The most recent revisions to the European medical devices directive have underlined the importance of clinical data in demonstrating compliance to the Essential Requirements. The first step in this process is the clinical literature evaluation, for which there are three key sources of guidance: ISO 14155-1 (2009) Annex A, MedDev 2.7.1 and GHTF document SG5/N1R8. However, many manufacturers are still unclear as to Notified Body and Competent Authority expectations, and how much data is required to demonstrate that a clinical investigation is not required.
As the majority of devices are launched on the basis of equivalence to existing devices, and since review of the literature is the first step in any clinical evaluation, this article will focus on issues related to evaluation by literature review rather than clinical trial. The most common problems Notified Body reviewers find in clinical literature evaluations are:
- Poorly established design equivalence
- Insufficient data provided to demonstrate the short to medium term safety and efficacy of the equivalent (and by extension subject) devices.
- Adequacy of postmarket surveillance including postmarket clinical follow-up (PMCF)
Whether or not there is any clinical data available for the devices themselves, the literature review should include any available data for similar devices. For this purpose, ‘similar’ can be approached from two perspectives. In terms of establishing baseline performance and relevant risks, equivalence can be considered fairly broadly and might include all devices with similar indications for use. For example, if trying to demonstrate that a particular design of collarless cementless hip stem is of a safe and effective design, the manufacturer might first consider the performance of hip replacements in general, how this performance compares with cementless hip stem performance and any generic risks and complications for these devices.
However, in establishing the adequacy of existing clinical data to demonstrate safety and performance for a specific subject device, more detail is required. For these purposes, the manufacturer should describe all known design differences and provide an evaluation of the potential impact of these differences on anticipated clinical safety and performance. This should include a detailed examination of differences in:
- Bone contacting geometry (e.g., impact on ease and reproducibility of surgical technique; extent of bone resection and therefore options for revision; load transmission and therefore, potential for bone resorption or cement fixation, etc.). Note that even seemingly small differences might affect performance. For example, a change in radius for a taper edge or groove patterning could affect stress in the cement mantle and subsequently, medium to long term fixation, or a slightly longer stem for a hip could result in increased thigh pain. Hence, all differences should be discussed.
- Articulation geometry (e.g., impact on range of motion, kinematics, wear, etc.)
- Surface finish and patterning (e.g., impact on cement fixation, bone ingrowth, subsidence, etc.)
- Coatings (e.g., impact of differences in density, porosity and surface roughness as well as possible differences in material specification)
- Material properties (although normally, devices made from different materials will not be considered equivalent, unless strong justification is provided)
As even small design differences can have a significant impact on clinical results, it would not normally be expected that a manufacturer could claim strong equivalence to several different devices.
Two additional factors should be considered when establishing design equivalence:
- Design evolution of the device with which equivalence is claimed. As a device design may incorporate incremental changes over several years, the manufacturer should establish the specifics of the design for which data is presented, especially if older studies are referenced.
- Design variants within a family. Often manufacturers will have several design variants within a family that have the same name, but would not be considered equivalent to one another. The clinical literature evaluation should demonstrate, as far as practically possible, that the devices for which data are presented are for the specific designs with which equivalence has been established.
As suggested by the above, when a dossier includes more than one distinct design variant, the clinical impact of these design differences on indications as well as safety and performance risks should be discussed.