The unfinished internal market for medical devices
The EU market for medical devices is not fully harmonized, which means that you cannot rely on a medical device being a medical device everywhere in the EU. That, essentially, is what the European Court’s recent judgment in the Lycocentre case confirms. Confirms? Yes, it basically confirms what we already knew and that is that each EU member state can second-guess the regulatory qualification of your product to suit it’s own needs. In my own legal practice I’ve found that this happens quite often.
What was the case about again?
It was about ‘Gynocaps’, a vaginal capsule containing live lactobacilli intended to restore balance to the vagina’s normal protective bacterial flora, which was marketed in Finland as a ‘medical device or accessory’, bearing a CE marking, until 2008. To paraphrase the European Court’s case summary:
‘The capsule is currently also marketed as a ‘medical device or accessory’ bearing a CE marking in a number of other Member States, including the Kingdom of Spain, the French Republic, the Italian Republic and the Republic of Austria. The European Medicines Agency (EMA) has not specifically adopted a position on the classification of vaginal preparations which, like Gynocaps, contain live lactobacilli. It has, however, taken the view that on the basis of its intended use and effects, a gynaecological tampon containing live lactobacilli satisfied the conditions for classification as a ‘medicinal product for human use’, within the meaning of Directive 2001/83. The Finnish authorities decided that the product was not a medical device but a medicinal product and therefore required marketing authorisation as a medicinal product. The Finnish authorities, therefore, decided to classify Gynocaps as a medicinal product within the meaning of Directive 2001/83 and notified the decision to the Commission because they interpreted Directive 93/42 as meaning that where a CE marking has been wrongly affixed, the safeguard clause procedure provided for in Article 8 of that directive was not applicable because the issue was not strictly one of non-compliance. Lycocentre appealed the decision but the Finnish court dismissed the action holding, inter alia, that according to the case-law of the Court of Justice, the fact that a product is classified, for example, as a foodstuff in one Member State does not preclude that same product from being classified in another Member State as a medicinal product. According to the Helsingin hallinto-oikeus, the Lääkelaitos was entitled to classify Gynocaps in Finland as a medicinal product, even though that preparation is marketed as a medical device in a number of other Member States of the Union.’
On appeal of that judgment the appellate court decided to stay the proceedings and to refer the following questions to the Court for a preliminary ruling:
- Does a definition given in one Member State in accordance with Directive 93/42, by which a product is regarded as a medical device or accessory in accordance with that directive and is provided with a CE marking, preclude the competent national authority of another Member State from defining the product concerned, on the basis of its pharmacological, immunological or metabolic effects, as a medicinal product in accordance with Article 1(2)(b) of Directive [2001/83]? – in other words: ‘can each member state qualify a ‘medical’ product differently?’
- If the answer to the previous question is in the negative, can that competent national authority define the product as a medicinal product observing only the procedures under Directive [2001/83] or is it necessary, prior to initiating procedures under that directive to define the product as a medicinal product, to follow the safeguard procedure in Article 8 of Directive [93/42] or to comply with the provisions of Article 18 concerning an unduly affixed CE marking? – in other words: ‘can each member state just do what they want or do they need to follow the safeguard and/or unduly affixed CE marking procedures in the MDD?’
- Does Directive [2001/83], Directive [93/42] or other European Union legislation (including legislation concerning the protection of human life and health and consumer protection) preclude products containing the same substance and having the same modes of action from being marketed in the same Member State both as medicinal products in accordance with Directive [2001/83], requiring a marketing authorisation, and as medical devices or accessories in accordance with Directive [93/42]?’ – in other words: ‘what happens when these differently qualified but identical/similar products start circulating in the internal market?’
Question 1: ‘Science can be different anywhere’
As to the question whether member states are allowed to qualify the same product differently, the Court first reiterates its case law on how such decisions must be completely evidence based and tailored to the specific product, but then goes on to hold that
“45 None the less, as Union law currently stands, until harmonisation of the measures necessary to ensure the protection of health is more complete, it will be difficult to avoid the existence of differences in the classification of products as between Member States in the context of Directive 2001/83 (see, to that effect, inter alia, Case C‑201/96 LTM  ECR I‑6147, paragraph 24, and Hecht-Pharma, paragraph 28).
46 As the Advocate General has stated in point 63 of her Opinion, asymmetries in scientific information, new scientific developments and differing assessments of risks to human health and the desired level of protection can explain why different decisions are taken by the competent authorities of two Member States as regards the classification of a product.
47 In addition, the fact that a product is classified as a medical device in accordance with Directive 93/42 in one Member State does not prevent it being classified, in another Member State, as a medicinal product in accordance with Directive 2001/83 if it displays the characteristics of such a product (see, by analogy, Case C‑150/00 Commission v Austria  ECR I‑3887, paragraph 60, and HLH Warenvertrieb and Orthica, paragraph 56).”
