Clinical evidence under the EU Medical Device Regulation: More firework on the way.

fireworkThe EU Medical Device Regulation (MDR) and the In Vitro Diagnostic device Regulation (IVDR) entered into force, local governments published guidance how to deal with the new medical device regulations, there is debate on the implementation of the MEDDEV 2.7/1 Rev. 4, and yet the number of posts from my end in 2017 have been minimal. I am afraid this is only symptomatic of what is happening in the clinical environment of medical devices in Europe: the demand for qualitative and manufacturer owned clinical evidence is rising, and medical device companies have only 2 years left to perform a gap analyses AND to address any gaps in their clinical data.

A snap shot of some key events from clinical perspective in 2017:

April 5, 2017 the EU adopts the new MDR and IVDR taking effect 2020 and 2022 for the MDR and IVDR respectively,

May 5, 2017 the final versions of the MDR and IVDR are published,

May 2017: UL publishes a white paper on medical device clinical investigations and ISO 14155,

May 25, 2017, the MDR comes into force

June 2017, EMA announces delay of the application of the clinical trial regulation until 2019 due to challenges with the EU portal.

August 2017, TUV SUD announces extension of the implementation timeline of MEDDEV 2.7/1 Rev 4.

Throughout 2017: several posts on the impact of the GDPR taking effect in May 2018, one includes a comparative table EU vs US you can find here,

December 12, 2017: Dutch Health Ministry and Health Inspection publish guidance on key changes for medical devices,

December 18, 2017, MHRA updates guidance on EU regulations for medical devices and in vitro diagnostic devices.

In short, one can only conclude that 2017 has been an exciting year for medical devices reflecting an environment of change with more firework to come, especially when looking at it from clinical evidence perspective: Changes that concentrate around a more thorough (premarket) clinical evaluation (refering to my previous blog post on that topic) with a more solid clinical evidence base, and a better, stricter postmarket surveillance program that includes Postmarket Clinical Follow-up studies. Medical devices companies better prepare themselves, and if not started at this point in time, perform a clinical evidence gap analysis and start collecting their own clinical data where needed NOW!

Please feel free reaching out in case of any questions regarding this post, or when you are looking for support with your medical device clinical evidence strategy.


Posted in Uncategorized | Tagged , , , , , , | Leave a comment


igz-conf-medical-devicesThe Dutch Health Inspection (IGZ) organized an invitational conference on clinical research with medical devices end of last year. Representatives of the different stakeholders, such as manufacturers, competent authorities, Ethics Committee’s, and health care providers were present during this conference and discussed the current system on this type of clinical studies. The summary and presentations that have been published on the IGZ’s website (see above link) are interesting to read, especially on the following items:

EU guidance and regulations

In his presentation regarding the regulatory environment on medical device clinical research, the Health Ministry representative, Van der Kroef, once more underlined that with version 4 of the MEDDEV 2.7/1 and the upcoming MDR (final text MDR just came out!), more clinical studies and more clinical expertise will be needed (also refer to an earlier post). The Health Ministry therefore calls for better alignment and cooperation between the different stakeholders.

Off-label use

Roles and responsibilities were discussed, and the IGZ emphasized that although by law the manufacturer/ sponsor has the ultimate responsibility, also other parties such as the investigators have theirs (be aware that the latter is also amended in the new version of ICH E6). Specifically off-label use was addressed in this context: Research physicians shall NOT apply a non-CE marked device which is not delivered according to the law, but manufacturers are hesitant to allow for studies outside of intended use. Following the FMS guidance could be a solution to guarantee a safe application of medical devices applied outside of their indication when devices are developed within a health care unit.

Patient Information Letter

consentConcern was expressed with respect to the patient information letter that informs the candidate study participant on the ins and outs of a clinical study. Requirements seem to differ for different rules and regulations, and are complicating its development. Web-based tools are considered helpful for the structure and completeness of the content, but, again, the manufacturer has the ultimate responsibility to ensure that the patient information is brief and clearly written (refer to my previous post on tips & tricks).

Observational Postmarket Studies

Thoughts were expressed that the new regulation will also impact the way clinical studies are submitted and reviewed, and options for a “WMO-light” have been discussed: review by an acknowledged Ethics Committee of studies that do not fall under the scope of the Medical Research Involving Human Subjects Act (WMO). The latter would include any Postmarket observational studies such as registries, so I think this is an important development to keep track of.


