Healthcare Operating Model Design: Why Architecture Is the Missing Ingredient
Digital Health Canada documented 175 AI initiatives across Canadian health systems in 2025. Sixty-five percent of Canadian hospitals had implemented AI-driven predictive analytics in their EHRs by 2024. For large hospitals in Ontario and British Columbia with more than 400 beds, adoption reached 92 percent. The Canadian healthcare system is not short of technology investment or AI enthusiasm.
It remains deeply fragmented.
Healthcare inefficiency does not stem from a lack of effort or a lack of technology. It stems from fragmentation: handoffs between departments, between providers, between payers, and between platforms that do not share a coherent view of the patient or a shared understanding of who is accountable for what at each stage of the care journey. A 2026 analysis from Digital Health News put this precisely: AI that can manage these transitions, flag inconsistencies, and surface relevant information at the right moment delivers disproportionate value precisely because the fragmentation is so structural and so costly.
Operating model design is the discipline that addresses fragmentation at its source. Not by deploying another technology layer on top of a fragmented foundation, but by designing the organizational structures, decision rights, accountability mechanisms, and care pathway logic that allow clinical and administrative functions to work as a coherent system rather than as a collection of independent units that happen to share the same building.
Why Healthcare Is a Distinctive Operating Model Challenge
The operating model design principles described in the general posts in this series apply in healthcare as they apply in any sector. The specific conditions of healthcare, particularly in the Canadian publicly funded system, create a version of the operating model challenge that differs from what most other sectors face in ways that are worth being explicit about.
Multiple Governance Structures With Overlapping Authority
A typical large Canadian hospital or health authority operates under federal funding obligations, provincial health ministry direction, regional health authority governance, professional college regulations for each clinical discipline, accreditation standards from Accreditation Canada, and the governance of its own board. Each of these structures has legitimate authority over some aspect of how the organization operates. None of them has unified authority over how the organization operates as a whole.
The practical consequence is that operating model decisions in healthcare require navigating a governance landscape that most other sectors do not encounter. A hospital that wants to redesign its patient flow operating model to reduce emergency department wait times needs to align changes with nursing union agreements, physician privilege structures, health authority performance frameworks, ministry reporting requirements, and its own board's strategic priorities, all of which may have different views about what the right operating model looks like. The operating model design work is as much a governance navigation problem as it is an organizational design problem.
The Clinical-Administrative Interface Is Poorly Designed in Most Institutions
The most consistent source of operating model dysfunction in health systems is the interface between clinical and administrative functions. Clinicians are trained to focus on patient outcomes and make decisions in real time with incomplete information. Administrative functions are designed to manage resources, comply with reporting requirements, and maintain financial sustainability. These two orientations are both necessary and in structural tension: a clinician's appropriate focus on the individual patient in front of them is in inherent tension with an administrator's appropriate focus on the population of patients across the system.
Most Canadian health systems have not deliberately designed the interface between these two orientations. They have allowed it to develop organically, producing an accumulated set of informal norms, workarounds, and escalation paths that are not documented, not consistently applied, and not optimized for either clinical outcomes or administrative efficiency. The result is that both clinicians and administrators spend significant time managing the interface between their functions rather than doing the work those functions exist to do.
Research published in the Journal of Medical Internet Research in 2024 identified this pattern specifically: coordination and communication gaps occur due to synchronous and asynchronous interactions through healthcare information systems, resulting in miscommunication and gaps in coordination among healthcare professionals. The extensive interaction with multiple healthcare systems increases the digital workload and contributes to clinician burnout. The technology is not causing the burnout. The poorly designed operating model that surrounds the technology is causing it, and the technology is making the dysfunction more visible by adding a digital interface to every step of an already fragmented process.
Care Pathways Cut Across Organizational Boundaries
A patient's care pathway for a chronic condition, cancer treatment, mental health support, or post-surgical recovery typically involves multiple organizations: primary care, specialist care, hospital-based acute care, diagnostic services, home care, and community support services. Each of these organizations has its own operating model, its own information systems, its own performance metrics, and its own accountability structure. None of them is accountable for the patient's experience of the pathway as a whole.
The patient experiences the pathway as a single journey. The system delivers it as a series of disconnected episodes. The discontinuities between episodes, the referrals that are not followed up, the discharge summaries that do not reach the primary care provider, the medication changes that are not communicated across the care team, are where most preventable harm and most unnecessary cost accumulate in the Canadian health system.
Designing operating models that span these organizational boundaries requires a different kind of architecture work than designing within a single organization. It requires agreement on shared accountabilities across organizational lines, standardized hand-off protocols that different organizations are willing to commit to, information sharing arrangements that navigate the provincial privacy legislation governing health information in each jurisdiction, and governance structures for the pathway as a whole that do not override the governance of the individual organizations involved but create accountability for the transitions between them.
The Three Operating Model Decisions That Most Health Systems Have Not Made Explicitly
Most Canadian health system operating model challenges can be traced to a small number of foundational decisions that have either never been made explicitly or have been made differently by different parts of the organization, producing structural inconsistency that compounds into the fragmentation observable at the patient and clinician level.
Who Is Accountable for the Care Pathway, Not Just the Episode
The most fundamental operating model decision in healthcare is whether anyone is accountable for a patient's care pathway across the full continuum of care, or whether accountability is segmented by care setting with the gaps between settings belonging to no one.
Most Canadian health systems operate in the second mode by default. The hospital is accountable for what happens during the admission. The primary care provider is accountable for what happens in their practice. The home care agency is accountable for what happens in the home. Nobody is explicitly accountable for whether the transitions between these settings happen correctly, whether the information that needs to flow between them flows accurately, or whether the patient's experience of moving through the system is coherent and safe.
