REGINA — A new study from the Fraser Institute focuses on the growing use of artificial intelligence in health care.
The study, released last week and titled How Implementing System-Wide Solutions Can Amplify the Impact of Artificial Intelligence on Health Care, concludes that adoption of AI in health care could provide enormous productivity gains, according to its news release. But that would only happen if health-care systems are redesigned with AI at their core, the report adds.
Avi Goldfarb, Rotman Chair in Artificial Intelligence and Healthcare at the Rotman School of Management at the University of Toronto, authored the study. In an email to SaskToday, Goldfarb outlined the motivation behind it.
”AI has potential to transform how we live and work,” said Goldfarb. “Its biggest impact might be to improve our health, but that will only be possible if we figure out how to use it effectively.”
Among the main findings, according to Goldfarb, is how AI is incorporated into applications.
“Many AI applications in health care to date involve ‘point solutions,’ where an AI tool is inserted into an existing workflow,” Goldfarb said.
“I argue that AI will only have a limited impact on improving our health when we change health-care workflows in what I call a ‘system solution’. For example, AI could enable nurses and pharmacists (rather than physicians) to diagnose. “
As one example, the Fraser Institute's news release pointed to an AI “scribe” as a tool already in use to automatically generate clinical notes from physician–patient conversations.
While the report notes this can save doctors time, it states that a system-wide solution could see AI record conversations in the exam room, structure clinical notes, update medical records, arrange follow-ups, co-ordinate with the pharmacy and flag relevant test results.
The report also points to machine-learning tools already in use. One example cited is a “self-driving lab,” where AI is able to design, run and adapt entire experiments in real time without human intervention, which it states could speed up the search for new drugs and treatments.
Goldfarb points to enormous benefits if AI is adopted in health care.
“The optimistic scenario is extraordinary, including faster diagnosis, better treatments, shorter wait times, and lower costs,” he said.
As for key recommendations from the study, Goldfarb points to the following:
“Health-care leaders should look for opportunities for system redesign. With powerful AI diagnosis tools, it should be possible to deliver better patient care with shorter wait times at lower cost. That will only be possible if we do more than insert AI tools into existing health-care workflows.”
The question, according to Goldfarb, is whether health-care leaders are willing to move ahead with such a redesign of the system.
”Achieving such change will require rethinking professional boundaries, payment models and regulation. This may be particularly difficult in Canada’s publicly-funded system.”