The Dutch Health Inspection together with the Health Ministry and CCMO will explore what elements of the discussed topics can be addressed, and are aiming for follow-up in the first half of 2017. I will follow the developments with interest and keep you posted. Stay tuned.

Do not hesitate contacting me to discuss the above or when looking for support for your medical device clinical studies. You can find my details in the upper right corner, click here to send me an e-mail.

Posted in Uncategorized | Tagged , , , , , , , , , , , | Leave a comment



Ensuring proper informed consent in clinical studies is a critical process, that protects the rights, safety, and welfare of human subjects involved. It will require even more attention under the new GDPR and with the globalization of clinical trials. Although most involved are aware of its relevance, yet there is hardly a process where I see the same issues over and over again during monitoring visits and audits. The annual GCP Inspectors Report indicates that ICF findings were 5th on the list in 2015. This post addresses the 5 most common findings as well as options for preventive measures:


The purpose of the informed consent process is to inform the candidate study participant on what it means to participate in clinical research and the concerning study in particular: what are the risks and possible benefits, what are you supposed to do and not to do as a participant, and what it is you can expect with respect to your health. All of this should be done in language understandable to the subject (or authorized representative) and in as few pages as possible, to ensure that the candidate study participant can make an informed decision.

Yet, I often come across patient information letters containing complex medical and scientific texts, clumsily translated from an English master, and sometimes even in a language different from the subjects own. To simplify the information letter provided to candidate study participants, local (hospital) information brochures on the concerning medical procedure and/ or disease can be of great help, as well as asking team members to read and provide feedback on the information letter, and using the readability statistics tool in word. And last but not least, ensure to have the information letter and consent available in all applicable local languages.


With the requirement to keep the information as understandable and lean as possible, it is easy to forget to address some of the required elements. What I often find missing, for example, is a reference to the (potential for) personal data being transferred to a country outside of the EU: Under the current EU data-protection regulations, and even more so the new GDPR, the definition of what is considered personal data is very broad and also concerns pseudo-anonymous health data, and one needs to be careful sending such data to third countries with different (typically less stringent) laws in that respect (also refer to my previous post on this topic). Third countries, by the way, include the US, and one should pay attention now that the Privacy Shield is at risk.


Of note in this respect is also, that even though an Ethics Committee has approved an Informed Consent with missing elements, this does not mean you are compliant with the EU data-protection regulations: ISO 14155-GCP clearly states that:

“If national or regional EC requirements are less strict than the requirements of this International Standard, the sponsor shall apply the requirements of this International Standard”, and the standard requires you comply with …”

Tools that can help to ensure your information letter and consent form contain all locally required elements are templates and checklists provided by the local Ethics Committees, as well as your own checklists that you build with time.

Study participant sign off

Good clinical practices are very clear with respect to the fact that the informed consent needs to be signed and dated by the subject him/ herself. ISO 14155 states that:

“… the informed consent shall … include personally dated signatures of the subject and the principal investigator or an authorized designee responsible for conducting the informed consent process”

Still, I find it happening over and over again that a consenting physician or study nurse dates the consent in an attempt to help the subject. Despite good intentions, this really is a ‘NO GO’. The sign off namely indicates that the subject was personally and timely involved, i.e. before the start of the study or data-collection, and therefore needs to be done by themselves. Besides emphasizing this during the site initiation, the most creative preventive measure I have seen so far is an embedded “stamp” on the consent page that says

“the patient needs to sign and date him/ herself”.


Less than 1/3 of treatments are studied in children, which is a big issue for the applicability of medical products. One of the reasons is that research with minors is only allowed when it meets specific requirements, which includes the informed consent process. In general (the regulations differ per EU country) the informed consent needs to be signed and dated by both parents or their legal guardian when the child is below a certain age, and by both parents or their legal guardian and the child when (s)he is considered capable to understand the informed consent process until the age of 18.  When the child becomes an adult (typically 18) in the course of the study, in The Netherlands (s)he needs to (re)sign the (adult) consent. Requirements in The Netherlands are expected to be broaden in this respect per March 2017, such that re-consent is not required anymore and the transition age becomes 16 years conform the new GDPR.


