Clinical Evaluation and Performance Evaluation of Medical Device Software

The requirements for the clinical evaluation and performance evaluation of medical device software (MDSW) under the EU Medical Devices Regulation (MDR) and In Vitro Diagnostic Medical Devices Regulation (IVDR) are defined in the MDCG 2020-1 guidance document issued by the Medical Device Coordination Group (MDCG). This guideline is addressed to manufacturers, notified bodies, and other stakeholders. Its objective is to provide a transparent, comprehensible, and robust framework for conducting clinical evaluation (under MDR) and performance evaluation (under IVDR) of medical device software in accordance with European regulation and for the protection of patients.

Accordingly, the clinical evaluation process (MDR) and performance evaluation process (IVDR) for medical device software should be designed as systematic and planned procedures that continuously generate, collect, analyse, and assess clinical data to demonstrate the safety and performance of the software with respect to its intended purpose. Manufacturers must provide three key types of evidence:

• scientific validity,
• technical or analytical performance, and
• clinical performance of the software.

Scientific validity demonstrates that the outputs of the software are associated with the intended clinical condition or physiological parameter. Technical performance ensures that the software functions as intended, while clinical performance demonstrates that the software delivers clinically relevant results under real-world conditions. The evaluation process includes planning, collection and appraisal of data, analysis, and ongoing documentation. Various data sources are used throughout the product lifecycle, including scientific literature, regulatory recommendations, databases, and expert opinions.

In accordance with MDR Article 61(1) and IVDR Article 56(1), the required level of clinical evidence must be appropriate to the intended purpose and characteristics of the medical device. For devices where demonstration of conformity with the General Safety and Performance Requirements (GSPRs) based on clinical data is not deemed appropriate (MDR Article 61(10)), the manufacturer must provide a detailed justification in the technical documentation as to why conformity is demonstrated solely on the basis of non-clinical testing methods such as performance evaluation, bench testing, preclinical evaluation, and usability assessment. This justification must be based on the outcomes of the risk management process. It should also include an assessment of the current clinical state of the art, including alternative diagnostic and therapeutic options known from the literature, and their relevance to the device under evaluation. Special attention must be paid to the device/body interaction, the intended clinical performance, and the manufacturer’s claims. Otherwise, similar principles apply as for the clinical evaluation of non-software medical devices.

Annex II of MDCG 2020-1 provides a structured overview of the requirements for the technical documentation of medical device software with regard to clinical evaluation (MDR) and performance evaluation (IVDR). The annex lists the essential elements that manufacturers must include in their technical documentation to demonstrate conformity with regulatory requirements. These include:

• Description of the software: Detailed information on the intended purpose, functions, and target population.
• Summary of clinical evaluation/performance evaluation: Presentation of the evaluation process, methods used, and key results.
• Evidence of scientific validity: Demonstration that the software outputs are associated with the intended clinical condition or parameter.
• Evidence of technical/analytical performance: Results from technical testing showing that the software functions as intended.
• Evidence of clinical performance: Data demonstrating that the software delivers the claimed clinical performance under real-world conditions.
• Risk and side-effect assessment: Presentation of identified risks, side effects, and measures to mitigate them.
• Summary of literature review: Overview of the scientific publications used and their relevance to the software.
• Conclusion on conformity: Summary assessment of whether and how the software meets the General Safety and Performance Requirements.

‍