Last year, OpenEvidence, a medical search engine founded by Canadian entrepreneur Daniel Nadler saw usage from more than 40% of physicians in the United States.

The power and promise of AI tools are undisputed.

But how do we ensure that the technology supports users with region-specific needs? In Canada, it is important that AI tools are developed including the institutions, health services, and clinical guidelines specific to the Canadian healthcare landscape.

The need to build national data assets to power AI health solutions is highlighted by vital services offered by agencies such as Ontario Health, Alberta Health Services, or Quebec’s INNESS. Patients across the country depend on these agencies for services such as reimbursement for medications, receiving specialized treatment like IV infusions at home, or placement in long-term care facilities.

Without prior experience with these services, patients may struggle to navigate the healthcare system.

In addition to its distinct healthcare institutions, Canada also produces its own clinical guidelines. Last year, researchers published the latest recommendations on blood pressure management highlighting the specific needs of Canadians. These recommendations were developed with consideration towards access, affordability and shortage risk of medications within the Canadian context.

Creating data assets built on top of the unique aspects of Canadian healthcare practice can shape AI tools to better support Canadians.

This effort was put into practice earlier this year, when the Centre for Effective Practice (CEP), a not-for-profit tasked with shaping better clinical decisions, partnered with Shoppers Drug Mart on PC ChatTM — a consumer-facing AI tool providing Canadians across the country a convenient place to ask health questions and get trusted plain-language information in real time. Understanding that health questions asked by Canadians require Canadian-specific guidance, CEP curated an evidence library that prioritized high-quality Canadian resources that were current, credible and contextualized to every province and territory.

The hope is that through this public-private partnership, Canadians will receive better guidance and answers when navigating their health or the healthcare system.

Public-private partnerships may be the best strategic option when building secure and trusted AI health solutions.

For example, AI scribes, which record and transcribe appointment interactions, help clinicians spend less time taking notes and dedicate more attention to the patient in front of them. As part of implementation, the companies behind the technology partnered with Canada Health Infoway, a not-for-profit institution. This partnership led to the launch of the AI Scribe Program, which provides clinicians with temporary funding for pre-qualified AI scribe tools.

For entrepreneurs and innovators developing AI health solutions, sourcing rigorously vetted data from a trusted authority is a necessary step. Not-for-profit organizations are trusted with a public mandate and their role as data stewards can protect the public from the dangers of accelerating AI development without independent oversight.

But creating high-quality national data assets requires significant investment. Not-for-profits are facing large mandates with a tight budget. To build and support the integration of Canadian-specific health approaches, governments and corporations must invest in the efforts of these institutions to create the building blocks that encompass the needs of all Canadians when it comes to AI — including rural, Indigenous, and demographically diverse populations.

AI in healthcare is advancing quickly, but the importance of quality and relevancy cannot be ignored. Within the healthcare context, where there is little margin for error, innovation must be coupled with the development of robust national data assets.

Comments

Be the first to comment on this!