not included in the original lit search. Sometimes you’ll do a great lit search, and for whatever reason, your poll just maybe misses one or two, and that’s fine. Now, if you miss more than a few, the reviewer might question if your lit search terms need to be adjusted if you’re being too restrictive. But otherwise, if it’s just a few, you’re fine. You’re actually showing transparency when you do that. And in some cases, you should justify your decisions. So, for example, most manufacturers only look at articles written in English. We saw that all the time. However, if you’re doing a lit search for a company that’s in China, where most of the devices are being sold and used, then it’s likely that the publications on that particular device may be published in a Chinese journal.
So just think about your device specifically and document your justifications if needed.
And then you will also need to explain how you plan to grade and appraise the articles. Now, we’ve already covered this topic previously, so this is just kind of a gentle reminder that you can choose whatever tool that you want to as long as it’s verified and validated. And we believe that brings us to the end of the presentation today, I know that was an awful lot of information. But yay, we powered through it.