Content area
Full text
Introduction
Randomized clinical trials (RCTs) provide the foundation of evidence-based medicine [1]. Randomly assigning participants to different therapeutic strategies is the best way to minimize sources of bias and allows inference of causality between interventions and their clinical outcomes [2, 3]. Well-designed and accurately conducted RCTs robustly inform clinical practice guidelines and decision-making processes, but barriers to their conduct include high costs related to excessive complexity in the governance of trials, and limited generalizability to patients receiving the intervention in daily clinical practice when only highly selected cohorts of patients are studied [4]. Large simple RCTs including a more representative patient population can address both problems.
The Medical Device Regulation (MDR) that came into effect in the European Union (EU) in May 2021 provides a regulatory framework that aims to balance the efficient approval of new medical devices (or technical iterations of existing devices) with demonstration of their safety. It requires evidence to be presented in a Clinical Evaluation Report (CER) that supports the intended use and safety of a medical device, before market approval, and then periodic reports on safety to be submitted thereafter [5]. The MDR imposes higher standards for generating and assessing evidence on the performance and safety of high-risk medical devices (class III and implantable devices) than was required under the previous medical device directives, and it specifies in particular that “clinical investigations shall be performed for implantable and other high-risk medical devices” [6]. However, more specific methodological aspects of clinical investigations (i.e., type and/or design of studies) are not addressed in detail in the MDR. This creates uncertainty about what is considered sufficient clinical evidence and which types of studies are appropriate—particularly for new high-risk devices, for which there is generally a dearth of information [7, 8].
The principles of simplifying the design and avoiding unnecessary distractions in the conduct of RCTs were developed many years ago [9]. They have been reconfirmed for “streamlined” RCTs of drugs [6, 7] and demonstrated to be feasible for conducting large and simple RCTs of high-risk medical devices [10, 11]. Nowadays, the increasing availability of routinely collected healthcare data (for example, in registries) and the continuing development of more powerful information and communication technologies provide new opportunities for applying these concepts much more widely.