In other words: there is no central authority to qualify medical devices, science can differ and authorities can have different policies as to what they consider application of science to products for regulatory purposes and they are allowed to differ in opinion.
This case puts a very precise finger on the sore spot of the role of science in the qualification of medical products and the EU’s paradoxical way of dealing with it in the light of division of competence between the EU and the member states. On the one hand we all want the best state of art clinical substantiation for medical devices to be fully scientific and we assume that science leads to the same conclusions everywhere. Actually, that is what the EU market access mechanisms for medical devices and medicinal products are based on. On the other hand the Court, following it’s AG, is ready to believe that for the purpose of regulatory qualification science can be different anywhere in the EU, just because member states have not fully surrendered their competence to each qualify a medical product for themselves. In other words, just because we don’t have a harmonized system in place member states can do whatever they like with different scientific views as an excuse. In a discussion I recently had with another expert we agreed that if you take this to its logic endpoint, gravity could go up in one member state and down in another and that’s OK for the Court because the law does not say it should pull towards the centre of the earth everywhere and member states are not willing to agree to that when it does not suit their purposes. That, I think, is a very bad way to deal with products that we would like to be as much scientific evidence based as possible – lives are at stake, remember? That is basically impossible if you can’t rely on underlying science being universal and member states to be willing to give it their very best scientific effort to conform to scientific state of art. To me, this invokes impressions of a world of parallel universes existing in the minds of different persons like in the cold war science fiction novel The Eye in the Sky of Philip K. Dick, in which each universe has its own laws of nature, leading to some very absurd results. We are in the exact same situation in medical devices borderline products. However, I think in today’s society we ought to be be far beyond discussions about whether the world is flat or round. Because if the member states can differ and the European Court will not decide on facts unless in direct actions (which do not really exist under the current medical devices directives), there is no way to resolve a debate on the mode of action of a particular product, which brings you back to a patchwork of different policy coloured opions about what science dictates.
In case of doubt they probably get it wrong
The Court makes a short sidestep to refer to the hierarchy clause in article 2 of Directive 2001/83 in discussing the relationship between medicinal products and medical devices regulation in the EU. This is a tool that member states often use to bail themselves out of a difficult qualification situation. In the word of the Court:
“Article 2(2) of Directive 2001/83 requires the application of Directive 2001/83, by providing that, where, taking into account all its characteristics, a product may fall within the definition of a ‘medicinal product’ and within the definition of a product covered by other Union legislation, it is the provisions of Directive 2001/83 on medicinal products that are to apply.”
What happens in practice? Case handlers at a competent authority will take issue with the regulatory qualification of a particular product as medical device, claim that there is “doubt” about the mode of action and therefore qualify the product as a medicinal product with reference to the hierarchy clause, without properly looking at all characteristics and without positively establishing that the product does actually fall within the scope of both Directive 2011/83 AND 93/42. This goes wrong more often than not, not only in my own experience but also in the experience of many others.
Question 2: Fortunately, there are procedures for this in the EU MDD
With regard to the procedural issue (can the national authorities act on their own, or should they apply the article 8 and/or 18 procedures in the MDD?), the Court finds that
“in order to classify as a medicinal product in accordance with Directive 2001/83 a product already classified in another Member State as a medical device bearing a CE marking in accordance with Directive 93/42, the competent authorities of a Member State must, before applying the classification procedure under Directive 2001/83, apply the procedure under Article 18 of Directive 93/42 and, where appropriate, the procedure under Article 8 of Directive 93/42.”
Good news, there are procedures to follow, and these procedures make the member states talk to each other, sort things out and then the outcome is binding for the whole EU market, right? Wrong. First, the outcome is not binding on the member states. Secondly, the process itself and the outcome of the process (the Helsinki procedure) are neither public nor transparent. And the manufacturer has no legal recourse against the qualification if the product, because (a) it is just a consensus reached between the member states and (b) the member states have agreed to make all the qualifications in the Manual on Borderline and Classification to relate to an ‘abstract’ situation, which however is a direct reference to the actual product that was assessed. So, while the Court refers to the MDD procedures, this will not help you a single bit if you are a manufacturer. Member states do as they like and if you don’t agree with the outcome of what they agree on a non-transparent and non-binding basis you can have fun litigating against the authorities in each and every member states that decides to make your qualification an issue. This is not only bad for individual companies, but also for patients relying on the devices concerned (because you often have to take them off the market to reinvent and apply for them as whatever other thing (usually medicinal product) the member states would like you to place it on the market). Even the Commission concluded in its Staff Working Paper that underlies the new medical devices regulation proposal that the current process is far from perfect:
“However, some controversial cases remain unsolved despite long discussions within the above mentioned Expert Group. The application of different regulatory regimes to the same product compromises both the protection of patient safety and the internal market. […] Moreover, even where a consensus is found, the consensus statements published in the Manual are not legally binding and competent authorities or national courts may decide at any moment not to follow them. This reduces the legal certainty and prompts criticism from stakeholders. […] The lack of uniform qualification (or classification) of a product across the EU creates a fragmentation of the internal market (as a manufacturer must follow different legal regimes in order to sell the same product in different Member States) and may put patient safety at risk.” (p. 18/19)
Yes, that’s right – and this politically correct language is just what the Commission is willing to put in writing publicly – there will be a world of frustration behind this. The Commission points to a lot of very well-deserved criticism about this process, which, as the Court puts it in the Lycocentre judgment, is the result of incomplete harmonisation “of the measures necessary to ensure the protection of health” (§ 45). The lack of harmonization is of course no reason for member states not to work together better and at least make sure they base their decisions solidly on the same scientific evidence, except that in practice it is.
