No trial has shown that suicide risk screening reduces suicide deaths. Say so plainly, and say so first.
A program that overclaims gets corrected in public by the one person in the room who knows the literature. This Module is written so that SAFER is never that program.
What is actually missing
Not the detection evidence. Screening instruments identify students, those students are mostly not already known, and screening increases service use. Module 2.1 covers that.
What is missing is the final link in the chain: a randomized trial showing that a school-based screening program reduces suicide attempts or deaths.
That is a very hard trial to run. Suicide deaths are, statistically, rare events, which means detecting a change in them requires enormous samples or very long follow-up. The absence of such a trial is partly a statement about the difficulty of the research and not only about the intervention.
The two findings that point the right way
One: a three-year follow-up. A multi-stage school screening program followed 9th grade students over three years and found an association between the program and a reduction in students reporting past-year suicidal thoughts and attempts.
It was nonrandomized. That has to be said every single time it is cited, including here, including by SAFER, including when it would be convenient not to.
Source: Torcasso G, Hilt LM. Suicide prevention among high school students: evaluation of a nonrandomized trial of a multi-stage suicide screening program. Child Youth Care Forum. 2017;46:35-49.
Two: the largest randomized trial of suicide prevention in adolescents tested screening directly. A cluster-randomized controlled trial published in The Lancet in 2015 ran several intervention arms across European schools, one of which was a professional screening arm.
The screening arm moved in the right direction and did not reach statistical significance.
Source: Wasserman D, Hoven CW, Wasserman C, et al. School-based suicide prevention programmes: the SEYLE cluster-randomised, controlled trial. Lancet. 2015;385(9977):1536-1544. [unverified – must be read in the original before any arm-level claim is published]
The argument SAFER makes, and it is an honest one
The largest randomized trial of adolescent suicide prevention tested screening. The screening arm moved in the right direction and did not reach significance.
The most likely reason is that the students the screen identified were not reliably connected to care. A screen that finds a student and then loses them is a list, not a service.
That is not a reason to doubt screening. It is the specific reason to think the question is still open, and it is exactly what SAFER was built to fix. Every High Risk student is seen the same day, within two hours, and goes home with a safety plan made in the same meeting.
SAFER is not claiming to have proved this. SAFER is claiming to be the version of the intervention worth testing.
That is why the evaluation partnership exists, and why its first question is whether the High Risk students were already known to staff. See Module 10.4.
Four other things SAFER does not know
1. Whether the instruments perform equally well across the populations in Unit 8. The C-SSRS Screener, PHQ-9A and GAD-7 have validation literatures, and those literatures are not evenly distributed across every group of students in an Oregon school building. See Module 4.1 and Module 8.1.
2. Whether the same-day standard survives scale. SAFER has met it 22 times out of 22 across 2026. That is a real operational achievement and it is also a small number. A building with one counsellor and 900 students is the test.
3. Whether twice-yearly screening is the right interval. It is a reasoned choice rather than an evidenced one. A student who screens Low in September can be High in February, which argues for more than once; staff capacity argues for not many more.
4. What happens after 60 days. SAFER tracks care continuation at 30 and 60 days. Beyond that, nobody in school-based screening has good data, including SAFER.
Critical Reflection. Of the four unknowns above, which one would change what your school does if the answer came back badly? That is the one to watch in your own data, and it is a better use of a staff meeting than re-litigating whether to screen.
What a school should say out loud
When a board asks whether this is proven, the defensible answer is not yes.
"Screening reliably finds students we would otherwise miss, and most of them are not in care. Whether a program like this lowers suicide deaths has not been shown in a trial, and the main reason earlier trials fell short was the gap between finding a student and getting them care. That gap is the part we built."
That answer survives the follow-up question. "The research shows screening saves lives" does not.
# UNIT 3 – PLANNING A SCREENING
SAFER Universal Screening Toolkit. Eight Modules.
The work before the week. Most of what makes a screening go well happens in the six weeks before anyone opens a laptop.
