BACKGROUND
Artificial Intelligence (AI) use cases in healthcare are on the rise with the potential of improving operational efficiency and care outcomes. However, successfully translating AI into practical everyday use and effective investment requires adoption from stakeholders and the healthcare industry at large.
OBJECTIVE
As adoption is a key factor of successful innovation proliferation, this scoping review is aimed at presenting an overview of the barriers and facilitators of AI adoption in healthcare.
METHODS
A scoping review was conducted using guidance provided by the Joanna Briggs Institute and the framework by Arksey and O’Malley. Medline®, IEEEXplore® and ScienceDirect® databases were searched to identify publications in English that reported on the barriers or facilitators of AI adoption in healthcare. This review focused on articles published between the periods of 2011 and September 2021 and did not limit the healthcare setting (hospital or community) nor the stakeholders that were studied (patients, clinicians, physicians, healthcare administrators). A thematic analysis was conducted on the selected articles to map factors associated with barriers and facilitators of AI adoption in healthcare.
RESULTS
1250 articles were identified in the initial search. After title and abstract reviews were conducted, 23 articles were found eligible for full text review. These articles were reviewed for barriers and facilitators of AI adoption in healthcare. The majority of the articles were empirical studies, literature reviews and reports or thought pieces. Approximately 18 categories of barriers and facilitators were identified. These were organized sequentially to provide considerations for AI development, implementation and the overall structure needed to facilitate adoption.
CONCLUSIONS
The literature review revealed that trust is impacted by many of the elements identified as barriers to adoption of AI. Trust is a significant catalyst of adoption. A governance structure can be a key facilitator, amongst others in ensuring all the elements identified as barriers are addressed appropriately. The findings demonstrate that the implementation of AI in healthcare is still, in some ways, away from the context of setting of regulatory and legal frameworks that would enable the adoption of AI. Further research is needed surrounding the frameworks/models/theories that could be used to enhance trust in AI thereby enabling adoption. Such research is necessary to provide the needed guidance to those implementers translating novel AI research from bench to bedside.