Background: In some randomized clinical trials, transradial access (TRA) compared with transfemoral access (TFA) was associated with lower mortality in patients with coronary artery disease undergoing invasive management. We analyzed the effects of TRA versus TFA across multicenter randomized clinical trials and whether these associations are modified by patient or procedural characteristics. Methods: We performed an individual patient data meta-analysis of multicenter randomized clinical trials comparing TRA with TFA among patients undergoing coronary angiography with or without percutaneous coronary intervention. The primary outcome was all-cause mortality and the co–primary outcome was major bleeding at 30 days. The primary analysis was conducted by 1-stage mixed-effects models on the basis of the intention-to-treat cohort. The effect of access site on mortality and major bleeding was assessed further by multivariable analysis. The relationship among access site, bleeding, and mortality was investigated by natural effect model mediation analysis with multivariable adjustment. Results: A total of 21 600 patients (10 775 TRA, 10 825 TFA) from 7 randomized clinical trials were included. The median age was 63.9 years, 31.9% were women, 95% presented with acute coronary syndrome, and 75.2% underwent percutaneous coronary intervention. All-cause mortality (1.6% versus 2.1%; hazard ratio, 0.77 [95% CI, 0.63–0.95]; P =0.012) and major bleeding (1.5% versus 2.7%; odds ratio, 0.55 [95% CI, 0.45–0.67]; P <0.001) were lower with TRA. Subgroup analyses for mortality showed consistent results, except for baseline hemoglobin level ( P _interaction =0.003), indicating that the benefit of TRA was substantial in patients with moderate or severe anemia, whereas it was not significant in patients with milder or no baseline anemia. After adjustment, TRA remained associated with 24% and 51% relative risk reduction of all-cause mortality and major bleeding, respectively. A mediation analysis showed that the benefit of TRA on mortality was only partially driven by major bleeding prevention and ancillary mechanisms are required to fully explain the causal association. Conclusions: TRA is associated with lower all-cause mortality and major bleeding at 30 days compared with TFA. The effect on mortality was driven by patients with anemia. The reduction in major bleeding only partially explains the mortality benefit. Registration: URL: https://www.crd.york.ac.uk/prospero ; Unique identifier: CRD42018109664.

Importance: The primary objective of any newly developed medical device using artificial intelligence (AI) is to ensure its safe and effective use in broader clinical practice.Objective: To evaluate key characteristics of AI-enabled medical devices approved by the US Food and Drug Administration (FDA) that are relevant to their clinical generalizability and are reported in the public domain.Design, Setting, and Participants: This cross-sectional study collected information on all AI-enabled medical devices that received FDA approval and were listed on the FDA website as of August 31, 2024.Main Outcomes and Measures: For each AI-enabled medical device, detailed information and key characteristics relevant for the generalizability of the devices at the time of approval were summarized, specifically examining clinical evaluation aspects, such as the presence and design of clinical performance studies, availability of discriminatory performance metrics, and age- and sex-specific data.Results: In total, 903 FDA-approved AI-enabled medical devices were analyzed, most of which became available in the last decade. The devices primarily related to the specialties of radiology (692 devices [76.6.%]), cardiovascular medicine (91 devices [10.1%]), and neurology (29 devices [3.2%]). Most devices were software only (664 devices [73.5%]), and only 6 devices (0.7%) were implantable. Detailed descriptions of development were absent from most publicly provided summaries. Clinical performance studies were reported for 505 devices (55.9%), while 218 devices (24.1%) explicitly stated no performance studies were conducted. Retrospective study designs were most common (193 studies [38.2%]), with only 41 studies (8.1%) being prospective and 12 studies (2.4%) randomized. Discriminatory performance metrics were reported in 200 of the available summaries (sensitivity: 183 devices [36.2%]; specificity: 176 devices [34.9%]; area under the curve: 82 devices [16.2%]). Among clinical studies, less than one-third provided sex-specific data (145 studies [28.7%]), and only 117 studies (23.2%) addressed age-related subgroups.Conclusions and Relevance: In this cross-sectional study, clinical performance studies at the time of approval were reported for approximately half of AI-enabled medical devices, yet the information was often insufficient for a comprehensive assessment of their clinical generalizability, emphasizing the need for ongoing monitoring and regular re-evaluation to identify and address unexpected performance changes during broader use.

