Clinical Evaluation under the EU MDR
Conformity assessment of a medical device under the MDR requires a demonstration that the device meets the general safety and performance requirements (GSPRs), including a clinical evaluation (Article 5).
The clinical evaluation should assess the performance and safety of the device under the normal conditions of its intended use, including any undesirable side-effects and the acceptability of the benefit-risk “based on clinical data providing sufficient clinical evidence” (Article 61.1).
As part of the clinical evaluation, clinical benefits of the device “in terms of a meaningful, measurable, patient-relevant clinical outcome(s)” are expected to be shown. These outcomes are identified from clinical data on the device.
Therefore, clinical data is a necessary and crucial requirement for successful conformity assessment.
Clinical data and sources for conformity assessment
Clinical data refers to information concerning safety or performance that is generated from the
- use of a device or
- a device for which equivalence to the device in question can be demonstrated,
and is sourced from
- clinical investigation(s) or
- other studies reported in scientific literature,
- reports published in peer reviewed scientific literature, as well as
- clinically relevant information coming from post-market surveillance, in particular the post-market clinical follow-up.
(Article 2 (48))
For establishing device conformity to the relevant GSPRs, the suitability, quality, and quantity of different clinical data sources and the required “level of clinical evidence” must be justified by the manufacturer in view of the characteristics of the device and its intended purpose device (Article 61.1). This is discussed in the MDCG 2020-6 guidance on legacy devices.
Clinical data sources have a “level of clinical evidence” weighting hierarchy (Appendix III of MDCG 2020-6), with high-quality data from the sources indicated above considered high “level of clinical evidence” or pivotal clinical data sources. Notably, for implantable and class III devices clinical data from these high-quality data sources must include data on the device under evaluation, not solely “equivalent” device data.
Lower weighted “level of clinical evidence” data sources can provide non-pivotal, supporting clinical evidence, and include compliance with standards/common specification, pre-clinical testing, and data from generic devices in the same class as the device under evaluation.
An exception is devices considered “well-establish technology” (WET), where lower weighted “level of clinical evidence” data sources, including from generic class devices, can be used to support their conformity assessment. WET devices must display a simple, common, and stable designs, with well-known performance and safety across the generic device group and a long history on the market, and commonly applies to class I and/or standard-of-care devices.
What constitutes an “equivalent” device and is this the same as a “predicate“ device?
In simplified terms, an “equivalent” device under the MDR is a single device with similar technical characteristics, and both the same biological and clinical characteristics, when compared to the device under evaluation. To confirm equivalence, a side-by-side comparison of technical, biological, and clinical parameters should be made, as indicated in Annex I of MDCG Guidance 2020-5. Importantly, differences in technical, biological, and clinical characteristics between the two devices that exist must be identified, and for each difference a scientific justification provided as to why these differences would not have a clinically significant impact on performance and safety.
Additionally, it must be “clearly demonstrated that manufacturers have sufficient levels of access” to the relevant equivalent device data. As a device’s data related to technical characteristics is commonly only detailed within internal documentation, access to this information is often unattainable. For implantable and class III devices, the extent of this level of access is “contractually-agreed full and ongoing access to the technical documentation” on the equivalent device.
Consequently, equivalence according to MDR should not be automatically equated with terms found in other regulatory jurisdictions and texts, such as “predicate” or “substantially equivalent” devices as used in the US. These devices may function to support approval in the US market, but the higher requirements for equivalence under the MDR mean that these devices, like many devices previously considered equivalent under the MDD, are now likely to only qualify as “similar” devices. “Similar” devices are defined in the MDR as devices with shared characteristics which can be considered to belong to the same generic device group.
These “similar” devices may provide input for various aspects of the clinical evaluation process and provide contextual support to observed clinical performance and safety of the device under evaluation, for example as a benchmark device when establishing the State of the Art. However, the clinical data from “similar” devices cannot be used as part of the clinical evidence demonstrating performance and safety for the device under evaluation.
What differentiates a clinical validation from usability/feasibility studies?
