drb2Well, the long awaited DRB (Dagmar Roth-Behrendt) report is here. I read through it as quickly as I could, answered a bunch of press questions and can now give you my preliminary view. I have poked fun at the process before with the ostrich picture but having read the proposal I am feeling more cynical than funny by now and did my share of facepalming today. You’ll understand by the time you’ve read this article. Here are my first thoughts on the rapporteur’s report.


Let’s start with PMA. It was announced. We feared it would be bad. And it is. Not even because of the   procedure as such but because of the underlying assumption that kills risk based regulation, the cornerstone of EU medical devices law. Why? The proposal divides the world in three groups: “innovative” devices, non-“innovative” but high risk devices and the rest. The first group is subject to EMA centralised authorisation, the second one to decentralised member state authorisation and the third subject to the ‘normal’ regime. There is some detail on what this would cover:

Centralised procedure:

– innovative implantable devices, – innovative devices referred to in Article 1(4) (combination products),

– innovative devices referred to in Article 1(5) and point 5.3. of Annex VII (Rule 11) (drug delivery devices), or

– innovative devices manufactured utilising tissues or cells of human or animal origin, or their derivatives, which are non-viable or are rendered nonviable.

Decentralised procedure:

– class III devices,

– non-innovative implantable devices,

– non-innovative devices referred to in Article 1(4) (combination products),

– non-innovative devices referred to in Article 1(5) and point 5.3. of Annex VII (Rule 11) (drug delivery devices), or

– non-innovative devices manufactured utilising tissues or cells of human or animal origin, or their derivatives, which are non-viable or are rendered nonviable.

By now you are probably wondering what this magic word “innovative” means. Well, the proposal actually does not tell us is what “innovative” is. Yes, really. It is left to the European Commission to define the products that should be classified as such by delegated act. So, while we have evolved IVD regulation from the cumbersome list based classification because that just did not work, we are inviting this right back in for the highest risk medical devices. And mind you, a delegated act procedure is not a light instrument, it’s political and requires both Council and Parliament input. The proposed text breathes what I have argued in relation to the scrutiny procedure: an irrational and non-evidence based fear of the innovative, just to appease the public opinion rather than educate the public. The procedures proposed are not surprising for PMA: they mirror the EMA medicinal products procedure for ‘innovative’ devices and the mutual recognition procedure for non-innovative but high risk devices. You will love the token surprise inspection during the application process. You will especially like the not so well thought through transitional regime:

“Devices referred to in paragraphs 1 and 2, and which are already on the Union market at the date of entry into force of this Regulation, shall be required to have a marketing authorisation, in accordance with the procedures set out in this Section, as from the expiry date of the validity of their certificate.”

Consequently, if your certificate expires after the regulation enters into forces at a date that allows less than the minimum time possible to get a marketing authorization after that mechanism entered into force: take your product off the market. This will not work of course.


A lot happens in the clinical field, like the mandatory introduction of Ethical Boards at member state level for approval of studies. You love it or you hate it, but at least it’s harmonisation. What is not so nice is the addition on clinical investigation methodology:

“As randomized controlled investigations usually generate a higher level of evidence for clinical efficacy and safety, the use of any other design or study has to be justified. Also the choice of the control intervention shall be justified. Both justifications shall be provided by independent experts with the necessary qualifications and expertise.”

This means that you will always have to argue in your study design not only why a particular design is appropriate, but now also why you are not doing a study that may be completely inappropriate for the device concerned (actually randomized studies are inappropriate for the majority of devices) – so you would have to explain why you are not using placebo implants, for example.

Also, post market clinical follow up plans that you already agreed with the notified body will be second guessed by third parties for class III devices (“[technical file contains] the PMCF plan and PMCF evaluation report, including a review of the PMCF evaluation report by an independent scientific body for class III medical devices“) (underlining added).

Another interesting one is that incident reports will need to contain information on the healthcare professional and patient involved if available to the manufacturer – and they will become accessible to healthcare practitioners and the public (although in the latter case the access level is not specified).

Systematic registries will be required for class IIb and III devices, where the Commission proposal only encouraged them.

Notified Bodies

As was to be expected after the 26 February ‘workshop‘ the notified bodies were not going to get it easier in the proposal. As I have flagged, what are the consequences of a notified body not adhering to an MDCG recommendation? Well, now we know: the member state responsible will have to justify this. At least this is a way to make member states responsible for what their notified bodies do, but I doubt if this is the right way because it assumes that the MDCG is always right while there is nothing in place to ensure that the MDCG always has more expertise than the notified body.

“As a consequence of the internal market, manufacturers are allowed to apply with a notified body established in another Member State than the one where the manufacturer is registered. However, in the view of improving transparency, if a manufacturer chooses to do so, it should inform the national authority of the Member State where it is registered of such an application.”, argues the rapporteur.

Making it more difficult for manufacturers to use notified bodies in other member states will likely raise eyebrows with the legal service of the Commission because it hinders the free movement of services, in my view unjustifiably. Indeed, I don’t see why member states could not receive this information from the Eudamed database. The idea of increasing transparency of fees charged and ensuring they’re comparable across members states sounds good on the surface, but the prospect of enacting national legislation reflects a desire to have government regulation on prices of notified bodies. On the one hand the proposal increases their overhead considerably and on the other provides for mechanisms to compete on price rather than quality. That’s not logical and won’t achieve the intended result. Companies can already compare costs because the mandatory services provided are the same. The proposal implements more control on notified bodies using subcontractors. I think that is a good thing (I have had some bad experiences with outsourced auditors in cases for clients), but this is something the notified bodies are fixing already themselves in their code. Putting it in the regulation will take care of free riders that do not want to subscribe to the code.