With the need for more signatures, unfortunately the chances for mistakes multiply, and I have seen several partially signed informed consents. Also the re-signing when becoming an adult is something that is easily forgotten, so I am happy to see the broadening of the Dutch regulations in this respect. Nevertheless such mistakes are best prevented, and it is key to know the local requirements when running a clinical study with minors, to ensure adequate training of the site study team during the initiation in this respect, and to install alerts in your EDC system to warn you ahead of time when a subject becomes of age.


Every-one involved in clinical studies is aware that the (latest) EC/ IRB approved version of the informed consent should be used. In spite of that, even the best study sites make the mistake to use an old or wrong version. Often because the consenting physician did not have the correct version available after an amendment, or because (s)he did not realize there were different languages available.

The only way to prevent this from happening is to ensure that all involved in informed consent process only have access to the latest, correct version, and that all are aware (by emphasizing this during the initiation) that multiple languages are available should this be expected.


With the above it certainly feels like I am stating the obvious, but apparently there is still room for improvement in the informed consent process. Too bad, since there are plenty of tools and options around to ensure that this key process in clinical studies is handled the right way, and prevention is much, much better than cure; Especially since corrections tend to lead to new mistakes.

Should a correction be necessary then, as a rule of thumb, follow good documentation practices and ensure that for transparency reasons any change, addition, or correction is dated, signed, and explained by the person doing so (could be the study participant!), and involve the concerning Ethics Committee in this process.

Please feel free to reach out in case you want to discuss any of the above or are looking for a monitor or auditor.

Posted in Uncategorized | Tagged , , , , , , , , , , , | 2 Comments


The year 2016 has been a very interesting one from clinical perspective with many developments in the EU medical device regulatory environment, and I am sure we are facing an even more interesting one coming up:

The latest draft of the MDR demands much more robust clinical evidence than before, and is expected to be finalized first half of 2017.

The new version (rev. 4) of the MEDDEV 2.7/1 has serious consequences for existing and new clinical evaluations regardless the class of the device involved.

Given its broad definition of personal data, the new GDPR will affect (retrospective) non-interventional clinical studies, since an informed consent is required regardless the way the health care data are collected.

And last but not least, at a more local level, the Senate of the Dutch parliament approved the proposal to amend the WMO. The amended legislation aims to broaden the possibilities for medical research with minors and incapacitated subjects, and will come into effect in March 2017.

I am looking forward to an exciting year ahead, but for now wish you happy holidays. See you next year.


Posted in Uncategorized | Tagged , , , , , , , , , , , , | Leave a comment


During the congress on clinical trials for medical devices in Berlin, I spoke on the mutual acceptance of clinical trial data. A short version of my presentation you can find on SlideShare:


Clinical trial environment

Probably one of the main developments over the last decades has been the introduction of electronic data capture, making the process of data collection, processing, and cleaning much more efficient in comparison with paper. Besides enhanced efficiency of data-processing, this opened up the world for clinical trials as the collection and cleaning of the data can be done from any location provided there is internet. In parallel we see that over a period of 5 years the proportion of clinical trials run in the US declined with almost 20%, whereas the proportion of clinical trials run in Japan and Europe rose with an ~8%.

Running your clinical studies elsewhere only is efficient when foreign clinical data are accepted in the intended region, so looking into the

Regulatory environment

an analysis by RAPS indicates that mutual recognition is facilitated by the fact that the different GCP standards for running clinical studies are becoming alike. The differences that still exist are minor and should not be a barrier for acceptance of clinical trial data collected elsewhere anymore. For medical devices studies it is worthwhile noticing that this includes ISO 14155: 2011, which in 2015 was acknowledged by the ICH regions to be able to serve as a global standard in that respect.

Standards specifically addressing acceptance of foreign clinical data include the ICH E5(R1) on Ethnic factors in the Acceptability of Foreign Clinical Data for medicinal studies, which to my opinion is also of interest for medical device studies as for example it mentions the use of bridging studies,

a clinical study in the new region aiming to provide information on pharmacodynamics, or clinical data on safety, efficacy, dosage and dose regimen in the new region to allow extrapolation of the foreign clinical data to the population in the new region.