Question 3: But what about similar products with a different qualification coming from other member states?
That’s a very reasonable question. If there is free movement, you would expect the products coming from other member states to be able to carry their qualification with them, because that would be what the MDD’s internal market clause of
“Member States shall not create any obstacle to the placing on the market or the putting into service within their territory of devices bearing the CE marking” (article 4 (1) MDD)
would seem to mean. No, says the Court: if
“within the same Member State, a product, which, while not identical to another product classified as a medicinal product, none the less has in common with it an identical substance and the same mode of action, cannot, in principle, be marketed as a medical device in accordance with Directive 93/42, unless as a result of another characteristic that is specific to that product and relevant for the purposes of Article 1(2)(a) of Directive 93/42, it must be classified and marketed as a medical device, which is a matter for the referring court to verify.”
In other words, we’re back to a situation as in HLH Wahrenvertrieb, which was about member states looking at active substances as such as a regulatory triage mechanism. Even though the Court is – I think – not coming reversing this case law, you’re still 1-0 behind as company if the member state can say: look, there’s an identical substance in there with the same mode of action as something that I have already considered a medicinal product before, good luck to you proving that “as a result of another characteristic that is specific to that product and relevant for the purposes of Article 1(2)(a) of Directive 93/42, it must be classified and marketed as a medical device”. The result of this judgment is that effectively the HLH Wahrenvertrieb burden of proof is shifted from the member state on to the manufacturer, because the Court says it’s OK not to look at the product individually in detail if another member state (on whatever grounds) has decided that the device is a medicinal product, and then the manufacturer can prove it is not a medicinal product. This is a step back in the protection of manufacturers against competent authorities that do not substantiate their arguments very well scientifically.
Looking forward: what will this look like under the new regulation?
I’ve complained a lot on this blog about how the new regulation will not be an improvement in a number respects. In the respect of borderline products it will be a huge improvement, because we will finally have a possibility for a central qualification that is binding on all member states in article 3 of the regulation. This article allows the Commission to do two things
- issue a qualification for a product or a group of products by delegated act, by its own initiative or after a request by a member state, and
- ensure that the science that can be different everywhere is less different as a result of sharing of expertise (which is basically what scientific development is about in the first place if you ask me)
The article looks like this in the Commission’s proposal:
“(1) The Commission may, at the request of a Member State or on its own initiative, by means of implementing acts, determine whether or not a specific product, or category or group of products, falls within the definitions of ‘medical device’ or ‘accessory to a medical device’. Those implementing acts shall be adopted in accordance with the examination procedure referred to in Article 88(3).
(2) The Commission shall ensure the sharing of expertise between Member States in the fields of medical devices, in vitro diagnostic medical devices, medicinal products, human tissues and cells, cosmetics, biocides, food and, if necessary, other products in order to determine the appropriate regulatory status of a product, or category or group of products.”
The qualification is by implementing act, which means that you most likely have no legal recourse against it as manufacturer, unless perhaps if you can meet the very high threshold of proof for being directly and individually concerned. Whether that is the case will depend on what the implementing act concerned looks like.
There were some minor tweaks by the Parliament, which proposed it should look like this (see picture above). No radical changes, except that (1) in case of a request of a member state the Commission must determine with an implementing act what the status of the product concerned is, (2) the MDCG and the MDAC get to weigh in with opinions and (3) the text makes it explicit that also borderline products can fall within it scope (that’s what the clause is for in the first place).
So, the internal market is not finished – but it may be less unfinished in the future
That’s what the new regulation has in store. The new article 3 will help, but the situation of member states taking different views to the detriment of patients and manufacturers will remain. The progress is that we will (hopefully) have a better way of overruling member states in individual cases, although the Commission will not be obliged to act on any requests by anyone except a member state.