Explore the vast range of titles available.

Familiar concepts from research ethics can guide the meaningful and rigorous translation of artificial intelligence (AI) tools into clinical practice.

Artificial intelligence as a medical device is increasingly being applied to healthcare for diagnosis, risk stratification and resource allocation. However, a growing body of evidence has highlighted the risk of algorithmic bias, which may perpetuate existing health inequity. This problem arises in part because of systemic inequalities in dataset curation, unequal opportunity to participate in research and inequalities of access. This study aims to explore existing standards, frameworks and best practices for ensuring adequate data diversity in health datasets. Exploring the body of existing literature and expert views is an important step towards the development of consensus-based guidelines. The study comprises two parts: a systematic review of existing standards, frameworks and best practices for healthcare datasets; and a survey and thematic analysis of stakeholder views of bias, health equity and best practices for artificial intelligence as a medical device. We found that the need for dataset diversity was well described in literature, and experts generally favored the development of a robust set of guidelines, but there were mixed views about how these could be implemented practically. The outputs of this study will be used to inform the development of standards for transparency of data diversity in health datasets (the STANDING Together initiative). A systematic review, combined with a stakeholder survey, presents an overview of current practices and recommendations for dataset curation in health, with specific focuses on data diversity and artificial intelligence-based applications.

ObjectivesGiven widespread interest in applying artificial intelligence (AI) to health data to improve patient care and health system efficiency, there is a need to understand the perspectives of the general public regarding the use of health data in AI research.DesignA qualitative study involving six focus groups with members of the public. Participants discussed their views about AI in general, then were asked to share their thoughts about three realistic health AI research scenarios. Data were analysed using qualitative description thematic analysis.SettingsTwo cities in Ontario, Canada: Sudbury (400 km north of Toronto) and Mississauga (part of the Greater Toronto Area).ParticipantsForty-one purposively sampled members of the public (21M:20F, 25–65 years, median age 40).ResultsParticipants had low levels of prior knowledge of AI and mixed, mostly negative, perceptions of AI in general. Most endorsed using data for health AI research when there is strong potential for public benefit, providing that concerns about privacy, commercial motives and other risks were addressed. Inductive thematic analysis identified AI-specific hopes (eg, potential for faster and more accurate analyses, ability to use more data), fears (eg, loss of human touch, skill depreciation from over-reliance on machines) and conditions (eg, human verification of computer-aided decisions, transparency). There were mixed views about whether data subject consent is required for health AI research, with most participants wanting to know if, how and by whom their data were used. Though it was not an objective of the study, realistic health AI scenarios were found to have an educational effect.ConclusionsNotwithstanding concerns and limited knowledge about AI in general, most members of the general public in six focus groups in Ontario, Canada perceived benefits from health AI and conditionally supported the use of health data for AI research.

The Consolidated Standards of Reporting Trials extension for Artificial Intelligence interventions (CONSORT-AI) was published in September 2020. Since its publication, several randomised controlled trials (RCTs) of AI interventions have been published but their completeness and transparency of reporting is unknown. This systematic review assesses the completeness of reporting of AI RCTs following publication of CONSORT-AI and provides a comprehensive summary of RCTs published in recent years. 65 RCTs were identified, mostly conducted in China (37%) and USA (18%). Median concordance with CONSORT-AI reporting was 90% (IQR 77–94%), although only 10 RCTs explicitly reported its use. Several items were consistently under-reported, including algorithm version, accessibility of the AI intervention or code, and references to a study protocol. Only 3 of 52 included journals explicitly endorsed or mandated CONSORT-AI. Despite a generally high concordance amongst recent AI RCTs, some AI-specific considerations remain systematically poorly reported. Further encouragement of CONSORT-AI adoption by journals and funders may enable more complete adoption of the full CONSORT-AI guidelines. The CONSORT-AI extension was developed to provide specific guidance for randomised controlled trials involving Artificial Intelligence (AI) interventions. Here, the authors show that since publication of CONSORT-AI, several AI-specific considerations remain systematically underreported.

