Assessing utility, impact and adoption challenges of AI-enabled prescription advisory tool for type 2 diabetes management: perspectives from endocrinologists in a tertiary hospital (Preprint) DOI
Sungwon Yoon, Hendra Goh, Phong Ching Lee

et al.

Published: July 17, 2023

BACKGROUND The clinical management of type 2 diabetes mellitus (T2DM) presents a significant challenge due to the constantly evolving practice guidelines and growing array drug classes available. Evidence suggests that artificial intelligence (AI)–enabled decision support systems (CDSSs) have proven be effective in assisting clinicians with informed decision-making. Despite merits AI-driven CDSSs, research gap exists concerning early-stage implementation adoption AI-enabled CDSSs T2DM management. OBJECTIVE This study aimed explore perspectives on use impact Prescription Advisory (APA) tool, developed using multi-institution registry implemented specialist endocrinology clinics, challenges its application. METHODS We conducted focus group discussions semistructured interview guide purposively selected endocrinologists from tertiary hospital. were audio-recorded transcribed verbatim. Data thematically analyzed. RESULTS A total 13 participated 4 discussions. Our findings suggest APA tool offered several useful features assist effectively managing T2DM. Specifically, viewed AI-generated medication alterations as good knowledge resource supporting clinician’s decision-making modifications at point care, particularly for patients comorbidities. complication risk prediction was seen positively impacting patient care by facilitating early doctor-patient communication initiating prompt responses. However, interpretability scores, concerns about overreliance automation bias, issues surrounding accountability liability hindered practice. CONCLUSIONS Although holds great potential valuable improving further efforts are required address clinicians’ improve tool’s acceptance applicability relevant contexts.

Language: Английский

What Are Humans Doing in the Loop? Co-Reasoning and Practical Judgment When Using Machine Learning-Driven Decision Aids DOI Creative Commons
Sabine Salloch, Andreas Eriksen

The American Journal of Bioethics, Journal Year: 2024, Volume and Issue: 24(9), P. 67 - 78

Published: May 20, 2024

Within the ethical debate on Machine Learning-driven decision support systems (ML_CDSS), notions such as "human in loop" or "meaningful human control" are often cited being necessary for legitimacy. In addition, principles usually serve major point of reference guidance documents, stating that conflicts between need to be weighed and balanced against each other. Starting from a neo-Kantian viewpoint inspired by Onora O'Neill, this article makes concrete suggestion how interpret role overcome perspective rivaling evaluation AI health care. We argue patients should perceived "fellow workers" epistemic partners interpretation ML_CDSS outputs. further highlight meaningful process integrating (rather than weighing balancing) is most appropriate medical AI.

Language: Английский

Citations

21
A Reflection Machine to Support Critical Reflection During Decision-Making DOI
Simon W. S. Fischer

Published: March 18, 2025

Language: Английский

Citations

0
Promoting Responsible Use of AI in African Healthcare: Strengthening Patients’ Moral Agency DOI Creative Commons
Edmund Terem Ugar

Asian Bioethics Review, Journal Year: 2025, Volume and Issue: unknown

Published: April 25, 2025

Language: Английский

Citations

0
Assessing the Utility, Impact, and Adoption Challenges of an Artificial Intelligence–Enabled Prescription Advisory Tool for Type 2 Diabetes Management: Qualitative Study DOI Creative Commons
Sungwon Yoon, Hendra Goh, Phong Ching Lee

et al.

JMIR Human Factors, Journal Year: 2024, Volume and Issue: 11, P. e50939 - e50939

Published: May 6, 2024

Background The clinical management of type 2 diabetes mellitus (T2DM) presents a significant challenge due to the constantly evolving practice guidelines and growing array drug classes available. Evidence suggests that artificial intelligence (AI)–enabled decision support systems (CDSSs) have proven be effective in assisting clinicians with informed decision-making. Despite merits AI-driven CDSSs, research gap exists concerning early-stage implementation adoption AI-enabled CDSSs T2DM management. Objective This study aimed explore perspectives on use impact Prescription Advisory (APA) tool, developed using multi-institution registry implemented specialist endocrinology clinics, challenges its application. Methods We conducted focus group discussions semistructured interview guide purposively selected endocrinologists from tertiary hospital. were audio-recorded transcribed verbatim. Data thematically analyzed. Results A total 13 participated 4 discussions. Our findings suggest APA tool offered several useful features assist effectively managing T2DM. Specifically, viewed AI-generated medication alterations as good knowledge resource supporting clinician’s decision-making modifications at point care, particularly for patients comorbidities. complication risk prediction was seen positively impacting patient care by facilitating early doctor-patient communication initiating prompt responses. However, interpretability scores, concerns about overreliance automation bias, issues surrounding accountability liability hindered practice. Conclusions Although holds great potential valuable improving further efforts are required address clinicians’ improve tool’s acceptance applicability relevant contexts.

