How to implement AI in primary care: co-design tools with users, research says

As Artificial Intelligence (AI) continues to advance and more health data becomes available through electronic health records, it is important to understand how AI or computer systems combined with datasets enable problem-solving in primary care settings. Previous studies have focused on developing AI tools; however, implementing them into practice remains a barrier. Researchers in Canada have identified trust as a key component in using AI in primary care and suggest “designing with, rather than for” stakeholders could promote the adoption of AI tools and patient-centred care.

In a study published in PLOS ONE, researchers gathered the perspectives of patients, primary care providers, and health system leaders through virtual deliberative dialogues to identify barriers and strategies to implement AI in primary care. The findings highlight four themes: (1) system and data readiness, (2) potential for bias and inequity, (3) regulation of AI and big data, and (4) importance of people as technology enablers.

Across all four themes, the common strategies are to co-design with stakeholders, seek user feedback throughout the process, and learn from pilot projects. Enabling systems to “talk to one another” and organizing a technology infrastructure to support AI tools are important for implementation.

The participants also raised concerns about the potential for bias and inequity in AI. Strategies to combat these issues include using representative data to feed algorithms, providing ethics training to users, and frequently re-assessing AI tools.

People, as important enablers of technology, need to understand the benefits and limitations of using AI. Getting familiar with a new system requires a robust adoption strategy to ease into using the tools. Moreover, promoting education on AI and selecting tools that align with users’ needs could promote AI use in primary care settings.

Health leaders and government players also play a role in supporting AI implementation. They need to demonstrate commitment to AI and “lead by example.” Developing national standards and creating regulations based on risk level could pave the way for systemic regulation of AI and big data; hence, building trust among patients and providers in using AI in primary care practices.

The study provides critical insights from patients, providers, and leaders about implementing AI in primary care practices. In a world where AI keeps on developing each day, the findings are crucial starting points to understand how providers can use this rapidly changing technology to improve patient care. Future research should engage with larger and more diverse populations to widely implement AI solutions in primary care.

Reference

Authors: Katrina Darcel^1,2, Tara Upshaw MHSc^1,3, Amy Craig-Neil MSc¹, Jillian Macklin MSc^1,2,4,5, Carolyn Steele Gray MA PhD^6,7, Timothy C. Y. Chan PhD⁸, Jennifer Gibson PhD^4,5, Andrew Pinto MD MSc^1,5,9,10*

Original title of paper: Implementing artificial intelligence in Canadian primary care: Barriers and strategies identified through a national deliberative dialogue

Journal: PLOS ONE

DOI: https://doi.org/10.1371/journal.pone.0281733

Affiliations:

1. Upstream Lab, MAP Centre for Urban Health Solutions, Unity Health Toronto, Toronto, Ontario
2. Undergraduate Medical Education, Temerty Faculty of Medicine, University of Toronto, Toronto, Ontario
3. Cumming School of Medicine, University of Calgary, Calgary, Alberta
4. Joint Centre for Bioethics, University of Toronto, Toronto, Ontario
5. Dalla Lana School of Public Heath, University of Toronto, Toronto, Ontario
6. Bridgepoint Collaboratory for Research and Innovation, Lunenfeld-Tanenbaum Research Institute, Sinai Health System, Toronto, Ontario
7. Institute of Health Policy, Management and Evaluation, Dalla Lana School of Public Health, University of Toronto, Toronto, Ontario
8. Department of Mechanical and Industrial Engineering, Faculty of Applied Science and Engineering, University of Toronto, Toronto, Ontario
9. Department of Family and Community Medicine, St. Michael’s Hospital, Toronto, Ontario
10. Department of Family and Community Medicine, Faculty of Medicine, University of Toronto, Toronto, Ontario