Page 4 of 18A.G. Fraser et al.their review. If a device needs to be removed from the market, it isthe resp<strong>on</strong>sibility of the NB to suspend its certificate.Coordinati<strong>on</strong>The Committee <strong>on</strong> <strong>Medical</strong> <strong>Devices</strong>, organized by the EC andcomposed of representatives of the Member States, has specificregulatory powers for the device directives. Policy is coordinatedby the <strong>Medical</strong> <strong>Devices</strong> Experts Group (MDEG) which is said to‘encompass all stakeholders’. 17 Its members include representativesfrom trade federati<strong>on</strong>s, CEN, CENELEC, NBs, and patients’organizati<strong>on</strong>s. There are currently no members from Europeanmedical associati<strong>on</strong>s, although the EC states that they have previouslybeen invited.The Notified Bodies group (NB-MED) and the Notified BodiesOperati<strong>on</strong> Group (NBOG) were established by the EC to improvecooperati<strong>on</strong> and performance of the NBs with the competentauthorities in the medical devices sector.The European Databank <strong>on</strong> <strong>Medical</strong> <strong>Devices</strong> (Eudamed) wascreated in 1998 to allow Member States to strengthen PMS byexchanging informati<strong>on</strong> about adverse events related to the useof medical devices. Reporting will become mandatory from May2011. 18 Competent authorities will have rapid access to the databasewhich will include names of registered manufacturers, datarelating to certificates, and reports from vigilance procedureswith the findings of clinical investigati<strong>on</strong>s. The European Databank<strong>on</strong> <strong>Medical</strong> <strong>Devices</strong> will implement the Global <strong>Medical</strong> DeviceNomenclature (GMDN) code; there is also a GMDN Agency,established under the auspices of CEN. There is no plan to openthe c<strong>on</strong>tents of Eudamed to the public.The USAIn the USA, the evaluati<strong>on</strong> and approval of medical devices is theresp<strong>on</strong>sibility of the Food and Drug Administrati<strong>on</strong> (FDA)through its Center for <strong>Devices</strong> and Radiological Health (CDRH),which has a Divisi<strong>on</strong> of <strong>Cardiovascular</strong> <strong>Devices</strong>. As in Europe,devices are categorized by risk into three classes, and higherlevels of evidence including clinical evidence are required forapproval of devices in Class III. In additi<strong>on</strong> to internal regulatoryreview by FDA pers<strong>on</strong>nel, advice is often sought from external,independent experts through well-established <strong>Medical</strong> <strong>Devices</strong>Advisory Committees (including a Circulatory System <strong>Devices</strong>Advisory Panel), 19 for example, for first-of-a-kind devices and fordevices that are expected to have a broad impact <strong>on</strong> publichealth. This provides a transparent mechanism for public reviewof issues relating to the approval of complex devices.Low-risk devices (class I) are approved by registrati<strong>on</strong>, aftercompliance with general c<strong>on</strong>trols such as regulati<strong>on</strong>s for goodmanufacturing practice. More complex, moderate-risk devices(class II) require additi<strong>on</strong>al special c<strong>on</strong>trols, such as specific labelling,compliance with requirements in guidance documents,device tracking, and design c<strong>on</strong>trols. The regulatory pathway forclass II products is through the Premarket Notificati<strong>on</strong> [‘510(k)’]programme, which is the pathway through which most medicaldevices are marketed. A manufacturer submits a ‘premarket notificati<strong>on</strong>’asserting that their new device is substantially equivalent(i.e. at least as safe and effective) as <strong>on</strong>e (‘the predicate’) that isalready legally <strong>on</strong> the market. If the FDA determines that theinformati<strong>on</strong> and performance testing dem<strong>on</strong>strate that this is thecase, then the device can be marketed in the USA. Ten to 15%of 510(k) submissi<strong>on</strong>s require data from human clinical studies tosupplement preclinical testing.High-risk devices (class III) follow the regulatory pathway forpremarket approval (PMA). This requires a comprehensive evaluati<strong>on</strong>including bench testing, preclinical animal studies, and clinicaldata, so that the device dem<strong>on</strong>strates a reas<strong>on</strong>able assurance ofsafety and effectiveness. The term ‘effectiveness’ means that thedevice will provide clinically significant benefits, and thus evaluati<strong>on</strong>focuses <strong>on</strong> clinical outcomes such as a reducti<strong>on</strong> in symptoms oradverse events. In c<strong>on</strong>trast, ‘performance’ (a standard EU regulatoryparameter) focuses <strong>on</strong> a device’s mechanism of acti<strong>on</strong> such asenlargement of the arterial lumen or enhanced myocardial bloodflow. Clinical data for a PMA are typically obtained from feasibilitystudies followed by a larger, ‘pivotal’ trial. In order to c<strong>on</strong>duct aclinical trial in the USA to assess the safety and effectiveness of a‘significant risk’ device and thereby provide evidence that willsupport a PMA or 510(k), the sp<strong>on</strong>sor must obtain FDA approvalof an Investigati<strong>on</strong>al Device Exempti<strong>on</strong> (IDE).In general, regulatory reviews by the FDA incorporate moredetailed technical standards and requirements for clinical evaluati<strong>on</strong>of devices than occurs in Europe, and they are more rigorous.Preclinical testing plays a more important role, with questi<strong>on</strong>sregarding l<strong>on</strong>g-term durability and local and systemic histopathologicalresp<strong>on</strong>ses being addressed in bench and animal studies. Sincethe FDA requires adequate preclinical safety data before theinitiati<strong>on</strong> of human trials, there may be a lag of several yearsafter the introducti<strong>on</strong> of some devices for clinical use withinEurope before they are used in patients in the USA. Many US companiesfirst get approval for their new medical devices in the EUand then use ‘OUS’ (outside the USA) data relating to safety andeffectiveness to support the initiati<strong>on</strong> of a pivotal IDE study atUS study sites.In a recent decisi<strong>on</strong>, the US Supreme Court judged that themanufacturer of a medical device could not be sued by a patient<strong>on</strong> the basis of alleged defects relating to its safety because thedevice had received PMA from the FDA. 20 The c<strong>on</strong>sequences ofthis decisi<strong>on</strong> are still unclear; for example, it is uncertain if it mayencourage manufacturers to submit more detailed claims fordevices when applying for regulatory approval.For permanently implanted class III devices such as cor<strong>on</strong>arystents and devices for occluding intracardiac shunts, the FDA typicallyrequires 5 years of clinical follow-up. Post-approval clinicalstudies that collect and report real-world outcomes associatedwith the use of novel devices are also comm<strong>on</strong>ly required. TheFDA has had a policy of ‘global transparency’ since 1994 and allmedical device reports (MDRs) c<strong>on</strong>cerning significant adverseevents are available <strong>on</strong> the Internet.United States and OUS device manufacturers can meet withFDA staff at multiple times during the development of a device,through the CDRH’s Pre-Submissi<strong>on</strong> Program. These informalmeetings are particularly useful for the sp<strong>on</strong>sors and the FDA toreach c<strong>on</strong>sensus <strong>on</strong> preclinical device testing, key elements ofIDE clinical trials, and requirements for PMA submissi<strong>on</strong>s. The510(k) program has been reviewed by the Institute of Medicine 21and it is being reviewed by the FDA. Recommendati<strong>on</strong>s whichDownloaded from http://eurheartj.oxfordjournals.org/ by guest <strong>on</strong> February 28, 2013