We Don’t Have an Interoperability Problem. We Have Several.
A new CMAJ study by Kappanam et al maps Canada’s health data interoperability landscape province by province, and their heat map tells a story worth sitting with. The provinces that score best share three things: small populations, a single EMR adopted — or at least designated — province-wide, and a unified Hospital Information System. Prince Edward Island is not Ontario. And that, unfortunately, is precisely the problem.
It would be easy to look at those results and conclude that size is destiny — that interoperability is simply harder when you have millions of patients, dozens of health regions, and decades of accumulated digital infrastructure. There is some truth to that. But the more important lesson from the map is structural: the provinces doing well made deliberate, system-level choices. The provinces doing poorly largely didn’t. That is a governance problem, not a geography problem.
The Words We Use Are Getting in Our Way
Before we can fix interoperability, we need to be precise about what we mean. The terms EHR, EMR, and HIS are routinely conflated in policy conversations — even in the academic literature — and that conflation makes the solutions feel murkier than they are.
In Ontario’s context, an EMR (Electronic Medical Record) is the system sitting in a physician’s office: the clinical desktop holding encounter notes, prescriptions, labs, and problem lists. A HIS (Hospital Information System) is the enterprise platform running an acute care facility. And the EHR — the Electronic Health Record — is neither of these. It is the set of provincial assets owned by Ontario Health and the Ministry of Health that allow data to flow between the EMRs in clinics and the HISs in hospitals. The EHR is the connective tissue, not the organ. This is where the fun lies.
This is where the fun lies.
When we treat these three things as interchangeable, we design policy interventions aimed at the wrong layer. And we haven’t even started talking about home care and pharmacy — two critical nodes in the patient journey that are largely invisible in interoperability discussions but deeply consequential to the people moving through our system.
Three Structural Problems That Aren’t Going Away on Their Own
1. EMR Consolidation Is a Multi-Decade Project
Canada’s physicians are, by and large, fully EMR-enabled. This is genuinely good news — and it has taken twenty years and enormous public investment to get here. And our provinces did well to get to where we are now given the landscape in place when they began. The bad news is that this success has created lock-in. Physicians in already-stressed clinical environments will not willingly or happily swap out a familiar system for an unfamiliar one, even if a province pays for it and mandates the transition. The disruption to daily clinic workflow, the necessary reduced patient volume during transition, the retraining burden, the data migration risk — these costs land squarely on individual practitioners in practices that have zero slack capacity to absorb them.
Costs land squarely on practices that have zero capacity to absorb them
This is not a failure of will on physicians’ part. It is a rational response to a system that routinely asks frontline clinicians to absorb the direct and indirect costs of system-level change. The realistic planning horizon for meaningful EMR consolidation — even with provincial coordination and full funding — is measured in decades, not years. Any interoperability strategy that depends on first achieving EMR standardization is a strategy that will never arrive. We have to work with the fragmented landscape we have, not the clean one we wish existed.
2. The HIS Vendor Market Is Fragmented by Design
Canada’s hospital information systems are dominated by three American vendors. That is not a coincidence, and it is not a market failure in the traditional sense — it is the vendors’ revenue model working exactly as intended. Each hospital system, and in many cases each individual hospital, runs its own separate instance of these platforms. The result, as the CMAJ study confirms, is that even within a single city, hospitals cannot reliably exchange data with each other, or with the community physicians trying to care for the same patients.
The vendors have little financial incentive to change this. Their support contracts scale with the number of distinct instances they maintain. Every new hospital system that signs on for its own deployment is a new revenue stream. Provinces that have pushed back against this model — Alberta with its unified Epic instance, PEI with its single-vendor approach across both community and hospital care — have demonstrated that consolidation is achievable. But it requires the kind of deliberate, sustained provincial will that most jurisdictions have not yet mustered.
3. The Data Still Travels by Fax
Perhaps the most damning finding in the CMAJ study is one that should surprise no one who works in clinical care: data exchange between primary care and specialists, and between hospitals and community settings, remains heavily dependent on fax in every single Canadian jurisdiction. Every one. In 2026. We were getting somewhere with ePrescribing, but we all know where that ended up.
