👉 When Housing Becomes Health Care

Kelowna Grapevine | February, 2026 | By Paul

I keep thinking about a “revolving door” in emergency rooms. Not as a metaphor for politics or budgets. A literal cycle that plays out every day: someone comes in sick, gets patched up, then gets discharged back to the same street corner or overcrowded shelter that made them sick in the first place. A few days later, they are back again.

In the health-care world, this cycle has a blunt nickname: the “doom loop.” The point is simple. When people are unhoused and medically fragile, the ER becomes the default safety net. It is not because the ER is the best place to heal. It is because the system often has nowhere else to put them.

Toronto has been testing a different approach, and the early results are hard to ignore. The project is called Dunn House, a four-storey modular building at 90 Dunn Ave. It opened in October 2024 with 51 studio apartments. Each unit is a self-contained home with a kitchen and bathroom, not a shelter cot and a shared hallway. Residents have their own door and their own space.

But the key feature is not the drywall. It is the support. Dunn House is designed for people who were frequent users of the University Health Network’s emergency departments and inpatient beds. The building is tied directly into health care, with on-site services and a care model built around the reality that housing is a medical intervention for some patients, not a “nice-to-have.”

The origin story matters because it shows how this can happen in expensive cities. The land is hospital-owned, and the project moved forward through an unusual partnership between the City of Toronto, University Health Network (UHN), United Way Greater Toronto, and the housing provider Fred Victor. The point is not that Toronto is special. The point is that the ingredients exist in most places, including B.C.: health authorities, city partners, and hospital land that is often used as surface parking.

Now the numbers. Early tracking focused on 48 residents. In the year before they moved in, those 48 people made 1,837 trips to the emergency department. After moving into Dunn House, emergency department visits dropped by 52%.

That is already significant, but the bigger system impact may be what happened to hospital bed days. When these residents did need to be admitted, the total length of their hospital stays dropped by 79%.

That second number gets at something most of us never see: “social admissions.” Sometimes a doctor keeps a patient in a hospital bed longer than medically necessary because discharging them is unsafe. If you are treating a serious infection and your patient has no place to keep a wound clean, no fridge for medication, and no stable shelter, discharge can be a gamble. A supportive home changes that. It lets hospitals use hospital beds for acute care again, instead of using them as the last resort shelter.

There is also a cold financial argument, and it lines up with the human one. Supportive housing runs at roughly $4,000 per person per month. The same individuals can cost many multiples of that when care shifts to hospitals, jail, or emergency shelters. I am not sharing those numbers to score points. I am sharing them because “doing nothing” is not neutral. It is expensive. It is also the option that keeps the doom loop running.

So what does this have to do with Kelowna?

We have our own version of the same pressure points: long ER waits, overworked staff, ambulance offload delays, and a visible crisis on the street. When health care and housing are treated as separate files, patients fall into the cracks. The discharge plan quietly assumes you have an address. When you do not, the entire plan collapses.

Dunn House is a case study in what it looks like to treat housing as a clinical tool for a specific group: medically complex, high-frequency ER users. It is not a complete answer to homelessness, and it is not meant to be. It is a targeted intervention aimed at breaking a very costly loop.

If you want a practical thought experiment for Kelowna, think smaller, not bigger. A four-storey modular building like Toronto’s Dunn House, housing about 50 people, would likely need a footprint in the range of 8,000 to 12,000 square feet. That is roughly the size of two to four standard residential lots near Kelowna General Hospital, where a typical lot is about 3,100 square feet.

Unlike a conventional condo, this kind of building would not require much parking. Most residents do not own vehicles. Parking needs would be limited to a small number of staff and service vehicles, all accommodated on-site. The land impact is measured in houses, not blocks, and the tradeoff is fewer emergency admissions, shorter hospital stays, and less pressure on a system that is already stretched.

Toronto is now doubling down. On January 20, 2026, governments and partners announced Dunn House Phase 2, adding 54 more rent-geared-to-income studio units aimed at at-risk seniors, using modular construction again. That is the tell. Pilot projects usually die quietly. This one is expanding because the early results are strong enough that decision-makers are willing to scale it.

I am not pretending this is simple. It requires coordination, funding, clinical partners, and public tolerance for a different kind of solution. But the logic is straightforward: if a roof plus support can cut ER use and shorten admissions for the highest-need patients, it is not charity. It is system design.

Innovation Spotlight: Could a targeted "Direct Cash Transfer" program for at-risk youth be the key to preventing chronic homelessness in Kelowna? Read how a recent $1,000-a-month experiment in Oregon achieved a 94% housing success rate.