European ParliamentNow that rapporteur Roth-Behrendt’s proposed amendments for the medical devices regulation proposal are out, it could not take long for  rapporteur Liese’s amendments for the IVD regulation proposal to also be published. And here they are.

Basically this proposal is not jerking the wheel in the other direction as the Roth-Behrendt proposal does, but is mainly carefully finetuning the Commission proposal from an ethical perspective. These amendments seem much better considered and less political than those of rapporteur Roth-Behrendt, that seemed to me largely copied from what was put in front of her by the reprocessing lobby. The first reactions to the Liese report that I heard are positive. I think the care that the rapporteur took to keep the proposal SME friendly is to be applauded. It does away with some of the translation formalities, applause for that as well.


Although the two proposals share a lot of procedure (the so-called ‘horizontal’ matters), rapporteur Liese turned out to follow the structure set up for general medical devices less than I expected. For example, PMA for the category of IVDs proposed to be subjected to scrutiny (class D) did not materialize. Additionally, he puts in an interesting and potentially far reaching proposal to change the definition of ‘medical device’ substantively, which affects both the IVD and general medical devices proposal (more about that below).

The rapporteur tabled a number of specific items that I’d like to discuss (also because the rapporteur himself considers these the important ones):

Companion diagnostics

Companion diagnostics are looked at more critically. They can’t be subject to the in-house exemption and must undergo design dossier review like class D devices.

Genetic testing and Informed consent

A significant amount of detail is added to the genetic testing part by adding a prior to use mandatory informed consent procedure for use of the device on the patient’s sample based on

  • “appropriate information on the nature, the significance and the implications of the genetic test”
  • provision by the physician to “the test subject concerned with appropriate and comprehensible genetic counselling without prejudging the outcome. The genetic counselling shall include medical, ethical, social, psychological and legal aspects”
  • while the ” consent shall be given explicitly in writing. The consent may be revoked at any time in writing or orally.”

Ethical as this may sound, I think that the preoccupation with any and all genetic testing makes e.g. applying cancer diagnostics based on genetic testing of tumor tissue overly burdensome. The consultation regime and resulting informed consent associated with any and all genetic testing should I think only apply to specific hereditary conditions which, admittedly, can have an enormous impact on patients’ lives and people should be counselled on that. However, I don’t see why the EU should intervene in this because it essentially is a matter of ethics that should be reserved to member states, just like the proposal clearly states in respect of other ethically sensitive matters like a member states decision whether to restrict the use of any specific type of in-vitro diagnostic device in relation to aspects that are not covered by this Regulation. In my view the rapporteur is not consistent here.

In addition, what happens when the consent is revoked after the test? Does that invalidate the test results? Can’t they be used anymore? What if medical treatment has been initiated based on these results? This invokes the all the problems that the General Data Protection Regulation proposal is now causing with the right to be forgotten in clinical trials.

Near patient testing

The conformity assessment route for near patient testing devices changes: design examination for near patient testing devices only for near patient testing devices in class C.


You can disagree with some of the solutions chosen, such as covering tests with indirect consequences to people’s health (nutrigetic and lifestyle tests) which I think fall outside the scope of the concept of medical device (based on the EU Court’s recent Brain Products case: if there is no direct medical intended purpose, even if there may be a relation to health, there is no medical device) and consequently cannot be IVDs because IVDs are a species of medical devices.

HOWEVER, the scope change to implement this seems to overreach and open some floodgates by amending the definition of medical device as such (apparently to get around the EU Court’s case law) by introducing the “indirect” medical purpose:

medical device’ means any instrument, apparatus, appliance, software, implant, reagent, material or other article, intended by the manufacturer to be used, alone or in combination, for human beings for one or more of the specific direct or indirect medical purposes of […]

– providing information concerning direct or indirect impacts on health”

That sounds like regulating things that are way remote from what is medical: a device with an indirect medical purpose providing information concerning indirect impact on health. That could be my intelligent fridge that tracks my food intake (too much soda drinks in the fridge ->  bad for health) or my Garmin Edge cycling computer that measures heart performance, calories burned and watts of power generated (generate few watts at high heart rate -> must be a medical issue). Again, this is at odds with the EU Court’s Brain Product judgment that says that the EU should only regulate devices with a directly intended medical purpose. I guess you can get around that by moving the goal posts. The question is if you should, because a lot of companies in sporting and general health products will suddenly find themselves producing medical devices. Just imagine what the accessory cloud around this could around these medical device would look like with the proposed amended definition of any device that “assists” a medical device in its intended purpose. Explosive proliferation of medical devices regulation – as devices lawyer I say: go for it!

In addition the amendment of the definition of manufacturer will also ripple through in the medical devices regulation. The rapporteur proposes an impressively amended concept of manufacturer, which includes also includes packaging, labeling and assembly, while the text is unclear about whether the packaging, labeling and assembly activities must be conducted with a view the devices concerned being placed on the market under that person’s own name (it is not clear if “with a view to their being placed on the market under that person’s own name or trademark” applies both to both sets of activities mentioned or only the latter):

“the natural or legal person with responsibility for the design, manufacture, packaging and labelling of a device before it is placed on the market under that person’s own name, regardless of whether those operations are carried out by that person or on that person’s behalf by a third party. The obligations of this Regulation to be met by manufactures also apply to natural or legal persons who assemble, package, process, fully refurbish or label one or more ready-made products and/or assign to them their intended purpose as devices with a view to their being placed on the market under that person’s own name or trademark.”

That definition clashes however blatantly with the ‘new’ New Approach system set up on the template of Decision 768/2008 that defines a general set of definitions and procedures to be used in CE mark regulation, including that of  ‘manufacturer’. If you want the clarify the concept, I think a mechanism like in the medical devices proposal of activities that are also subject to control as if you are a manufacturer (translation and repacking in that case) would be a much better option. Anyway, there is an opportunity to improve the drafting here, and the manufacturer concept should of course be as consistent with other CE statutes (last but not least the medical devices regulation!) as possible.


The proposal provides for the possibility to conduct inspections at interventional clinical performance studies sites. Normally national health supervision authorities will have these powers, but the regulation will fill in any gaps because they have the possibility and will have to monitor such studies. The rapporteur clarifies that this way member states have no excuse to let this slide.

Beefing up of interventional clinical performance study requirements

These are supplemented with a specific regime for incapacited and for minor test subjects – the most vulnerable study subjects.

Transitional Period

The rapporteur does not beat around the bush and wants a shorter transitional period, 2 years instead of 5 years.

All in all

The proposed amendments generally make sense, except for the medical devices definition amendment and the genetic testing counseling. Both of these look nice, but will resort in a lot of practical problems because of their lack of preciseness.

Update 18 April 11:30 re transitional period

EDMA pointed out to me that the limitation of the transitional period to 2 years does not apply to all aspects of the proposed regulation, but only to

“the requirement for member states to cooperate with one another and a faster implementation of the new vigilance provisions (in line with the implementation of the new UDI provisions). Compliance to the new classification and conformity assessments linked to them remain at 5 years as well as many of the related provisions.”

That does make a quite a difference, so many thanks to EDMA for that observation.