- Design
- Prespecified secondary analysis of a randomised trial, with an observational historical control group
- Population
- 227,502 US veterans who contacted the Veterans Crisis Line
- Primary outcome
- Suicide mortality within 12 months of the index contact
- Effect
- 0.29% with caring letters vs 0.27% historical controls (P = .31); peer 0.21% vs clinician 0.22% (P = .95)
This prespecified secondary analysis followed US veterans who contacted the Veterans Crisis Line. Between 2020 and 2021, 87,926 received nine caring letters over a year, randomly signed by either a clinician or a peer veteran. Their suicide mortality over 12 months was compared with 139,576 veterans who contacted the line in the preceding years and received no letters.
There were 625 suicides in total. The suicide rate was 0.29% in the letter group and 0.27% in the historical controls, a difference that was not statistically significant. Whether a clinician or a peer signed the letters made no difference.
Caring contacts have been promoted on the strength of an older randomised trial that showed fewer suicide deaths. This large programme did not reproduce that signal at the mortality level, although the parent trial reported more health-care use after letters.
The comparison with no letters was not randomised, and crisis-line callers changed across years, including during the pandemic. So this does not prove letters are useless. It does say that a low-cost, light-touch contact is unlikely on its own to lower deaths in a high-risk group, and it should not replace safety planning, means restriction and timely follow-up.
- Keep caring contacts as one part of follow-up, not as the main suicide prevention step.
- Pair any outreach with a safety plan and a booked appointment after a crisis contact.
- Ask about access to lethal means at every follow-up after a crisis.
- Who signs the letter, clinician or peer, made no difference to deaths in this analysis.
- Track engagement with care after a crisis; it may be where letters help.
Why it matters
It challenges the assumption that a cheap, scalable contact by itself reduces suicide deaths after a crisis.
Don't overread it
The letters-versus-none comparison used a historical control group, so it cannot rule out a benefit.
The statistics, in plain English
A suicide rate of 0.29% against 0.27% means about 29 against 27 deaths per 10,000 people over a year. Suicide is rare enough that even 227,000 people may not detect a small benefit. Because the no-letter group came from earlier years, any difference, or lack of one, could reflect changes over time rather than the letters.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for psychiatry, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free