opinion: the next ai breakthrough in health will be better connected care
5 minute read
by living quietly inside the electronic medical records (emr), pharmacy solutions, claims and virtual care platforms teams already rely on, ai can practically reduce administrative burden, close coordination gaps and support continuous patient care.
adobe stock
stay up to date on this topic
last year, much of the conversation around health-care ai focused on the efficiencies gained through solutions that support documentation, such as ai scribes.
that made sense. for clinicians buried in paperwork, ai scribes are one of the first digital tools to show real value in optimizing workflows. they reduce typing, speed up note-taking and help doctors spend more time focused on the person in front of them.
that matters. but it’s only a first step in fixing a system where patients often feel like a data point rather than a person.
patients don’t experience health care as data entry. they experience it as a connected journey and care about how smoothly their health information moves with them. they want to know whether a medication change is reflected across a care team, whether a referral goes out on time, and whether their pharmacist, physician and broader care team are working from the same chart.
that is where the next opportunity for ai lives.
canada will not realize the real benefits of health-care ai by layering more standalone tools onto already fragmented systems. the next phase must be seamless and invisible, with ai embedded directly into the daily workflows of clinicians, pharmacists and care teams. success means no separate tools, no heavy training, and no added steps to the workflow. by living quietly inside the electronic medical records (emr), pharmacy solutions, claims and virtual care platforms teams already rely on, ai can practically reduce administrative burden, close coordination gaps and support continuous patient care.
the next step is coordination
the burden on care teams remains enormous. canadian physicians still report spending roughly nine hours a week on administrative work, and nearly half of those tasks are considered unnecessary. those tasks add up to 19.8 million hours a year, almost the equivalent of 9,100 full-time physicians. every hour a clinician spends on administration is an hour taken away from patient care. at the same time, two in five general practitioners and two in three specialists already report using ai tools to help with tasks.
advertisement
in other words, clinicians are not waiting to see whether ai can help. the real question is what kind of ai will actually improve care by shifting the focus back to the patient journey.
the best health-care ai should feel less like a new application and more like a natural extension of the workflow. it should anticipate the next administrative step, surface relevant context, and reduce effort without forcing clinicians or staff to stop, switch tools or relearn how to work.
that means surfacing the right patient context before a visit, so the patient does not have to re-explain the details of why they are seeking care, matching incoming documents to the right chart, drafting a referral letter for clinician review, or routing follow-up tasks to the right team member instead of leaving them in limbo, creating longer wait times for patients to receive the care they need.
a recent telus health 2026 ai discussion paper makes this point clearly: the most useful applications are not flashy or autonomous. they are practical. a patient shouldn’t have to wait for documentation to manually move from one inbox to another; they deserve a system where their care moves as fast as their lives do. at telus health, ai within an emr is implemented with this in mind. an ai-powered inbox triage automatically matches incoming documents to patients in an average of 1.2 seconds, reducing manual routing time. concurrently, referral letter agents draft suggested referral letters and titles directly within the emr during patient-clinician interactions for immediate clinician review.
these applications are useful because they reduce cognitive load, close gaps in coordination and help care teams stay on top of what matters, without taking control away from them. they also work because they are embedded where the work already happens. that matters. in health care, adoption depends not only on what technology can do, but on how naturally it fits into a busy clinical environment. if a tool requires too much training, too many clicks or too much workflow change, it will struggle to deliver value at scale.
advertisement
this is especially important in a health-care system like canada’s, where fragmentation is operational, not solely technical. information often sits in different systems, provinces and care settings. clinicians and staff are left to bridge those gaps manually, which is inefficient for doctors and frustrating for patients, who are left attempting to navigate an often confusing system on their own.
done well, better integration can change that. it can help care teams catch issues earlier, improve continuity and deliver more effective care across the patient journey. it can also help break down geographic silos, so broader patterns and diverse use cases inform how care is delivered, not just how it is recorded.
technology should support care teams, not replace them
health care does not need ai making autonomous decisions in high-risk situations. it needs ai handling the narrow, repeatable, time-consuming tasks that pull clinicians away from care. human oversight, trust and clear boundaries have to remain non-negotiable.
that is not a limitation. it is what makes these tools workable in the real world. the most successful healthcare ai will be practical, supervised and easy to use. it should not ask care teams to adapt to the technology. the technology should adapt to them.
this is also why interoperability matters so much. it is not only a technical goal, but a care-delivery issue. if ai cannot connect the dots across emrs, pharmacy, claims and acute care facilities such as hospitals, it will never deliver more than incremental improvements.
the real promise of health-care ai is not faster documentation for its own sake. it is better coordinated care and fewer missed handoffs. it is less time spent navigating systems and more time spent using clinical judgment where it matters most. ultimately, it ensures that technology handles the data entry, and clinicians focus on the human experience.
advertisement
canada has a real opportunity. but we will only realize it if we stop treating ai as a series of disconnected features and start embedding it where care actually happens.
for care teams, meaningful adoption means ai that is seamless, convenient and trusted enough to become part of the natural flow of work. for patients, it means a system that feels more connected, less repetitive and easier to navigate.
ratcho batchvarov is vice president of provider solutions at telus health.
share story
