Europeans are left to their own devices

I read the group of articles on the regulation of medical devices with great interest. From the point of view of someone who cares about patient safety and practising evidence-based medicine, both US and EU regulatory regimes have obvious flaws. The EU system of approval by agreement between manufacturer and a commercial regulatory body, under conditions of almost total commercial secrecy, overseen in a "hands off" manner by national regulatory authorities, seems extraordinary, in its lack of transparency and the obvious inherent risks of collusion. The fact that the entire system is based on device safety and functionality, with no efficacy requirement is also very surprising when it is applied to invasive treatment technology with significant inherent risk potential.

Whilst it is true that the EU system's faults appear more egregious, it is misleading to hold up the FDA as a shining example. The fact that 90% of devices which obtain FDA approval do so through the 510k process is recognised by the FDA themselves as a major problem, and one that they are actively trying to address. Equipment passed via 510k is essentially subjected to the same type of regulatory questions as those posed by EU regulation, so Dr Shuren's remarks about the superiority of US regulation only apply to 10% of the US market. The MAUDE database described by Dr Shuren in Cohen and Billingsley's article as keeping the FDA in check is essentially a voluntary registry, in which the manufacturer gets to decide whether an adverse event was related to their device (and should therefore be logged) or not : MAUDE's weaknesses were in fact recently enumerated very recently in the BMJ(1). Cohen et al are quite right to bemoan the impossibility of independent evaluation of device failure rates and harms in Europe, but the fact is that the situation in terms of our ability to understand what is really happening is not a great deal better in the US. Both regimes need to be updated to deal with the reality of a new world in which there are now myriad devices whose invasiveness, capabilities and potential for harm are hugely greater than 20 years ago. In developing rules for evaluating this type of innovation, however, the natural history of the development process needs to be taken into account. In this regard the IDEAL framework(2) may prove a more useful template than an approach based on the pharmaceutical model, since devices critically require a "tinkering" stage during their early development (IDEAL stage 2a) and their performance is often affected by operator learning curves (IDEAL stage 2b).

Mario and colleagues' response, warning against more regulation, seems muddled, and based on a number of strange assumptions. After outlining in their first paragraphs just why power in this area appears to belong to manufacturers, they conclude that it is in the hands of administrative bodies. They then erect a "straw man" to demolish, by postulating a regulatory requirement for an immediate RCT of coronary angioplasty versus surgery. An appropriate paradigm such as IDEAL would not call for this, but for a logical progression of studies, addressing different questions using different methods as the innovation process develops. In this regard I would disagree with Freemantle's otherwise very sensible article. Certainly Class 3 devices should be subjected to rigorous evaluation, but an RCT too early in the development process is likely to fail or produce contentious results which are difficult to interpret, for well recognised reasons(3). Later in their article, Mario et al suggest that patients will be protected by clinicians, who can be relied on to make correct decisions about the morass of evidence they are presented with in the ultra-competitive field they choose as their example. This seems over-optimistic.

Much of the discussion in all of the articles is based on a false dichotomy between approval and non-approval, whereas in many situations, the appropriate course to ensure the accumulation of good evidence, minimise risk and allow innovation, would appear to be approval of use under strictly defined conditions of continuing evaluation in well- designed prospective studies. Rare and long-term harms cannot be reliably detected by such early phase studies - nor indeed by randomised trials. What is needed is a framework of ongoing evaluation alongside innovation, whereby appropriate methods are used at each stage in the product life- cycle up to and including that of long term surveillance. This could, with great benefit, be developed on an international basis, as Fraser et al suggest, but much thought will be required about how to develop the registries and ethical frameworks needed to allow essential early phase innovative work to continue without compromising patient safety.

1. Lenzer J, Brownlee S. Why the FDA can't protect the public. BMJ. 2010 Nov 2;341:c4753. doi: 10.1136/bmj.c4753.

2. McCulloch P, Altman DG, Campbell WB, Flum DR, Glasziou P, Marshall JC, Nicholl J for the Balliol Collaboration. No surgical innovation without evaluation: the IDEAL recommendations. Lancet. 2009 Sep 26;374(9695):1105-12.

3. Ergina PL, Cook JA, Blazeby JM, Boutron I, Clavien PA, Reeves BC, Seiler CM for the Balliol Collaboration. Challenges in evaluating surgical innovation. Lancet. 2009 Sep 26;374(9695):1097-104.

Competing interests: None declared