The explicit operating model decision here is to assign pathway-level accountability to a defined role for defined patient populations. This is the operating model underpinning of primary care reform, of integrated care networks, of bundled payment models, and of population health management programs. The technology platform often gets designed before this accountability question is answered, which is why most care coordination technology deployments underperform: the information system is designed to support a care coordination model that the operating model has not yet defined.
How Clinical and Resource Allocation Decisions Are Made and by Whom
The most contentious operating model question in healthcare is the allocation of decision rights between clinical and administrative authority. Clinicians reasonably assert that clinical decisions should be made by clinicians. Administrators reasonably assert that resource allocation decisions require administrative oversight. The tension is real and is not resolved by asserting the primacy of either perspective.
The operating model design question is not which perspective is right. It is how to design a decision architecture that allows clinical decisions to be made by clinicians within resource constraints that are transparent and consistently applied, and that allows resource allocation decisions to be made with sufficient clinical input to avoid decisions that generate false economies by shifting cost from visible budget lines to invisible clinical outcomes.
Health systems that have designed this interface explicitly, through structured forums for joint clinical-administrative decision-making, through transparent data about the cost and quality implications of different clinical practice patterns, and through governance mechanisms that create accountability for both clinical outcomes and resource utilization, consistently outperform those where the interface is managed through informal negotiation and political escalation.
What Technology Enables Versus What Technology Replaces
The third foundational operating model decision is the most immediately relevant in 2026, given the pace of AI and digital health investment: what organizational capabilities and human roles does technology enable at higher performance, and what does it replace, and who makes that decision and how?
Most health system digital transformation programs have answered this question implicitly rather than explicitly, by deploying technology and then assessing what changed. The implicit approach has produced the AI-driven efficiencies documented in the research alongside the clinician burnout documented in the same research, because the technology was deployed into workflows that were designed for human execution and the operating model around those workflows was not redesigned to take advantage of what the technology made possible.
The explicit approach defines the target operating model before the technology is selected: what does this workflow need to accomplish for patients and for the organization, what are the human capabilities that are essential to that accomplishment, what are the tasks within the workflow that technology can perform more reliably and efficiently than humans, and what does the workflow look like when it is redesigned around that division of labor? That sequencing produces technology deployments that reduce administrative burden and improve decision support rather than technology deployments that add digital workload to already overburdened clinical staff.
Where Operating Model Design Produces the Most Immediate Return in Healthcare
Three specific application areas consistently produce the most immediate and measurable return from operating model design work in Canadian health systems, based on the pattern of documented outcomes in the sector.
Emergency department and patient flow. ED overcrowding and access block are operating model problems. The bottleneck is rarely the emergency department itself. It is almost always a downstream system constraint, insufficient acute care beds, insufficient home care capacity to support discharge, insufficient primary care capacity to divert non-urgent presentations, that produces congestion in the ED because the ED is the end of the accessible pathway for patients who have no other entry point. An operating model that addresses only the ED will not solve the access problem. One that maps the full patient flow pathway from presentation through discharge through post-acute care, assigns accountability for each transition, and designs the interfaces between them can produce measurable improvements in patient wait times, length of stay, and staff experience simultaneously.
AI and digital health governance. Health Canada's 2025 to 2026 Departmental Plan identifies interoperability and digital modernization as central policy objectives. CIHI published its AI strategic foundation in March 2026 with explicit emphasis on ethical, transparent, and responsible AI implementation. Canada's AIDA legislation, advancing through Parliament, will create compliance obligations for high-impact AI systems used in healthcare. The operating model question for health systems investing in AI is not what AI tools to deploy. It is what governance structure ensures that AI deployments are evaluated consistently, that high-impact AI systems are subject to appropriate oversight, that compliance obligations are met, and that the clinical and administrative stakeholders who need to trust AI-generated insights have the context they need to evaluate them appropriately.
Clinical service line integration. Most Canadian health systems have clinical service lines, programs, or centres that operate with significant clinical autonomy and limited administrative integration. The operating model for clinical service lines, the governance structure for how service line leadership relates to administrative authority, how service line performance is measured, and how resources are allocated within and across service lines, is often underdeveloped relative to the clinical capabilities within those service lines. Operating model design that clarifies service line accountability, standardizes the performance framework across service lines, and designs the administrative support model for service line leadership releases clinical capacity that is currently absorbed by administrative navigation and escalation work.
The Practical Entry Point
Healthcare operating model design at the enterprise level is a multi-year program. The practical entry point is a specific problem that is causing visible pain for a specific leader who has the authority and motivation to address it.
The most productive entry points in Canadian health systems in 2026 are the problems that are simultaneously clinical, administrative, and technological in nature: the ones where a technology investment has been made, the technology is working as designed, and the intended improvement in patient outcomes or operational efficiency has not materialized because the operating model around the technology was not redesigned to realize what the technology makes possible. These problems are common, they are increasingly visible as health system AI investments multiply, and they are problems that neither clinical leadership nor IT leadership can solve alone because the root cause is the operating model that sits between them.
A focused operating model diagnostic in the specific care area or administrative domain where this pattern is most acute, scoped to sixty to ninety days and designed to produce specific operating model design recommendations rather than a comprehensive transformation roadmap, produces both the organizational understanding of what is actually causing the problem and the credibility to design and implement the changes that will address it.
Talk to Us
ClarityArc works with healthcare organizations on operating model design, capability mapping, and the governance architecture required to make digital health and AI investments produce their intended outcomes. If your health system is navigating transformation and the operating model is not keeping pace with the technology, we are ready to help.
Get in Touch