This paper has been accepted and is currently in production.
It will appear shortly on 10.2196/101509
The final accepted version (not copyedited yet) is in this tab.
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Artificial Intelligence-Enabled Digital Health Tool Implementation in Clinical Practice Settings: An Umbrella Review
ABSTRACT
Background:
Artificial intelligence (AI) has the potential to improve clinical decision-making, efficiency, and patient outcomes, yet adoption in routine clinical practice has been slower than anticipated. While prior studies have identified factors that influence AI implementation, a comprehensive synthesis grounded in implementation science frameworks is lacking.
Objective:
To qualitatively and comprehensively describe barriers and facilitators to AI implementation in clinical care settings using an implementation science determinants framework the Consolidated Framework for Implementation Research (CFIR).
Methods:
We conducted an umbrella review examining determinants of AI implementation in clinical care settings. Following PRIOR reporting guidelines, we systematically searched PubMed, Scopus, and CINAHL for reviews published between January 2013 and July 2025. Barriers and facilitators were deductively mapped to the updated CFIR and inductively analyzed to identify AI-specific constructs not captured by CFIR. Review quality was assessed using the Joanna Briggs Institute checklist.
Results:
Twenty-two reviews published between 2021 and 2025 were included. Determinants of AI implementation spanned all CFIR domains. Prominent barriers included limited evidence of clinical impact, poor interpretability, workflow incompatibility, insufficient infrastructure, inadequate training, and lack of clear regulatory and liability frameworks. Key facilitators included perceived relative advantage, co-design with clinicians, robust IT infrastructure, access to training and technical support, and cross-sector partnerships. Across domains, we identified emergent AI-specific determinants—such as transparency, trust, accountability, model bias, data governance, model lifecycle maintenance, and clinician concerns regarding professional identity, skill erosion, and loss of the human element in care—that are not fully represented within CFIR.
Conclusions:
AI implementation in clinical settings is shaped by complex, multilevel, and AI-specific determinants that extend beyond traditional digital health considerations. Implementation science frameworks can be adapted to incorporate these constructs. Tailored, theory-informed strategies addressing technical, organizational, psychological, and regulatory factors are essential to support safe, equitable, and sustainable integration of AI into clinical care.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.