What makes it interesting is that this can help to extrapolate clinical data from the complete clinical evidence package collected elsewhere to the new region at relatively low cost.

Standards specifically for medical devices standards are the FDA guidance’s for drugs and devices in March 2001, and one (draft) specifically for medical devices in April 2016, and the  Japanese notification #479 issued in March 2006. In Europe the new version of MEDDEV 2.7/1, a topic recently blogged on, addresses that in case a clinical trial is conducted outside of the EU, an analysis whether data are transferable to the EU population is needed.

What the different standards have in common regarding the requirements for mutual recognition of clinical trial data are that

  1. Trials should be performed according to internationally acknowledged standards
  2. Clinical data should be transferable to the local population and clinical condition, and the
  3. Methodology should meet the local standards including scientific appraisal

With the GCPs becoming alike, conformance with international standards has become relatively speaking easy, but other factors such as ethnic differences may well hinder data transportability.

Local differences

Anatomical factors such as growth, muscularity, and size may well impact safety or efficacy, and underlying disease or different physiology could cause a different tolerability or allergy.

It is known, for example, that Asian eyes, or more specifically the anterior parts of the eye, are smaller than those of Caucasians. So clinical data from Asians on the performance of intra ocular lenses are not 1:1 transferable to Europeans as a lens suitable for an Asian may be too small for a European and start rotating. Thereby impacting efficacy, so size does matter.

Also the efficacy of barrier creams has been shown to depend on the type of skin applied to:  Caucasians respond different to the same barrier cream as compared to Africans, indicating that physiological differences can affect the performance of a device.

Another aspect that can affect acceptance of foreign clinical data concerns the study methodology. The local Standard of Care can impact the choice of the treatment in the comparator arm of a RCT and such may well differ from one region the other. In addition, study endpoints may differ depending on local requirements, the FDA for example requires effectiveness data whereas the EU requires performance data.

So with the above in mind what is it that we actually see happening with respect to the use of foreign clinical data?

Mutual acceptance in practice

Interestingly data from the Japanese Pharmaceuticals and Medical Devices Agency indicate that the acceptance of sole foreign data for PMDA submissions in the period from 2006 to 2011 is substantial, and after an increase in the first 2 years seems to stabilize around 60%.


Which contrasts with US data on PMA approvals (personal analyses, see above figure) in May, June, and July in 2001, 2006, 2011 and 2016, as that indicates that the use of foreign data, although slightly increasing over that period, is still less than 40%. This in spite of the fact that the overall proportion of clinical trials executed in the US declined to less than one third.

The challenge in Europe is that the medical device regulatory bodies are far less transparent on the (origin of) clinical data used for CE marking, so I can only speak of own experience that under MEDDEV 2.7/1 the approach has been much more the applicability of the clinical data rather than its origin. Subsequent local implementation, however, is often hindered by lack of local data, as global trials typically are not taking into account the local health care system and/ or reimbursement requirements, and you often end up doing another local post-market study for the same reasons as to why foreign data are not considered acceptable for market approval outside of Europe.


In conclusion, the clinical trial environment as well as the regulatory environment favors mutual use of clinical data. Internet and electronic databases facilitate clinical trial globalization, and standards address the way to deal with foreign data. The changes and proportion of clinical trials as they are seen in the trial execution, however, are not identical  to those seen in the acceptance of medical device foreign data in Japan but especially the US.

It seems that factors such as differences in the population, local practice, and requirements still outweigh the possible benefits of reduced duplication and faster access to innovative care. Although the system is slow, I doubt whether we will see further substantial changes and wonder whether in medical devices we should formally implement the concept of bridging studies similar to pharma.

Update 2018

February 21, 2018, the FDA published its final ruling on the acceptance of data from clinical investigations for medical devices. This will open up the options for use of clinical data collected outside of the US to support an IDE, 510k, or PMA submission. The rule will take effect February 21, 2019.

Posted in Uncategorized | Tagged , , , , | 1 Comment


consentAs mentioned in my previous post on the MEDDEV 2.7/1 revision 4, the second quarter of 2016 is characterized by several big steps from clinical evidence perspective: Besides the publication of the new version of the MEDDEV 2.7/1, consensus was reached on the new Medical Device Regulation (MDR), and in May the General Data Protection Regulation (GDPR) entered into force.

