The growing interest among healthcare professionals in incorporating AI tools into their practice has increased the need for provider and patient-centered approaches that prioritize the responsible implementation of AI in clinical settings.
By: Marielle Boutin
To address this issue, IHPME’s HIVE Lab is developing a global research network dedicated to promoting patient, caregiver, and clinician-centered AI implementation practices.
The initiative, “Advancing Global Collaboration for the Responsible and Human-Centered Deployment of Artificial Intelligence in Healthcare,” is a collaboration between HIVE Lab, University College London and other international partners and is supported by a Horizon Europe Partnership Development Grant from U of T.
Collaborators on the project include IHPME faculty members Dr. Zahra Shakeri – Director of HIVE Lab and project co-lead, and Dr. Jeremy Petch – founder of CREATE and the Director of Digital Health Innovation at Hamilton Health Sciences. Also co-leading this effort is Dr. Susanne Gaube, Assistant Professor in Human Factors in Healthcare at UCL’s Global Business School for Health (GBSH).
In healthcare settings, AI models are actively being explored to support administrative tasks, disease detection, image labelling, and health outcome predictions. Despite remarkable technical advancements over the past several years, health systems continue to struggle with integrating these tools into everyday care. Unpredictability and error rates are a continued concern among stakeholders, as it can increase the risk of care denial, stigma, misinformation, and health disparities.
Healthcare professionals are expected to critically evaluate AI outputs, oftentimes without knowing how the models work, highlighting the need for patient-informed support, such as the new Global Collaboration.
As noted by Dr. Shakeri, the successful implementation of AI in healthcare must put the experiences of clinicians and patients above the technology itself.
“People, not algorithms, decide whether AI succeeds at the bedside,” says Dr. Shakeri. “Clinicians already juggle complex data and tight timelines and patients rightly demand tools that respect their dignity and privacy. Our research starts by engaging these stakeholders, understanding their needs, and identifying foundational challenges that will influence the successful use of these tools, such as the ability of clinicians to provide effective oversight of these tools.”
This approach also reflects HIVE Lab’s mandate to create transparent, inclusive, and data-informed tools that advance public health and help reduce health disparities.
Throughout the project timeline, researchers will come together through virtual meetings, partnerships, international events, and roundtable discussions to discover how people, systems, and technology have and continue to shape the use of AI in healthcare with the goal of lowering risks, reducing bias, and building trust in AI-assisted care.
According to Konrad Samsel, Project Manager for HIVE Lab, the project fosters cross-disciplinary and international collaboration by bringing together diverse voices to ensure outputs are actionable for both academics and clinical practitioners.
“Health systems continue to navigate challenges with clinical AI deployment, and we expect our partnerships and future research to play a key role in guiding best practices,” says Samsel. “While our long-term goal is to grow this into a larger international effort, we also want to spark broader academic interest in deployment research and show how collaboration across disciplines is needed to advance the sustainable adoption of AI in healthcare.”
One of the key barriers faced, as Dr. Petch mentions, is integration, notably translating lab-developed research into systems that can be used in a complex clinical setting.
“I often see really interesting AI research that struggles when it comes time to implement, because it was developed in a university lab, so hasn’t been developed to integrate smoothly into clinical practice,” says Dt. Petch. “Sometimes that’s because of technical integration problems – the AI system hasn’t been designed to interoperate with a hospital’s EHR – but just as often research teams haven’t considered how their tools will integrate smoothly into clinical workflows or complex clinical pathways. Very often the result is really exciting AI research that can’t be easily adopted in the real world.”
To curb this challenge, health leaders must continue to build capacity to manage and monitor these tools.
“AI is still a very new technology when not managed effectively it can pose significant risks. Whether through development of their own resources or strategic partnerships with universities and AI centres, health system leaders are going to need to ensure that they can implement AI tools in a safe and effective manner,” says Dr. Petch.
As the interest in AI in healthcare continues to grow, initiatives like the Global Collaboration are essential to ensuring the lived experiences of patients and caregivers are central to these new systems. Through international collaborations and human-centred design, this is just one step to developing equitable and sustainable AI solutions.
Sign up for IHPME Connect.
Keep up to date with IHPME’s News & Research, Events & Program, Recognition, e-newsletter.
Subscribe to Connect Newsletter
Get in Contact
Communications
Marielle Boutin
Email Address: ihpme.communications@utoronto.ca