A small interview study of two Japanese psychiatric hospitals suggests that crisis planning may influence more than individual treatment decisions. At one hospital, staff described a broader shift toward patients expressing their preferences and professionals responding to them.
The study does not establish that peer-led planning reduced restraint, compulsory admission or psychiatric symptoms. Researchers interviewed 13 people in October and November 2025, and most accounts came from one hospital with an unusually strong commitment to reducing restraint.
Three conditions stood out
At the main hospital, researchers identified three conditions associated with the approach: the hospital accepting responsibility for clinical risk, the gradual introduction of peer workers’ lived experience into the team, and patients writing their own crisis plans.
A crisis plan is prepared while a person is well. It can set out early warning signs, responses that may help or worsen a situation, and preferred treatment. In this model, peer support workers—staff with lived experience of psychiatric illness and treatment—helped patients write plans in their own words.
The source notes that clinicians often write crisis plans themselves. That can turn a document intended to express a person’s rights and preferences into a checklist centered mainly on symptoms and clinical risk.
From being managed to requesting care
Staff at the main hospital described patients moving from being managed by professionals to asking for specific forms of care. Professionals then responded to those requests, according to the interview accounts.
Staff also described patients having more freedom to move around the unit, access to a vending machine and permission to keep phone charging cables. The hospital had a ward-wide zero-restraint policy, and restraint equipment had been removed.
Two nurses said episodes of agitation had decreased significantly. That was a staff impression, however, rather than a rate measured by the researchers.
Why the comparison hospital matters
The comparison hospital used professional staff without peer workers in patient care. The study description said crisis plans helped individual patients there but did not lead to hospital-wide change.
That contrast is part of the researchers’ explanation for why the plans may have had a wider effect at the main hospital. They tentatively proposed that change may require formal organizational responsibility for clinical risk, a protected staff member who can bridge different perspectives, enough peer workers with crisis-plan experience, and continuing discussion between professional and lived-experience viewpoints.
What earlier trials found
The interview study was designed to explore how change might happen, rather than to test outcomes. Its findings sit alongside mixed evidence from randomized trials.
- A meta-analysis of five randomized trials found a 25% lower risk of compulsory admission with crisis planning, with a risk ratio of 0.75 and a 95% confidence interval of 0.61 to 0.93.
- A 2004 English trial involving 160 people reported compulsory admission or treatment for 13% of participants offered a joint crisis plan, compared with 27% receiving usual care.
- The 2013 CRIMSON trial, which randomized 569 people, found no significant reduction in compulsory treatment.
- A 2022 French trial involving 394 people with schizophrenia, bipolar I disorder or schizoaffective disorder reported compulsory admission for 27% of those offered peer-worker help with advance directives, compared with 39.9% receiving usual care over 12 months.
- A 2012 US trial of 519 adults found that an eight-week, peer-taught Wellness Recovery Action Planning program reduced depression and anxiety symptoms more than usual services.
These trials examined different programs, populations and outcomes. The Japanese interviews therefore add context about organizational conditions, but they do not resolve whether the approach improves measurable outcomes across hospitals.
A promising account with important gaps
The study is especially limited by its small and uneven sample. Twelve of the 13 interviews came from one hospital. The comparison hospital contributed only one informant, a day-care unit chief, and no nurses or peer workers there were interviewed.
No patients were interviewed, so service-user experiences were conveyed through staff and peer workers. The accounts were retrospective, rather than observations recorded as events occurred. The researchers also did not measure restraint use, compulsory admissions, symptoms or incidents.
As a result, the study cannot show whether all three identified conditions were necessary, whether they simply added to one another, or whether other differences between the hospitals explained the contrast. It also remains unclear whether the bridging role described in the study could be transferred to other settings.
For readers and services considering crisis planning, the central relevance is the distinction between having a plan and creating the conditions for that plan to be heard. In this study, staff linked broader change to patient authorship, peer involvement and organizational responsibility—but those links remain a qualitative explanation rather than proof of cause and effect.
AI tools were used to assist with the preparation of this article under our editorial standards.
Leave a Reply