Identifying and using similar devices
A device's clinical documents are expected to say which other devices on the market resemble it and what their data show, even where the device's own evidence does not rest on them. A similar device needs less to qualify than an equivalent one, and its data serve the clinical evaluation plan, the clinical evaluation report and the post-market clinical follow-up documents in ways the regulation and its guidance set out.
Similar devices and equivalent devices
Under the MDR, similar devices are devices belonging to the same generic device group, which the regulation defines as "a set of devices having the same or similar intended purposes or a commonality of technology allowing them to be classified in a generic manner not reflecting specific characteristics". Put more plainly, any device can serve as a similar device if it has the same or a similar intended purpose, or is based on a common technology, without the exhaustive demonstration of technical, biological and clinical characteristics that is obligatory to claim equivalence. Another key difference is access: a manufacturer is not required to have "sufficient levels of access to the data relating to devices" for a similar device, as it is for an equivalent one. Claiming similarity is therefore more straightforward than claiming equivalence.
An Ilizarov lower extremity ring external fixator, for example, stabilises broken bones with rigid rods, pins and connectors inserted through the skin into the bone. By "same or similar intended purpose", its similar devices are the various devices used to fix a broken upper or lower leg; by "commonality of technology", they are any external fixator. The definition joins its two elements with OR, not AND, which makes them theoretically mutually exclusive, and in 2023 how this would be used in practice, in preparing clinical documents and in reviewing them, remained to be seen.
What similar-device data can be used for
Where a clinical evaluation cannot be based on equivalence, MDCG 2020-5, a Medical Device Coordination Group guidance document, lists six ways similar devices may be useful: making sure the risk management system is comprehensive by identifying relevant hazards and clinical risks; understanding the state of the art, the natural course of the disease and the alternative treatment options available; helping define the scope of the clinical evaluation by identifying design features in similar devices that pose special performance or safety concerns; providing input to the design of a clinical investigation or of post-market clinical follow-up, and to the post-market surveillance system; identifying relevant and specified clinical outcome parameters for the intended clinical benefits from the published data on the similar devices; and defining minimum requirements for a quantified clinical benefit considered clinically relevant, or identifying acceptable occurrence rates of risks and adverse events.
Similar-device data can supply threshold or benchmark values against which the clinical performance and safety of the manufacturer's own device are evaluated, and they are expected to be used that way. The clinical evaluation plan and the post-market clinical follow-up (PMCF) plan are good documents to set those benchmarks in. The data also strengthen the clinical conclusions and let manufacturers make important decisions in both the pre-market and post-market phases.
For legacy devices based on well-established technologies, MDCG 2020-6, another of the group's guidance documents, gives three uses of similar-device data: data from similar devices may be used, for example, to demonstrate ubiquity of design, lack of novelty, and the known safety and performance profile of a generic group of devices; for well-established technologies the clinical evaluation can be based on data from similar devices, under the conditions in its paragraph 6.5(e); and similar-device data may be important in establishing whether the device under evaluation and its similar devices belong to the group of devices considered well-established technologies.
Similar-device data can benefit every kind of clinical evaluation — one based on the manufacturer's own clinical data, on an equivalent device's, or on no clinical data — in both the pre-market and post-market stages, and could further support legacy devices based on well-established technologies.
What the regulation and the notified body expect
Under the MDR, manufacturers are expected to identify similar devices and discuss their data in the clinical documents, especially clinical evaluation reports and PMCF evaluation reports. MDR Annex II 1.2(b) requires the technical documentation to include "an overview of identified similar devices available on the Union or international markets, where such devices exist", and the clinical evaluation assessment report template in MDCG 2020-13, at section J, has the notified body check for similar-device data. Manufacturers are therefore expected to search for similar devices and to identify and include them in their technical documentation where they exist. Where none exist, which happens only in exceptional cases such as a unique intended purpose and technology, the manufacturer can state so clearly in the clinical evaluation report to manage the reviewer's expectations from the start.
A stepwise approach
In 2023 the process of identifying and selecting similar devices was not well defined, which left manufacturers several possible approaches. One stepwise approach follows the structure MEDDEV 2.7/1 revision 4 sets for clinical evaluation reports; it is not exhaustive, and several other approaches can be used.
- Identification and selection (stage 0, scoping). First identify the similar devices, which can easily be the most challenging step: the selection has to rest on scientifically sound rationales such as the devices' market share, the quality and quantity of the available data, and their known safety and performance profiles. Then justify the choice — why five devices and not 20 — by showing strict compliance with the MDR definition, same or similar intended purpose or commonality of technology, and with the criteria in section E of MDCG 2020-7: product name, intended purpose, intended users, intended patient population, medical condition and indication. This must be done at the level of the PMCF plan and should ideally be done in the clinical evaluation plan or report as well.
- Data identification (stage 1). Once the devices are selected, all the data available on them or known to the manufacturer must be searched for and discussed in the clinical documents; the sources may include scientific literature such as peer-reviewed articles and Cochrane systematic reviews, post-market clinical follow-up studies, post-market surveillance data, and registries and other real-world evidence.
- Appraisal (stage 2). This step could be optional; like the clinical data, similar-device data can be appraised against suitability and contribution criteria.
- Analysis (stage 3). The data must be analysed in the PMCF evaluation report, ideally also in the clinical evaluation report. In the PMCF evaluation report, manufacturers are encouraged to follow section D of MDCG 2020-8; in the clinical evaluation report the data can be discussed separately from the state of the art or as part of it, or in an independent set of documents with appropriate interfaces to the clinical documents. The conclusions drawn from the data must then be used to update risk management and post-market surveillance activities, including post-market clinical follow-up.
Like the rest of the clinical evaluation, similar-device data have to be updated continuously and systematically.
Similar-device data and patient safety
All this is a lot of work, but one author expects the industry to agree that it is for the better. In that author's view, had similar-device data been analysed and the so-called class effect identified earlier, several thousand patients could have been spared the complications of metal-on-metal hips. The goal of the effort is patient safety first.