did housing first participants still do better than usual care seven years later?
“Even seven years later those that were in the house and first did so much better than those that got usual care”
a guest, a psychiatrist and senior scientist at CAMH said on PsychEd: Educational Psychiatry Podcast, 2026-09-01, at 26:23. hear it in context → the speaker cited: the At Home/Chez Soi study.
the episode: PsychEd Episode 78: Mental Health and Homelessness with Dr. Vicky Stergiopoulos (at 26:23). we link the show's own page; we do not re-host audio.
first, the part that holds
the core of this is true. in the toronto arm of the at home/chez soi trial, people randomized to housing first were still stably housed on more days than the usual-care group six years in, and the gap was largest for people with high support needs.
what we found when we went and looked
the toronto extension study (lancet psychiatry, 2019) followed 414 of the 575 original participants for a median of 5.4 years. in year 6, the high-needs housing first group was stably housed 85.51% of days against 60.33% for usual care (rate ratio 1.42). in the moderate-needs group it was 88.16% against 78.22% (rate ratio 1.13), down from rate ratios of 3.02 and 2.40 in year 1. the only paper we found that reports seven-year outcomes is the toronto acute-care paper (journal of urban health, 2021), and its results are mixed: the high-needs group (housing first with assertive community treatment, an intensive team-based outreach service) had fewer emergency visits and fewer days in hospital. the moderate-needs group (housing first with intensive case management, one worker per person) had more hospitalizations (relative rate 1.69) and more emergency visits (1.42) than usual care. so 'better' holds for housing stability up to year 6, but 'seven years' and 'so much better' go further than the record does.
assertion by assertion
| assertion | verdict | what the record says |
|---|---|---|
| people randomized to housing first in at home/chez soi stayed better off than usual care over long-term follow-up | confirmed | on the primary outcome, days stably housed, housing first beat usual care in every needs group through year 6 at the toronto site: high needs 85.51% vs 60.33% of days (rate ratio 1.42, 95% ci 1.19-1.69), moderate needs 88.16% vs 78.22% (1.13, 1.01-1.26). this covers housing stability only. we could not confirm long-term differences in symptoms, quality of life or substance use from the records we read. [1] |
| the advantage was still there seven years later | overstated | the housing-stability extension runs to year 6, with a median follow-up of 5.4 years. the only seven-year outcomes we found are hospitalizations and emergency visits (journal of urban health, 2021), and those are mixed rather than uniformly better. [1][2] |
| housing first participants did 'so much better' than usual care | overstated | the gap shrank a lot over time. year-1 rate ratios of 3.02 (high needs) and 2.40 (moderate needs) fell to 1.42 and 1.13 by year 6, because many people in usual care also found housing; by then the moderate-needs gap was about 10 percentage points. over 7 years, housing first with intensive case management (moderate needs) had more hospitalizations (relative rate 1.69, 95% ci 1.09-2.60) and more emergency visits (1.42, 1.01-2.01) than usual care. housing first with assertive community treatment (high needs) had fewer emergency visits (0.57, 0.34-0.95) and fewer days in hospital (0.32, 0.16-0.63), and no difference in hospitalization rates. [1][2] |
| the long-term data come from the whole five-city at home/chez soi trial | misattributed | both long-term papers we read cover only the toronto site. the housing paper is phase 2 (414 of 575 phase 1 participants); the acute-care paper covers 527 participants. they do not report results for winnipeg, montreal, vancouver or moncton. the speaker did not say whole trial; we include this because listeners may hear 'the at home/chez soi study' that way. [1][2] |
both directions
for the speaker: a randomized design with long follow-up found that housing first's advantage in stable housing lasted to year 6 in both needs groups, with confidence intervals above 1. the authors describe the effect as 'enduring' and larger for people with high support needs. at seven years, the high-needs group (housing first with assertive community treatment) had fewer emergency visits and spent far fewer days in hospital than usual care.
against: no published outcome we read lasts to seven years and is clearly better on every measure. housing stability ends at year 6, and the median follow-up was 5.4 years. the advantage shrank over time; for moderate-needs participants, year 6 meant 88% vs 78% of days housed. the one set of seven-year results is mixed: the moderate-needs group (housing first with intensive case management) had more hospitalizations and emergency visits than usual care. 'so much better' reads as a broad, large effect, and the record shows a lasting housing effect that got smaller, plus mixed results on acute care.
what this cannot say
we read pubmed abstract records and the pmc full text of the acute-care paper, not the full lancet psychiatry article. so we cannot say what that paper reports at year 6 on secondary outcomes like quality of life, community functioning, symptoms or substance use; there may be gains there that we did not check. more hospital or emergency use is not automatically a worse outcome, since it can mean people are finally getting care they had been missing. 19% of phase 1 participants did not join the extension. everything here is from one resource-rich city. we do not know which specific outcome the speaker meant at 26:23.
sources, each one fetched and read
- Long-term effects of rent supplements and mental health support services on housing and health outcomes of homeless adults with mental illness: extension study of the At Home/Chez Soi randomised controlled trial. Lancet Psychiatry, 2019. PMID 31601530 (PubMed abstract record) What we read there: 414 (81%) of 575 phase 1 participants joined phase 2, with a median follow-up of 5.4 years. rate ratio of days stably housed: moderate needs 2.40 (2.03-2.83) in year 1 falling to 1.13 (1.01-1.26) in year 6; high needs 3.02 (2.43-3.75) falling to 1.42 (1.19-1.69). in year 6: high needs 85.51% vs 60.33% of days stably housed, moderate needs 88.16% vs 78.22%. 'rent supplements and mental health support services had an enduring positive effect on housing stability.' Fetched 2026-10-22 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/31601530/
- The Effect of a Housing First Intervention on Acute Health Care Utilization among Homeless Adults with Mental Illness: Long-term Outcomes of the At Home/Chez-Soi Randomized Pragmatic Trial. J Urban Health. 2021;98(4):505-515 What we read there: toronto site, 7 years of follow-up, 527 participants. high needs (housing first + assertive community treatment): no effect on hospitalizations, emergency visits rrr 0.57 (0.34-0.95), length of stay rrr 0.32 (0.16-0.63). moderate needs (housing first + intensive case management): hospitalizations rrr 1.69 (1.09-2.60), emergency visits rrr 1.42 (1.01-2.01), no effect on length of stay. Fetched 2026-10-22 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC8382791/
speakers are described by role, not named. one sentence is quoted under fair use, with attribution and a link to the original at the timestamp; we do not re-host audio or video. nothing on this page is a reason to start, stop or change a medication. if you have the file behind anything marked unverifiable, the research board is where people can answer. adam