Artificial intelligence (AI) has vast potential for improving healthcare delivery, but concerns regarding biases in these systems have raised important questions regarding fairness when deployed clinically. Most prior studies on fairness in clinical AI focus solely on performance disparities between subpopulations, which often fall short of connecting the technical outputs of AI systems with sociotechnical outcomes. In this work, we present a simulation-based approach to explore how statistical definitions of algorithmic fairness translate to fairness in long-term outcomes, using AI-assisted breast cancer screening as a case example. We evaluate four fairness criteria and their impact on mortality rates and socioeconomic disparities, while also considering how clinical decision makers’ reliance on AI and patients’ access to healthcare affect outcomes. Our results highlight how algorithmic fairness does not directly translate into fair and equitable outcomes, underscoring the importance of integrating sociotechnical perspectives to gain a holistic understanding of fairness in healthcare AI. Artificial intelligence (AI) can greatly improve healthcare delivery and outcomes, but potential embedded biases can affect fairness in clinical deployment. Here, the authors develop a simulation-based approach to explore which formalisations of AI algorithmic fairness translate into long-term outcome fairness, with a focus on breast cancer.

IntroductionAny tool that can reduce the administrative burden on healthcare providers while preserving safe, accountable and high-quality medical documentation is of immense value both to healthcare institutions and consumers. The key question we need to answer is whether a prospective tool can reduce these burdens while maintaining (and, ideally elevating) quality documentation standards. The goal of this study is to describe the local performance of a large language model-based documentation assistive tool to draft safe, high-quality documentation in the Child Development Unit at the Women’s and Children’s Hospital. By generating local evidence of performance, we can assess the suitability of the artificial intelligence (AI) Scribe and inform a larger interventional study protocol and establish evidence-based governance.Methods and analysisUsing an algorithmic audit framework developed specific to our context, we will compare clinician-written clinical notes to AI-generated notes produced in parallel to the standard of care (ie, a ‘silent’ or translational trial paradigm). We will compare the time required to review clinical documentation per the standard of care compared with the AI-supported workflow with consideration to the accuracy of the final documentation. Finally, we will qualitatively describe AI-generated notes and compare them to the current standard to identify specific areas where clinical guidelines (eg, performance information, risk mitigation) would support appropriate clinical use.Ethics and disseminationEthics approval has been obtained by the Women’s and Children’s Health Network Human Research Ethics Committee (HREC) (HRE00067) and the South Australian Aboriginal HREC (#04-25-1185). This protocol offers an accessible example for health institutions looking to apply an evidence-based approach to AI Scribe assessment that prioritises clinical documentation standards. We will publish our study results in an academic journal and include a publicly accessible summary for the general public on the Women’s and Children’s website.Trial registration number10.17605/OSF.IO/P6TM5.

As more artificial intelligence (AI) applications are integrated into healthcare, there is an urgent need for standardization and quality-control measures to ensure a safe and successful transition of these novel tools into clinical practice. We describe the role of the silent trial, which evaluates an AI model on prospective patients in real-time, while the end-users (i.e., clinicians) are blinded to predictions such that they do not influence clinical decision-making. We present our experience in evaluating a previously developed AI model to predict obstructive hydronephrosis in infants using the silent trial. Although the initial model performed poorly on the silent trial dataset (AUC 0.90 to 0.50), the model was refined by exploring issues related to dataset drift, bias, feasibility, and stakeholder attitudes. Specifically, we found a shift in distribution of age, laterality of obstructed kidneys, and change in imaging format. After correction of these issues, model performance improved and remained robust across two independent silent trial datasets (AUC 0.85–0.91). Furthermore, a gap in patient knowledge on how the AI model would be used to augment their care was identified. These concerns helped inform the patient-centered design for the user-interface of the final AI model. Overall, the silent trial serves as an essential bridge between initial model development and clinical trials assessment to evaluate the safety, reliability, and feasibility of the AI model in a minimal risk environment. Future clinical AI applications should make efforts to incorporate this important step prior to embarking on a full-scale clinical trial.