Clinical validation is a term relating to Medical Device Software / Software as a Medical Device (MDSW/SaMD) defined by the International Medical Device Regulators Forum (IMDRF) as the measurement of a medical device software’s ability to yield a clinically meaningful output associated to the target use of software output in the target health care situation or condition (Guidance on Clinical Evaluation of SaMD [N41/2017])
Clinical validation is required for all MDSW, as it demonstrates that the software’s performance in intended clinical use settings, reflects its demonstrated analytical/technical performance. It is performed on the MDSW prior to entering the market, and continues after market entry. In contrast, the analytical / technical verification and validation activities take place during software development to establish the software’s technical performance, primarily accuracy and reliability, prior to clinical validation.
Usability studies/testing (also referred to as human factors testing) relates to activities performed to demonstrate compliance to the international medical device usability standard: IEC 62366. It assesses the ability of intended users of a medical device to employ the device, in an expected use context. Its goal is to identify risks or limitations that may impede the specified users from appropriately using the device. Outcomes from usability testing allow the manufacturer to take action to eradicate or minimize these aspects in the final device and its use. Once EN IEC 62366 is harmonised with the MDR, it can be used to demonstrate compliance with MDR usability requirements.
Feasibility studies (also referred to as exploratory studies) are small scale clinical trials conducted in the early stages of the medical device development to establish preliminary safety and effectiveness of the device. Feasibility studies inform the design of pivotal or confirmatory clinical investigations to generate clinical data that demonstrates device safety and performance. Both feasibility and pivotal clinical trials are expected to be performed according to article 62, annex XV and Good Clinical Practice (EN ISO 14155:2020). Notably, Class III and implantable devices require device data from manufacturer sponsored clinical investigations.
Can clinical data from studies performed outside the EU be used in the MDR conformity assessment procedure?
The MDR indicates that clinical investigations are expected to be performed in EU Member States with the aim to generate data for device conformity assessment (Article 62). Studies must be conducted according to Good Clinical Practice (EN ISO 14155:2020), have ethics approval, informed consent, and respect additional local EU Member State requirements in which they are performed.
The MDR does not specifically address clinical investigations performed in different regulatory jurisdictions. However, the IMDRF guidance on clinical evaluation (N56:2019, Appendix D) does, stating that “when clinical investigations are conducted ethically in accordance with applicable Good Clinical Practice (GCP), the clinical data should be accepted for consideration in any jurisdiction.” However, the guideline stresses that this is dependent on the clinical investigation being conducted to the same regulatory standards, and reflect the population and intended use of the jurisdiction in which it will be marketed.
Accordingly, clinical investigations performed outside the EU would need to be performed according to ISO 14155:2020 (and MDR Annex XV), and ensure any differences have no clinically significant impact on the generated clinical data. Therefore, any non-EU clinical data generated must be critically assessed and a scientific justification provided as to why any regional, intrinsic, and extrinsic factors, like regional population genetics or regional diseases, have not impacted the validity of the resulting clinical data and conclusions, relative to the intended population and intended clinical use setting in the EU.
What sample size do I need to include in my study?
Sample size is dependent on many variables and therefore a fixed sample size is not specified under MDR.
Sample size is calculated to support the generation of reliable and statistically robust results, including by factoring in variability that is unexpected or unrelated to the device, for example enrolled participants dropping-out of the study or being lost to follow-up. Error sources such as these will impact on the statistical significance of assessed performance and safety outcomes/endpoints for the device.
Accordingly, the MDR specifies that sample size is to be determined based on power calculations of the clinical investigation (Annex XV).
The power of a clinical investigation is the probability that the study will detect a true predetermined difference in measurement (endpoint) between two groups. The power is impacted by the magnitude of the expected difference in the endpoints, sample size and the probability of false positive error or false negative errors.
To determine the sample size needed to confirm a true statistically significant difference of a specified endpoint, an assumption must be made on the expected magnitude of the endpoint to be measured. This requires careful consideration of pre-clinical or exploratory clinical data for the device and existing published clinical data for similar/comparator devices. Poorly defined endpoints can lead to various negative outcomes, including study failure through the samples size being too small, or to unnecessary costs due to the samples size being too large.
How can Decomplix help?
Understanding the requirements regarding clinical data is crucial for accessing your target markets. For example, do you have questions on what clinical data are required for your devices to successfully undergo the conformity assessment in the EU? Our team of experts will be happy to support you in developing the regulatory strategy and specific questions on clinical data, clinical evaluation reports and clinical studies.