This part of the proposal seems to have been completely written by the reprocessing lobby. Disclosure: I am normally on the OEM side, so you would expect me to say something like this. However, if I were that industry I would be worried about such a stunning success in getting very controversial points across and written down as proposal. The report proposes to introduce the assumption that EVERY device (yes – “every” as in “none excluded”) is capable of being re-used but that economic considerations prevent manufacturers from admitting this. Here is the rapporteur writing:

“Reuse of medical devices was very common until the 1980s, when manufacturers started more systematically to label their devices as single-use. The current situation is that there are too many devices labelled as single-use while they could be reprocessed, as manufacturers want to avoid bearing the responsibility in case the reprocessing of a device would pose a danger to a patient. Sometimes, improper labelling is the result of economic considerations.”

Let’s apply this reasoning to medical practice: opening up people to take a look inside has served us well for centuries for diagnostic purposes, now we are using MRI and CT scanners because these naughty manufacturers just want to make a profit. So, let’s get rid of these scanners because they just add to costs.

The rapporteur totally ignores that there might be a reason that has to do with, let’s say usability and risk/performance ratio, that devices get built for single use. Catheters that have to be capable of being reprocessed will be built sturdier and be less friendly to the patient’s vascular system, just to mention an argument that even lawyers understand. This is so absurd that a Star Trek quote is in order (doctor McCoy  observes a primitive treatment method in Star Trek IV): “My God, man! Drilling holes in his head isn’t the answer! Now put away your butcher knives and let me save this patient before it’s too late!”

The burden of proof is on the manufacturer to prove that a device cannot be reprocessed. The manufacturer has to prove this, and in case of class III devices, even commission a SCENHIR opinion to back this up. Essentially the reprocessing industry gets all clinical substantiation for free from the OEMs, who have to provide clinical support for activities that they do not intend for their products. Sounds very logical, right? The problems are compounded because the whole reprocessing regime will apply to everything in the proposed definition of reprocessing (which was not amended): “the process carried out on a used device in order to allow its safe reuse including cleaning, disinfection, sterilisation and related procedures, as well as testing and restoration of the technical and functional safety of the used device”. Consequently, the fairly balanced regime for reprocessing from the Commission’s proposal is blown up to include all current and normal activities to prepare intended re-usable devices for intended re-use. I know that I will be buying my own single use devices for my family and will be taking them to the hospital if this proposal clears.

Isn’t there anything good in there?

Actually, there is. The report proposes to involve healthcare practitioners more in the incident reporting process, one of the points on which I think the Commission proposal clearly fell short.

Now what?

This is just a preliminary view, but the big picture emerges: largely token measures that will not solve the problem they purport to solve. The rapporteur’s report will be the basis for the ENVI committee’s vote, which in turn is the basis for the vote by the parliament. There are however also other parliamentary committees involved, such as IMCO (internal market), which issued a report that includes no PMA at all. Further, the member states have to agree for the proposal to become law. So far (as far as I know), only France has been lobbying for PMA. Industry can still lobby and they certainly will. I think this report proposes a regulation that nobody should want because it is especially onerous for the SMEs in the higher risk medical devices. It is prone to create a sub-optimal regulation as we have with the ATMP regulation, which has very little to show for it in terms of authorised products and includes a lot more SME stimulation than the devices regulation proposal that actually includes no SME incentives. EMA is now canvassing biotech SMEs for some time already to please please take a chance with the process. Is that what we want for devices?

Update 16 April 9:49

You know that feeling when you write something, send it off in a hurry and then look at it the next day? I had that feeling this morning, so here are some updates (apart from the spelling mistakes I already corrected in the text above – apologies for those).

Here is the IMCO committee report link (it’s not up on OEIL yet, so you’ll have to get it from my Dropbox for the moment, I hope your security allows it). You’ll see that IMCO’s report stays closer to the Commission draft, but makes the MDCG scrutiny opinion binding on the notified body, beefs up the clinical expertise behind the MDGC (if you have to go MDCG, this is a sensible thing) and provides for a scientific advice procedure for manufacturers that have scrutiny liable devices (hopefully good for SMEs, works like the pharma system in EU).

The notified bodies are not very happy  at all – they have a closed huddle today to see how they should deal with this.

As @robertmdproject pointed out to me: don’t forget that the proposal seeks to redefine some core technical and clinical concepts in medical devices regulation:

  • “performance”: “any technical characteristics, any effects and any benefit of the device when used for the intended purpose and in accordance with the instructions of use;”
  • “benefit”: “the positive health impact of a medical device based on clinical and non-clinical data”; and
  • “safety”: “the avoidance of risk or harm caused by the medical device or associated with its use”

You can imagine the impact this will have on for example the way we will have to interpret harmonised standards like EN 14155, the medical devices GCP standard that contains a definition of “clinical performance’ and risk management standards like EN 14971 (contains a definition of safety different from the one proposed (“freedom from unacceptable risk”) and describes risk/benefit analysis in great detail). How are we going to work with definitions that conflict with harmonised standards? I honestly have no idea.

The IVD regulation proposal by rapporteur Peter Liese is not out yet. It is expected in the coming days and I’ll report on that when is comes out. But you can expect all the procedural stuff to be identical or similar, like PMA for class D IVDs. The IMCO report re the IVD regulation proposal proposed to clarify the regulation of in-house (home brew) exemption/regulation and include predictive genetic tests. I expect this to find its way to Peter Liese’s report too, given the preoccupation with genetic testing we have seen.