Class of 2023: Project offers a glimpse of how to make artificial intelligence succeed in the health-care system

Integrating new technology is a disruptive puzzle for many organizations. For a year, Melanie King had a front-row seat as she researched how a group of health-care professionals learned to apply a new artificial intelligence tool in their work — and succeeded. How did they do it?

King’s search for answers helped her earn a Doctor of Education from the Werklund School of Education in the Class of 2023. And her research project changed the trajectory of her career at the University of Calgary, where she’s now a director.

Along the way, she overturned misconceptions that AI replaces people and takes the human factor out of health care. “If anything,” she says, “we’re seeing how the tools we can create with AI give us the ability to be even more personalized in medicine.”

Curious about personal experiences of health care

King developed a curiosity about personal perspectives of the health-care experience early in her career. As a nurse, King did stints in Africa and northern Canada, where she noticed that the context and environment of each community profoundly affected the lives and experiences of patients and health-care workers alike.

King completed a master’s degree in nursing (2016) from UCalgary, then went on to manage programs at the university that help people change how they conduct health research. Her work nurtured another deep curiosity: how learning takes place alongside the significant changes seen by health practitioners throughout their careers.

Sparked by her growing interest in the role of data in health care, King continued her academic journey, enrolling in the Doctor of Education program at Werklund. “Everybody says, study something you’re interested in, or that aligns with your work,” she explains. That’s exactly what she did.

AI tool translates reams of data

The AI tool that would become a focus of King’s case-study research is a cardiac registry, a powerful database that combines and compares information entered from multiple sources to help doctors predict and prevent cardiac disease and provide more personalized care. AI can quickly and accurately translate reams of data, using computational algorithms to flag patient risk factors or other patterns that may be missed by manual methods.

King noticed that the role of learning in workplace adoption of AI within health care had received little scholarly attention so she moved ahead with her own transdisciplinary research. King’s twin areas of interest — social context and learning during a time of change — informed her qualitative research approach. “And I was genuinely curious and interested in the role of data to improve patient outcomes,” she says.

In her doctoral thesis, King explored the perspectives of 16 health-care professionals who had administrative, project or leadership roles in implementing the registry. The study identified four overall themes of impact among the participants: the people involved, the environment where the change takes place, the design of the innovation, and how participants learned new processes and skills.

‘People play a central role’

The single ingredient most critical to the success of the registry project was the people, especially the project champions. “Findings from the study revealed that people play a central role in the success of a change event and project implementation.” Her research insights will be useful in the future for health-care leaders implementing similar projects in complex settings.  

She credits her supervisor, Dr. Jennifer Lock, PhD, vice-dean at the Werklund School of Education, for her mentorship, advice, and “boundless positivity” during her research.

King’s PhD studies and her career merged about three years ago, when the Libin Cardiovascular Institute hired her as its director of program development and implementation for the Precision Medicine Initiative. In the role, she facilitates integration, governance, and analysis of data for cardiovascular research, in collaboration with the Department of Cardiac Sciences, the Centre for Health Informatics, and the Mozell Family Analysis Core at the Cumming School of Medicine.

“If we can impact patient care by creating these tools, for example, identify a patient in the ER who has an elevated risk for a complication or severe outcome and intervene before that happens, I think we have a moral obligation to pursue this kind of work.”