The measure thatisn’t there.
Return to work after a mental-health leave depends less on the clinic than on the workplace a person comes back to — and above all on whether it feels safe to not yet be fully okay. Researchers have a name for that, and validated ways to measure it. We screened nearly 5,000 studies to ask one question: has anyone measured it in workers returning from leave? No one had. Not once. Not in Canada, not anywhere.
If you are living through a disability claim yourself, you are welcome here, and support is available now if you need it.
The field has circled this idea for fifteen years without measuring its core variable.
There is broad agreement that the social environment a worker returns into shapes whether the return holds: the supervisor, the team, whether it is safe to ask for an adjustment or to say today is hard. The construct that names it — psychological safety — is well established and measurable, including with a Canadian-built instrument. So we asked a plain question: in the research on returning to work after a mental-health-related leave, is psychological safety actually measured — or only gestured at? The wider evidence base, including the Canadian court record, sits on The Evidence.
A structured screen of 4,995 studies, with one test built in.
We ran a structured, AI-assisted screen of 4,995 studies of mental-health-related work disability, with a deliberate test built in: separate the studies that merely describe workers as “supported” from those that actually measure psychological safety as a defined construct. We then verified the result four independent ways, including a check on the construct that comes closest — psychosocial safety climate. No counterexample turned up.
This is an evidence scan, not a peer-reviewed systematic review. A registered scoping review, built to full methodological standard with academic partners, is in progress. Read how we verify our evidence.
The two literatures do not intersect.
The return-to-work literature and the psychological-safety-measurement literature run on separate tracks. Of nearly 5,000 papers, 695 studied return to work and 35 measured psychological safety as a construct. Zero did both. None were Canadian.
Of 4,995 studies screened, none both measured psychological safety and studied return to work
Drawn as sets, the two bodies of research share no members: not a single study sits in the space where they would overlap.
The return-to-work and psychological-safety literatures are disjoint
It is not that the field ignores the workplace — it reaches for the idea without ever measuring it. Among the 695 return-to-work studies, 138 gesture at “feeling safe,” a “safe climate,” or “support,” 557 do not address it, and not one measures psychological safety.
Within the 695 return-to-work studies, psychological safety is named but never measured
A real gap, not a quirk of one search.
The return-to-work field does measure the workplace — through well-established constructs like supervisor support, the demand–control–support model, effort–reward imbalance, organizational justice, and return-to-work self-efficacy. Psychological safety — the felt safety to take an interpersonal risk, to not yet be okay — is distinct from all of these, and it is the one that has not been measured in this population. That is why this is not just “support” relabeled.
Return-to-work research measures the workplace, through every construct except this one
And where psychological-safety-type ideas have been studied, it is on the prevention side — what keeps people from going on leave — never carried downstream into the return.
Psychological safety is studied to prevent leave, never carried into the return
We can’t hold practice to something no one can measure. So the gap defines a program.
Each step below is publishable on its own and produces a free public good. Together they climb the gap from measurement to mechanism to intervention.
- 01
Build the measure
A validated Return-to-Work Psychological Safety measure for workers coming back from a mental-health leave — the first of its kind, and the seed of the standard.
- 02
Test the mechanism
Whether psychological safety predicts who returns successfully, and whether it is the reason supportive workplaces produce more durable returns — not just a correlate.
- 03
Test the intervention
Whether deliberately raising team and supervisor psychological safety makes returns hold, evaluated against usual disability management.
A peer-reviewed scoping review is underway with academic partners. The measure, once built, is the instrument behind better screening, better employer guidance, and a standard worth teaching to.
Where this is strong, and where it isn’t.
This was an AI-assisted screen, so completeness is established by triangulation, not proof — but four independent searches agreed on the same empty result. “Measured as a construct” is a deliberately strict bar; a reader who counts any “support” question as psychological safety would dispute the zero, which is exactly why the distinction between psychological safety and its neighbours matters here. The peer-reviewed version will redo the searches transparently across named databases.
Claimant evidence is used in aggregate only; we never name, quote, or profile an individual. NAPSDM is an independent federal not-for-profit corporation. See How we verify and Sources.
A standard worth holding practice to starts with measuring what matters.
This is one strand of NAPSDM’s research. See the wider evidence base, or read how every claim is verified.
This page is licensed under Creative Commons Attribution-ShareAlike 4.0 (CC BY-SA 4.0) (opens in a new tab). Any worker, union, clinician, employer, insurer or practitioner may use, adapt and share it with attribution to NAPSDM, provided any adapted version carries the same licence and passes the same freedom on.