Towards an evidence based medical device introduction?

Medical devices, and particularly their introduction into the healthcare system are under scrutiny. Recent adverse experiences with devices such as the metal-metal hip implant and the PIP implant have accelerated the public awareness that there is insufficient safeguard against possible side effects of innovative medical devices. After a national broadcast of a consumer program revealing the ease to acquire a CE mark, the Dutch ministry of health was triggered to inform the parliament about new regulations to prevent future disasters.

The Dutch implant register involves linking three major registries (Orthopedic, Cardiovascular, Breast), which is a good start, but not quite inclusive. Still, the implant register is a good step towards post-marketing surveillance, especially for safety. The proposal also involves the right to know of the patient about the received implant. The side downside being that this limits blinded studies, something which is already a great lack in medical device research.

The main problem however is the introduction into the market of new devices, and the clinical evidence supporting their effectiveness and safety. Researching medical devices is very different than for pharmaceutical interventions. Complicating aspects are long observation tmes (>10 years), methodological issues regarding randomisation and blinding, and the technology driven innovation cycle. That said, a lot can be learned from the pharmaceutical standards in being more reserved for new interventions / substances without clinical evidence.

The European efforts are focused on improving the notified body system. While the system has been shown not to work properly, the fundamental issue will not be solved by just auditing the notified bodies. A comprehensive medical device regulation should not be based on safety solely, but in balance with clinical effectiveness.

Improvements in the updated EU medical device regulation include the recognition of the need for clinical data, established in controlled investgations in the target population. While the possibility to use observational data is acknowledged, no further requirements are made as to the quality of these studies. Only large observational studies with sufficient safeguard against confounding and bias can be used as a valid alternative to randomized studies. A major drawback has been the ease with which devices that resemble existing devices receive a CE mark and are allowed to the market.

The succes of the new directive is dependent on how carefull the notified bodies will evaluate each and every new device and require clinical evidence.a

Onbepaald