Background: Female sex was not included among the high bleeding risk (HBR) criteria by the Academic Research Consortium (ARC) as it remains unclear whether it constitutes an HBR condition after percutaneous coronary intervention. We investigated whether female sex associates with HBR and assessed the performance of ARC HBR criteria separately in women and men. Methods and Results: Among all consecutive patients undergoing percutaneous coronary intervention between 2009 and 2018, bleeding occurrences up to 1 year were prospectively collected and centrally adjudicated. All but one of the originally defined ARC HBR criteria were assessed, and the ARC HBR score generated accordingly. Among 16 821 patients, 25.6% were women. Compared with men, women were older and had lower creatinine clearance and hemoglobin values. After adjustment, female sex was independently associated with access‐site (adjusted hazard ratio, 2.14; 95% CI, 1.22–3.74; P =0.008) but not with overall or non–access‐site 1‐year Bleeding Academic Research Consortium 3 or 5 bleeding. This association remained consistent when the femoral but not the radial approach was chosen. The ARC HBR score discrimination, using the original criteria, was lower among women than men (c‐index 0.644 versus 0.688; P =0.048), whereas a revised ARC HBR score, in which age, creatinine clearance, and hemoglobin were modeled as continuous rather than dichotomized variables, performed similarly in both sexes. Conclusions: Female sex is an independent predictor for access‐site bleeding but not for overall bleeding events at 1 year after percutaneous coronary intervention. The ARC HBR framework shows an overall good performance in both sexes, yet is lower in women than men, attributable to dichotomization of age, creatinine clearance, and hemoglobin values, which are differently distributed between sexes. Registration: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT02241291.

Objective Given the complexities of testing the translational capability of new artificial intelligence (AI) tools, we aimed to map the pathways of training/validation/testing in development process and external validation of AI tools evaluated in dedicated randomised controlled trials (AI-RCTs). Methods We searched for peer-reviewed protocols and completed AI-RCTs evaluating the clinical effectiveness of AI tools and identified development and validation studies of AI tools. We collected detailed information, and evaluated patterns of development and external validation of AI tools. Results We found 23 AI-RCTs evaluating the clinical impact of 18 unique AI tools (2009–2021). Standard-of-care interventions were used in the control arms in all but one AI-RCT. Investigators did not provide access to the software code of the AI tool in any of the studies. Considering the primary outcome, the results were in favour of the AI intervention in 82% of the completed AI-RCTs (14 out of 17). We identified significant variation in the patterns of development, external validation and clinical evaluation approaches among different AI tools. A published development study was found only for 10 of the 18 AI tools. Median time from the publication of a development study to the respective AI-RCT was 1.4 years (IQR 0.2–2.2). Conclusions We found significant variation in the patterns of development and validation for AI tools before their evaluation in dedicated AI-RCTs. Published peer-reviewed protocols and completed AI-RCTs were also heterogeneous in design and reporting. Upcoming guidelines providing guidance for the development and clinical translation process aim to improve these aspects.

The European Union Medical Device Regulation 2017/745 challenges key stakeholders to follow transparent and rigorous approaches to the clinical evaluation of medical devices. The purpose of this study is a systematic evaluation of published clinical evidence underlying selected high-risk cardiovascular medical devices before and after market access in the European Union (CE marking) between 2000 and 2021. Prespecified strategies were applied to identify published studies of prospective design evaluating 71 high-risk cardiovascular devices in 7 different classes (bioresorbable coronary scaffolds, left atrial appendage occlusion devices, transcatheter aortic valve implantation systems, transcatheter mitral valve repair/replacement systems, surgical aortic and mitral heart valves, leadless pacemakers, subcutaneous implantable cardioverter-defibrillator). The search time-span covered 20 years (2000-2021). Details of study design, patient population, intervention(s) and primary outcome(s) were summarized, and assessed with respect to timing of the corresponding CE-mark approval. At least one prospective clinical trial was identified for 70% (50/71) of the prespecifed devices. Overall, 473 reports of 308 prospectively designed studies (enrolling 97,886 individuals) were deemed eligible, including 81% (251/308) prospective non-randomized clinical trials (66,186 individuals) and 19% (57/308) randomized clinical trials (31,700 individuals). Preregistration of the study protocol was available in 49% (150/308) studies, and 16% (48/308) had a peer-reviewed publicly available protocol. Device-related adverse events were evaluated in 82% (253/308) of studies. An outcome adjudication process was reported in 39% (120/308) of the studies. Sample size was larger for randomized in comparison to non-randomized trials (median of 304 versus 100 individuals, p<0.001). No randomized clinical trial published before CE-mark approval for any of the devices was identified. Non-randomized clinical trials were predominantly published after the corresponding CE-mark approval of the device under evaluation (89%, 224/251). Sample sizes were smaller for studies published before (median of 34 individuals) than after (median of 135 individuals) CE-mark approval (p50 individuals) and those with longer recruitment periods were more likely to be published after CE-mark approval, and were more frequent during the period 2016-2021. The quantity and quality of publicly available data from prospective clinical investigations across selected categories of cardiovascular devices, before and after CE approval during the period 2000-2021, was deemed insufficient. The majority of studies were non-randomized, with increased risk of bias, and performed in small populations without provision of power calculations, and none of the reviewed devices had randomized trial results published prior to CE mark certification.