This is not a technology problem. We have had the technical capacity to move structured clinical data electronically for years. Alberta’s Connect Care — a single Epic instance now spanning over 650 AHS facilities — and PEI’s province-wide Cerner deployment, integrated with a single community EMR, have both demonstrated that deliberate consolidation is achievable at a provincial scale. It is worth noting that even Connect Care has its limits: community physicians in Alberta are not natively on the system but access it indirectly through the existing Netcare provincial repository. That gap matters, and it illustrates just how hard the final mile of integration remains — even in the best Canadian examples.
The problem is that the conditions required for automation — common data standards, clear governance, aligned incentives, and shared infrastructure — do not yet exist in most of Canada. And because they don’t, the movement of information from one care setting to another remains dependent on individual practitioners making it happen, encounter by encounter, fax by fax.
Encounter by encounter, fax by fax
That is not a sustainable model. It is a model that creates gaps, delays, duplications, and harm — and it places the burden of those failures on the very patients the system is meant to serve.
The Path Forward Requires More Than Good Intentions
The CMAJ authors call for national legislation, strengthened governance, and aligned incentives. They are right on all counts. But the devil — as always — is in the details.
The devil — as always — is in the details.
Standards must be mandated, not voluntary. Voluntary interoperability has been Canada’s default for twenty-five years, and the heat map shows us exactly what that has produced. The Digital Health Interoperability Task Force — convened in 2024 by the Canadian Medical Association and Canada Health Infoway, on which I served — called for a five-year national health data interoperability plan and was explicit that standards without enforcement are suggestions, not infrastructure. Bill S-5, the Connected Care for Canadians Act, reintroduced in the Senate in February 2026, is an important step — it addresses vendor data blocking and establishes baseline interoperability requirements. But it is not the finish line. Federal legislation sets a floor; provinces must build the walls and the roof. Health care delivery is, like it or not, still primarily a provincial endeavour, and provincial governments need to pick up where federal legislation leaves off with their own complementary regulatory action.
Workflow automation is not a luxury — it is the precondition for uptake. The movement of clinical data between systems cannot depend on clinicians manually initiating transfers at the point of care. Automation must be built into the infrastructure, with data following the patient as a default rather than as an exception. The safeguards for privacy and security must be embedded from the beginning — not bolted on afterward — and responsibility for errors must be shared across the system, not placed solely on the backs of physicians and nurses already operating at capacity.
The incentive structure has to change. The CMAJ study found that providers receive no compensation for meaningful use and exchange of clinical data. That needs to change. If we want interoperable workflows, we have to make interoperability worth a clinician’s time. And if we want vendors to build toward open standards rather than proprietary lock-in, we have to make that a clear condition of doing business in Canada — not a checkbox in a procurement document. As well, currently even if we have interoperability via data standards and data exchanges, the infrastructure for where that data will actually go, and the guardrails and governance around “for what reason” have yet to be built.
What Actually Has to Land
The Kannappan et al. study estimates that disconnected health data costs Canadian taxpayers more than $9.4 billion annually. That is not an abstraction. That is the price we pay, year after year, for a system that cannot reliably tell your emergency department physician what medications your family doctor prescribed last month.
The solutions are known. They are not easy, but they are not mysterious. A common data and communication standard — HL7 FHIR (Fast Healthcare Interoperability Resources) is the international consensus choice, and Canadian EMR vendors are at varying but accelerating stages of adoption — needs to be legislated as the mandatory exchange standard and enforced as a condition of operating in publicly funded health care. Provincial EHR layers need to be resourced and governed with the seriousness they deserve. Data movement for specific reasons, not just for movement’s sake, has to be outlined. Vendor practices that block data portability need to be prohibited and penalized. And the clinicians who have been asked to absorb the costs of every previous wave of digital health change need to be brought into the design of this one, not handed another implementation on top of a full clinical day.
The CMAJ heat map shows us where we are. We already know where we need to go. The question — as it has been for the better part of two decades — is whether we have the political (both big “P” and small “p” political) will to close the distance between those two points.
That is not a technology question. It never was.
Credit for all embedded graphics goes to the brilliant graphic arts folks like Hugh Macleod at gapingvoid.com.
