Any attempt to wrap the FDA vs CE Mark debate up into one short article is ambitious and likely to provide a superficial treatment of its subject. We’ll probably prove that here. Bloomberg Week and other organs have similarly embarked on the task in articles which includes all the issues and all the players, mixing them all up, blurring the lines between regulatory reform differences across the pond, device-related issues and downright criminality in order to create a tale of global angst about the world of medical device regulation. In fairness to the authors, they’re not alone in joining the party. Self-interested commentators from the FDA themselves to umpteen lobbyists have weighed in, leaving the man in the street throwing his hands in the air and believing that clinicians are using Christmas Cracker gifts to take out their appendices.
Let’s get in the regulatory helicopter
In our view we need to separate the issues before we can move to a progressive new regime, if that’s what’s required. Just bear in mind however that, to draw a financial parallel, the complex US Sarbanes Oxley regulations which were introduced to prevent another Enron, are widely judged not to have been adequate to do just that. Let’s not make the same mistake again and leave ourselves with a burdensome regime which is no better suited for purpose than today’s, but mightily more burdensome.
It is highly likely that the scandal over “dodgy” breast implants made in France may make it harder for medical device makers to argue their case for US reforms to permit faster approval for high-risk products, like their European counterparts. US Politicians and “nay sayers” will obviously make hay with this. In Europe the “reformists” will similarly claim PIP to demonstrate a fundamental problem with the Medical Devices Directive(MDD) and its implementation and then tar all medical devices and occasional high profile problems with the same brush.
But we’re not so sure of the connection, and even less convinced by the reference to device problems in general and the link with regulatory process. It’s unhelpful to pick examples like DePuy’s ASR all-metal hip, because that was approved both sides of the pond. Surely better to look at what we like from FDA land and what we don’t, what we like from CE land and what we don’t, and come up with sensible, pragmatic regulatory regime change which ensures as far as reasonably possible that patients benefit from the best technology for the job. This doesn’t necessarily mean ultra-tough rules, because such a regime may actually prevent new and brilliant technologies coming to market. But it should be an opportunity to iron out the inadequacies of the current rules.
So what’s in and what’s out?
Spot checks: These will keep ’em on their toes
Back to breast implants for a minute, this scandal doesn’t really prove that EU regulations are “weaker” than FDA’s per se. PIP looks like the deliberate act of a company to deceive the regulators. It’s disingenuous to suggest that such a deception could not occur under FDA’s rules nor that PIP chose not to go into USA because it knew the rules would catch them. The factor most likely to mitigate against them getting caught in EU, but not in US is the existence of unnannounced spot checks in USA. So that measure would be on our to-do list.
Company/Notified Body/Competent Authority relationship: Ironing out conflicts of interest
More fundamental is the Company/Notified Body/Competent Authority axis. We don’t think in its current form this should work, but it mostly does. So how can we take the “mostly” and make it an “always”? Firstly, we don’t like the conflict of interest and agree with The British Medical Journal(BMJ) on this one. According to Deborah Cohen, an editor at the London-based publication; “The European system sets up “an inherent conflict of interest” where companies can shop for the most amenable reviewers”.
Admittedly it’s probably a very cost-effective way of doing things. The government doesn’t have to do all work and incur all the expense, just keep its hand on the tiller via its competent authority policeman overseeing a small army of Notified Bodies. The Notified Bodies are businesses which need to make a profit, but that needn’t mean there is a totally free market approach to this. Could the competent authority keep a tighter reign on their activity in the same way a driving examiner is himself examined? This falls down somewhat because the existence of 27 member states would mean we are relying on 27 FDAs to behave the same way… and again national self-interest could creep in.
Perhaps to address the conflict of interest NBs should only have limited allowable tenure over their clients. The argument against this is that a degree of local knowledge and cooperation is most likely to result in a quality system and product approval process which is efficient and transparent with a degree of trust.
Classification rules: Class I is under-regulated
Still more fundamental are the classification rules. Taking the lowest risk of product, Class I reusable surgical instruments as an example. Should a regulator be examining a file on every pair of scissors or should he/she be satisfied of the adequacy of the company’s quality system and technical filing capability (all of which IS auditable). The problem with the existing situation is that low cost imported instruments arrive bearing a CE mark and frequently when asked the supplier sends one or two pieces of photocopied paper which it refers to as its technical file or certificate of compliance. Someone somehwere has to conjure up an “essential requirements checklist”, but again if the photocopier’s handy, what fits one, fits many. Class I rules need a look because the self-certification route is definitely not overseen by NBs to the extent it should be.
So would we adopt the current 510(k) rules? Not in my book. 510(k) works on the principle of predicacy, so a product can be placed on the market if FDA can be convinced that it doesn’t differ too significantly from a pre-existing “predicate” device. This means a new gizmo could reach the market based on its similarity to a device designed and approved under a prehistoric and different set of rules. That doesn’t make sense. And then there’s the issue of predicate creep, a sort of “Chinese whispers” where a product is approved based on a predicate, but not being identical introduces a new feature. Another new product can be approved on the back of product 2, and so on and so on until product 10 looks nothing like product 1 and would not be accepted for 510(k) on that basis, yet is on the market. FDA is reviewing its 510(k) rules and it’s heartening to see these issues being addressed, as well as the one of “split predicacy” under which a product can nick predicate features from more than one device.
Above Class I
Up the chain to higher risk products, we’d vote for independent expert panel reviews, but accept the argument that waiting for these to happen is a major brake on the process in FDA land. We do believe their output is both pragmatic, informative for non clinical regulators, thorough and relevant.
Freedom of Information
It’s hard for medical professionals to establish basic regulatory facts about products. Transparency/availability of information is on a different level entirely in USA with no apparent associated problem. We fervently believe an obligation to publish approvals, recalls, warnings and other information would raise the game over here.
Safety vs Efficacy
“Safety versus effectiveness” is another commonly used argument to support FDA vs CE Marking. While its true that CE rules place more emphasis on safety, it’s not true to say there is no treatment of effectiveness. Quality Systems include design processes which are audited frequently to ensure that design objectives of new products are laid out, that the device is working towards fulfilment of those objectives at every stage, that external (and relevant) views are being sought and that the end product is validated against its original design goals and effectiveness. The difference lies in the extent of work required to demonstrate effectiveness; in USA this frequently requires a multicentre study even under 510(k) rules. We think representative expert panels should set the requirements because after all these are people who make clinical decisions based on clinical evidence every day, so should be best placed to dictate what evidence they need.
Homogeneity: Could FDA and EU Authorities collaborate to avoid duplication?
Sadly this is probably cloud cuckoo land, but imagine if a company only had to do one study which would be acceptable on both sides of the pond. This will likely never happen.