Characteristics of included studies.

Background
Regulation and postmarket surveillance of medical devices have been criticized for being too lenient as compared to drug regulation and postmarket surveillance. Little is known about the factors that determine which medical devices are chosen for implementation among similar medical devices.
Objective
Our aim was to systematically identify and characterize empirical studies on medical device regulation, implementation and postmarket surveillance, and to describe the recurring themes and trends in the studies.
Methods
The scoping review was preregistered, with the protocol publicly available (https://osf.io/mx36f). We followed the JBI guidelines for scoping reviews and reported the review in accordance with PRISMA-ScR guidelines. Last searches were done in MEDLINE and Embase through Ovid on 8th of February 2024. We included primary studies with empirical data, and we excluded any secondary studies such as editorials, opinion papers or systematic reviews using bibliographic databases as the primary data source. We were interested in studies that examined medical devices approved by the U.S. Food and Drug administration (FDA) and European Union (EU), and any studies on the decision-making process regarding medical device implementation.
We described study characteristics and mapped them graphically. Recurring themes were presented in a table. Furthermore, we reported conclusions from identified essential studies and provided a summary of the main results. Graphs and descriptive statistics were done in R version 4.3.2, package ggplot2.
Results
We screened 3862 titles/abstracts, after which 368 records were assessed in full-text, yielding 139 studies included in the review. Out of these, 68 studies (49%) examined approval, 40 studies (29%) examined postmarket surveillance, 17 studies (12%) implementation and 14 studies (10%) both approval and postmarket surveillance. The studies were published between 2003–2024 and consisted of 77 cross-sectional studies (55%), 35 cohort studies (25%), 20 qualitative studies (14%) and seven mixed-methods studies (5%). As data source, 90 studies (65%) used FDA, 25 studies (18%) other data sources and 24 studies (17%) interviewees through semi-structured interviews. Nine out of the 139 studies investigated regulatory approval within the EU. Predominantly, the studies reported that the available clinical evidence for medical device approval was considered inadequate, making it difficult for stakeholders to evaluate the suitability of a medical device for implementation.
Conclusions
Studies on medical devices are mainly conducted using FDA device databases, since restricted access to publicly available data has hindered research within the EU. Research on how and why specific medical devices are chosen and adopted into clinical practice is limited. We suggest that evidence on medical device efficacy and harms should be strengthened through higher demands from regulatory agencies and improvement of accessible registries.