Language: Английский

Citations

3
Pitfalls of Artificial Intelligence in Medicine DOI Creative Commons
Bakheet Aldosari, Abdullah Alanazi

Studies in health technology and informatics, Journal Year: 2024, Volume and Issue: unknown

Published: Aug. 22, 2024

Artificial Intelligence (AI) offers great promise for healthcare, but integrating it comes with challenges. Over-reliance on AI systems can lead to automation bias, necessitating human oversight. Ethical considerations, transparency, and collaboration between healthcare providers developers are crucial. Pursuing ethical frameworks, bias mitigation techniques, transparency measures is key advancing AI’s role in while upholding patient safety quality care.

Language: Английский

Citations

2
From Angels to Artificial Agents? AI as a Mirror for Human (Im)perfections DOI Creative Commons
Pim Haselager

Zygon®, Journal Year: 2024, Volume and Issue: 0(0)

Published: May 18, 2024

Artificial intelligence (AI) systems paradoxically combine high levels of certain types and cognitive capacities (pattern recognition, reasoning, learning, memory, perception, etc.) with an absence understanding sentience (feeling, emotion). Apparently, it is possible to make great progress in modeling smartness without making towards genuinely what all the clever reasoning about. This relevant when dealing AI programs that produce potentially convincing propositional output on religious topics. article suggests genuine cannot amount authentic religiosity. Comparing ourselves other entities, (in)animate or (super)natural, has always been a way for humans understand better. Throughout ages, many different beings agents have functioned as tools self-examination, presenting us mirrors reflect at least some our characteristics, capacities, (im)perfections. The recent provides exciting, though sometimes worrisome, cases newly informed look ourselves. Thus, may profound effects how we regard others proud claim are smartest species planet turn out not mean much. Inspired by example Thomas Aquinas, comparison nearest neighbors extended chain being—namely, animals, angels, AI—may deepen appreciation features homo sapiens share organisms.

Language: Английский

Citations

1
A randomized controlled trial on evaluating clinician-supervised generative AI for decision support DOI

Rayan Ebnali Harari,

Abdullah Al-Taweel,

Tareq Ahram

et al.

International Journal of Medical Informatics, Journal Year: 2024, Volume and Issue: 195, P. 105701 - 105701

Published: Nov. 29, 2024

Language: Английский

Citations

1
AI @ Work: Human Empowerment or Disempowerment? DOI Creative Commons
Sabine T. Koeszegi

Published: Dec. 20, 2023

Abstract Recent advancements in generative AI systems fuel expectations that will free workers to resolve creative, complex, and rewarding tasks by automating routine repetitive work. Furthermore, algorithmic decision (ADS) improve quality providing real-time information insights, analyzing vast amounts of data, generating recommendations support decision-making. In this narrative, empowers achievements they could not reach without the technology. However, using work contexts may also lead changes workers’ roles identities, leading feelings reduced self-efficacy lower confidence their abilities a sense diminished value workplace, ethical decision-making abilities, professional integrity. Initial empirical findings on impact context point essential design aspects determine which narratives becomes reality. This chapter presents these initial makes suggestions.

Language: Английский

Citations

2
Generating a decision support system for states in the USA via machine learning DOI
Hüseyin Ünözkan

Expert Systems with Applications, Journal Year: 2024, Volume and Issue: 246, P. 123259 - 123259

Published: Jan. 21, 2024

Language: Английский

Citations

0
AI Feedback Loops in Aotearoa New Zealand: A Human-centered Creative Problem-Solving Approach DOI
Victoria Agyepong, Céline Cattoën,

Kwasi Adusei-Fosu

et al.

Published: May 10, 2024

Artificial Intelligence (AI) can have unintended consequences in systems where they are deployed. Researchers found that by increasing contextual understanding of AI feedback loops, cause and effect systems, especially high-risk applications like health, biosecurity, conservation, justice transport, tools learn to improve over time leverage wider neural networks. This paper fills the knowledge gap on how consider varying competencies human-AI teams identify leveraging eight disciplines outside commerce computer science. The study academic actors from more than one discipline tend relevant sources applications. recommends integration human lived experiences, generated partial exposure ideas non-academic actors, decision making natural environment reduce incidence misinformed decision-making,

Language: Английский

Citations

0