More than 6.5 million Canadians don’t have a family doctor, up from 4.5 million in 2019. As the gap widens, governments are scrambling to find solutions. Quebec is a good example. It has proposed one of the most sweeping health care reforms in a generation. But will it work? With the right use of AI, it might—and there’s a lesson here for all Canadians.

Quebec’s Proposed Reforms

Quebec Health Minister Christian Dubé proposes a system in which only patients with significant health problems, such as chronic illnesses or mental health issues, will be assigned a family doctor. Healthier people will use the services of a new organization, Santé Québec, to help them make doctor’s appointments when needed.

In essence, the Dubé proposal would declare family doctors a scarce resource and establish a triage system to ration access to them. Administrative control of the new doctor-patient relationship would be centralized in Santé Québec.

Doctors and patient organizations don’t like the reforms. They think centralization will weaken local health agencies, making it harder for them to respond to specific community needs and interact with policymakers.

These objections aren’t new. The debate over whether to centralize or decentralize healthcare has been going on for decades. Centralization appeals to decision makers because it gives them a bird’s-eye view of the system, allowing them to make strategic decisions to avoid bottlenecks or align resources with broader goals.

On the other hand, opponents fear that centralization will distance decision-makers from events on the ground, which, in turn, will weaken responsiveness, stifle innovation, and allow inequalities to go unchecked.

In Quebec, this debate is more practical than ideological. Most parties want to find the right balance between the two approaches. The disagreement is mainly over the degree of centralization needed to ensure that triage works well—and that is where AI and data sharing are opening new doors.

AI enables decentralization

Canadian governments recognizes that AI and data sharing are the way of the future. Led by Ottawa, they have already agreed to work together to use these tools to improve services. However, their efforts remain task-specific. 

For example, most provinces are developing electronic health records (EHRs), which consolidate patient information into a single file online. This allows health care providers to coordinate their care, ensuring, for example, that a doctor prescribing medication has access to the patient’s complete medical history.

However, while EHRs can help improve patient care, the data they contain can also be used for other purposes, such as predicting staff shortages, managing patient flow, and identifying underused resources—which in turn can help administrators find a better balance between centralizing and decentralizing the system. Basically, the more accurate, timely, and nuanced the information, the more administrators can use it to fine-tune the system.

Scandinavian countries, such as Denmark, are already doing this. Administrators regularly analyze patient data to help their hospitals and clinics avoid bottlenecks and find ways to improve healthcare delivery across regional networks. This promotes more localized decision-making without fragmenting the national system.

Quebec isn’t there yet. The government still sees its health system as too big, opaque and unresponsive to decentralize a task like triage. It believes that doing so would lead to fragmentation, and so it has decided that the best way to implement the new system is from the top down. 

While there is lots of historical evidence to support this, things are changing. AI is transforming how large, complex systems are managed. Leveraging critical data sources like EHRs allows health administrators to generate an increasingly nuanced picture of system performance in real time. Indeed, with the right data, AI could generate a dynamic model of the inner workings of the system, much like an MRI does for the body. Local health organizations could then use this to help them make decisions that are aligned with the goals of the larger national network.

Next steps

The lesson from countries like Denmark is that the long-standing debate over centralization vs. decentralization is entering a new phase. A truly “networked” approach is becoming a viable option and that could be the game-changer that large, rickety, aging systems like healthcare need to thrive in the future.

That said, there are still hurdles to overcome, particularly in health care. The prevailing view in Canada is that patient data should be used only for direct patient services. Data sharing, whether through EHRs or other sources, is tightly controlled.

Fortunately, new AI techniques like federated learning offer solutions that allow healthcare organizations to collaborate and benefit from each other’s data without compromising patient privacy. However, for a robust data sharing system to take hold, Canada needs a culture shift, supported by greater public understanding of AI and higher levels of data literacy. This is the policy challenge ahead.

At the same time, governments like Quebec's, which are willing to explore radical ideas like rationing family doctors, show that policymakers are increasingly open to new approaches to saving their health care systems.

Quebec is well positioned to take the lead in moving toward AI and data sharing to achieve a better balance between centralization and decentralization. If it succeeds, all  of Canada will be the better for it.

Don Lenihan PhD is an expert in public engagement with a long-standing focus on how digital technologies are transforming societies, governments, and governance. This column appears weekly.