As Canadian healthcare moves toward integrated care, it lacks “North Star” metrics, relying instead on hospital-centric indicators that miss the patient experience. Building on the integrated care framework by Anne Wojtak and Jodeme Goldhar, this paper argues for shifting from measuring system activity to measuring the burden placed on patients and caregivers. It proposes three new metrics: the System Friction Ratio, the Caregiver Latency Index and the Patient-Reported Burden Score. To achieve true population health, we must stop counting hospital-focused administrative outputs and start measuring the friction patients and families face when navigating the system.

Introduction

In the corporate world, a leadership team disciplined by an overarching strategy champions a “North Star” results-based metric intrinsic to its goals. For a software company, it might be annual recurring revenue or profit per employee. For a retailer, it can be revenue per square metre.

It is the single number that cuts through the noise, the one dial on the dashboard that tells you, unequivocally, if the machine is working.

In Canadian healthcare, seeking the promised land of “integrated care,” we are still searching for our North Star metrics. We are drowning in data but starving for clarity on integration indicators that matter. We count what is easy to count, not necessarily what matters most to integration.

Ask hospital administrators how integration is going, and they will likely point to well-established continuity-of-care metrics: readmission rates, discharge times or the number of hand-offs to home care. These are the “streetlight” metrics – we look there because that is where the light is. The data reside in the hospital information system, so that is what we measure.

But as Anne Wojtak and Jodeme Goldhar contend in their new framework for the current issue of Healthcare Quarterly, integration is not a single construct (Wojtak and Goldhar 2026). It is a continuum. It starts with condition-based pathways (fixing a hip), moves to multi-condition care for the chronically ill and, ultimately, aims for population-based health.

Our measurement disconnect is stark: we are trying to manage a population-based transformation using condition-based metrics. If our goal is merely to fix hips faster, then counting discharge times is fine. But if our goal is true population health, then our current metrics may be misleading. They measure the system's activity, not the reality of the people it serves.

As Wojtak and Goldhar note, “population health cannot be advanced without integration as a foundational enabler” (Wojtak and Goldhar 2026). Yet we lack the metrics to tell us if that foundation is actually being built.

We need a new “One Metric.” A metric that does not just tell us if the gears are turning, but if the machine is actually serving its purpose. How should we think about this One Metric?

The Trap of the “Euphemism Economy”

In search of better integration metrics, we must also be wary of what Robert W. Marotta calls the “euphemism economy” that has taken root in the sector (Marotta 2026). We use terms like “Alternate Level of Care” (ALC) to describe patients who are essentially stranded. We track ALC days as a metric of efficiency, but we rarely ask the deeper question: Why is the patient there? Is it because the “integrated” system failed to provide a safe landing spot in the community?

When we rely on euphemisms, we obscure the reality of the patient experience. We need metrics that cut through the jargon and expose the friction in the system.

The “Rainbow” of Complexity

Academic frameworks, such as the widely cited “Rainbow Model of Integrated Care,” correctly identify that integration needs to occur across multiple levels – from the clinical and professional to the organizational and normative (Valentijn et al. 2013). However, systematic reviews reveal a persistent trap: we excel at measuring the structures and processes of these layers – signed agreements and shared information technology systems – but consistently fail to measure the outcomes attributable to them (Strandberg-Larsen and Krasnik 2009).

This complexity spawns dashboards with myriad indicators that track the mechanics of the system rather than the reality of the patient. We need to sift through this “rainbow” of process metrics to isolate the singular measure of impact at the population level. If the goal is normative integration – shared values and culture – the ultimate test is not whether the organizations are talking to one another volubly and frequently, but whether the patient suffers the friction of their silence.

The “System Friction Ratio”

My proposal for a North Star metric is the System Friction Ratio. It asks: for every hour of clinical care received, how many minutes did the patient or caregiver spend managing that care?

• The numerator: Time spent scheduling, travelling, waiting on hold, filling out duplicate forms and coordinating between providers.
• The denominator: Time spent in actual receipt of care or support.
• In a strongly integrated system, this ratio approaches a low baseline. In our current system, for a complex patient – the “multi-condition” patient described by Wojtak and Goldhar (2026) – it is often 1:1 or worse. If a daughter spends three hours on the phone to arrange one hour of home care for her mother, the System Friction Ratio is 3. That suggests a failure of integration, regardless of whether the home care worker eventually showed up and delivered quality care.

This metric forces us to see the “invisible work” of caregiving. It acknowledges that the caregiver is a critical node in the network – often the only node that connects the others.

The “Caregiver Latency Index”

For a second, more operational metric, let us look at the Caregiver Latency Index.

Traditional wait times measure the time from referral to appointment. But from a population health perspective, the clock starts ticking the moment the need arises, not when paperwork gets filed.

The Caregiver Latency Index would measure the time gap between a caregiver identifying a crisis (e.g., “Mom is wandering at night”) and the system providing a stabilizing intervention.

In a fragmented system, this latency is measured in weeks of stress and emergency department visits. In an integrated system, this latency should be measured in hours.

Listening to the System

Finally, we need to listen – literally. We should be auditing the “voice” of our primary care system. Do we track the outbound messages on the answering machines of family doctors? Do they say, “If this is an emergency, go to the ER,” or do they offer a warm hand-off to an integrated partner?

And do we track the actual hours of service provided by primary care in a given geography, versus the theoretical hours? These are the granular, unglamorous metrics that reveal the truth about access and integration.

The “Patient-Reported Burden Score”

As my colleague Zayna Khayat has suggested, we can also leverage artificial intelligence and survey systems to capture the patient's voice directly (Z. Khayat, personal communication, January 23, 2026). She proposes a Patient-Reported Burden Score, sampled quarterly for high-usage patients (i.e., patients with services billed to provincial insurance more than four times monthly). It would ask questions like: “On a scale of 1–10, how confident are you that the people caring for you understand your story?”; “Do you feel like you are the integrator of your care, or the system is?”

This aligns well with the need to measure the experience of integration, not just the mechanics. If the patient feels like the integrator, the system has failed, no matter what the readmission rates say.

Measuring What Matters for Integration

If we keep measuring hospital activity, we will keep building a system optimized for hospitals.

It is time to stop looking under the streetlight. We need to walk into the dark, messy reality of the patients' and caregivers' lives and start measuring the friction they face.

We need to measure the burden we place on their shoulders. We need to measure whether they are living their lives or managing the system with insufficient supports.

That is the “One Metric” that proves integration is working.

Comments

Be the first to comment on this!