Since data protection and the informed consent for clinical studies are inseparable, in this post I want to share my thoughts on some key elements from clinical study perspective:

Health Care Data

Clinical studies concern any systematic collection of health data, and the definition of health data under the new GDPR is very broad, including

“all data pertaining to the health status of a data subject which reveal information relating to the past, current or future physical or mental health status of the data subject”.

Under the new GDPR such health data are subject to strict requirements: It prohibits processing of data concerning health, unless subject has given explicit consent to the processing of those personal data for one or more specified purposes.

This is particularly of note since this implies that informed consent is also applicable for registries or other non-interventional studies, whether these are prospective or retrospective. Therefore a substantial impact on epidemiologic studies can be expected where collection of informed consent can be very challenging due to scale for example.


In principle data collected within the context of clinical studies are anonymized before they are being processed. As addressed in another blog on transparency, however, the challenge with current technologies, especially when medical devices are involved, is that personal identifiers are often embedded within the diagnostic tools or the investigational devices (think images, ECG strips, …) to facilitate processing. Also the GDPR definition of what should be considered identifiable is very broad, and not just looking at a name or date of birth:

“an identifiable natural person is one who can be identified, directly or indirectly, in particular by reference to an identifier such as a name, an identification number, location data, an online identifier or to one or more factors specific to the physical, physiological, genetic, mental, economic, cultural or social identity of that natural person”

So besides we are collecting health data, this makes an informed consent an in-disputable part of any clinical studies, interventional or not, and even more so when medical devices are involved.

When in addition the study will be monitored and/ or there is a likelihood that non health care professionals will be reviewing the study patient hospital files, this needs to be mentioned in the patient information and the data-subject will need to consent to this.

Informed Consent

The informed consent process is not an easy part of clinical studies, and the GDPR indicates consent needs be specific and unambiguous, i.e. the

“Consent should cover all processing activities carried out for the same purpose or purposes. When the processing has multiple purposes, consent should be given for all of them”,

meaning for example that when blood or tissue samples are collected and stored for possible future use, subject’s consent is required for this, and research such as with Henrietta Lacks’ cervical cancer cells is a no-go.

The typical cut-off age for giving consent in clinical studies is 18 years, and I find it interesting to see that the GDPR indicates a child can give consent to the processing of his or her personal data at the age of 16, and below that

“such processing shall be lawful only if and to the extent that consent is given or authorised by the holder of parental responsibility over the child.”,

implying that – at least for observational studies – the age for giving consent could be lowered.

Also of note is that the GDPR refers to the Clinical Trial Regulation for medicinal products for the consenting to participation in research activities. Where does that leave us with medical device studies? Should we follow the Clinical Trial Regulation, or is it OK if we follow ISO 14155 – GCP? Since provisions with respect to the informed consent process are very much alike, I do not expect any issues, but still…

Consent withdrawal

Any study participant has the right to withdraw consent at any time without any explanation or any consequences, and the GDPR once more emphasizes this. Of note in this respect I find, that in the clause on the right to be forgotten (article 7), it says that erasure of data does not apply when processing of data is necessary for

“ … scientific or historical research purposes or statistical purposes in accordance with Article 89(1) in so far as the right referred to in paragraph 1 is likely to render impossible or seriously impair the achievement of the objectives of that processing; …”

In other words, in case a study participant withdraws consent, processing of anonymous data collected until that point in time is still possible if such is needed for the objective of the clinical study.

Data-processing outside of the EU

With the globalization of clinical studies in mind, it is of note that the GDPR is very comprehensive on transfer of personal data outside of the EU, since third countries are likely to have less strict regulations around data-protection. Aspects to consider in this respect are the location of the study sponsor and the database, but also of possible core-labs. When, for example, clinical study images of EU subjects are sent to a core-lab outside of the EU for screening purposes, the subject should be informed and consent to this specific aspect of the study before such is done.


In conclusion, generally speaking I do not expect many changes for clinical studies under the new GDPR. Most of the elements addressed above typically are already taken into account in the clinical study informed consent. The main challenge will be for retrospective and epidemiological studies: to obtain subject’s consent or ensure fully anonymous data processing.

Posted in Uncategorized | Tagged , , , , , | 3 Comments