A valid clinical association (scientific validity) determines the extent to which the output data of the SaMD are clinically accepted based on the input data, particularly in the context of the clinical situations or conditions that are identified in the intended purpose. The assessment of the clinical association considers whether the SaMD is part of a ‘well-established clinical association’ or a novel clinical association and may be based on published literature, professional guidelines, proof of concept studies, or clinical investigations.

For example, publicly available data to support the validity of the clinical association are collected, evaluated and analyzed in the state-of-the-art (SOTA) chapter of the CER. If the clinical association is found to be valid, no further evidence is required. If there is a gap in the information that can be found in the published literature and professional guidelines, own clinical data must be collected or reanalyzed to provide additional input.

Technical performance is the demonstration of the SaMD’s ability to accurately, reliably, and precisely generate the intended output from the input data (according to MDCG 2020-1). The basic technical performance, such as availability, confidentiality, accuracy, limit of detection, cut-off value(s), measuring interval, absence of inacceptable cybersecurity vulnerabilities, is confirmed by verification and validation (V&V) tests based on the software specifications.

Clinical performance can be demonstrated by testing the SaMD in the target population and for the intended use. Depending on the device characteristics and intended use, the methodology may include pre-clinical testing or clinical investigation. Clinical performance validation demonstrates clinical benefit in terms of measurable and patient-relevant outcome parameters (e.g., sensitivity, specificity, positive/negative predictive value, positive/negative likelihood ratio, usability, confidence interval(s)). If no clinical benefit is claimed for the SaMD, relevant clinical outcomes may be obtained through usability testing.

The clinical evaluation must also consider the risk-benefit ratio in the context of SOTA practices for diagnosis, treatment, or patient management.

The post-market surveillance requirements according to MDR are applicable for SaMD. The manufacturer should actively and continuously monitor the safety, effectiveness, and performance of the SaMD. The unique level of connectivity of SaMD facilitates access to real-world performance data which gives additional insight in the clinical performance data and allows the manufacturer to develop new claims and to anticipate further improvement in the device.