Physical space emerged as a modifiable lever in reducing coercive care
A rapid systematic review in BMJ Open examined whether the built environment of mental health facilities can help reduce the use of seclusion and physical restraint. The question is clinically direct: when inpatient units are designed differently, do coercive interventions become less necessary? Across 35 peer-reviewed studies, a pattern emerged around calmer, more private, and more therapeutic spaces, although the underlying evidence base was weak and uneven.
The review focused on architectural and environmental features in mental health settings rather than on staffing models, medication protocols, or policy alone. That matters because seclusion and restraint are often treated as operational responses to acute behavioural escalation, yet the physical setting can either dampen or intensify arousal. Crowded wards, harsh acoustics, a lack of privacy, and limited access to restorative spaces can heighten agitation in vulnerable patients. In contrast, environments that allow movement, offer sensory modulation, and create room for recovery may reduce the chance that distress turns into imminent risk.
Gardens, sensory rooms and private spaces clustered around fewer coercive episodes
Several themes recurred across the included studies. Facilities with access to gardens or recreational areas were repeatedly linked with more favourable outcomes. Sensory rooms and comfort rooms also appeared in the evidence as potentially useful features, especially when designed to give patients an alternative to escalating conflict. Private, uncrowded, and calm spaces formed another common thread.
These design elements are plausible from a behavioural and neurobiological standpoint. Acute psychiatric distress is often amplified by sensory overload, loss of control, and social crowding. Spaces that reduce noise, visual clutter, and interpersonal density may help lower autonomic arousal and improve a person’s ability to self-regulate. A sensory room can serve as a structured de-escalation environment; a garden can create a brief but meaningful break from confinement; privacy can reduce the confrontational dynamics that sometimes precede restrictive interventions.
A quieter ward is not a cosmetic improvement.
It may alter the threshold at which distress becomes containment.
The review does not support a single magic feature. Instead, the signal points to a layered environment in which multiple small design decisions appear to work in the same direction. Open sightlines, access to outdoor space, dedicated calming areas, and the avoidance of crowded shared settings may each reduce the conditions under which staff are forced toward seclusion or restraint. In practice, that suggests architecture functions less as a standalone solution and more as a structural support for trauma-informed care.
Low-quality evidence limits certainty, but the direction of effect is clinically important
Despite the consistency of the broad themes, the review found that the overall quality of the studies was low. That finding is central to interpreting the results. Many of the included studies were observational, and the literature appears to be vulnerable to confounding from unit culture, staffing levels, service mix, leadership, and concurrent quality-improvement initiatives. A modern ward redesign is rarely implemented in isolation, which makes it difficult to separate the effect of the building from the effect of broader practice change.
The review therefore offers preliminary rather than definitive evidence. Even so, the clinical implication is meaningful because seclusion and restraint carry known harms, including psychological trauma, injury risk, erosion of trust, and the possibility of reinforcing fear-driven behaviour. If the environment itself contributes to escalation, then redesign becomes part of prevention rather than an aesthetic afterthought. In that sense, architecture may influence the frequency with which staff reach for coercive measures during crisis care.
The included studies collectively suggest that environmental change may be most effective when it operates alongside therapeutic programming and staff training. A sensory room will not offset a punitive culture. A garden will not solve chronic understaffing. A quiet space will not replace skilled de-escalation. But when these features are embedded in a ward that already values patient autonomy and emotional safety, they may give both patients and staff another way through moments that otherwise end in restraint.
Co-design with consumers may determine whether design changes translate into safer care
The review closes on an important practical point: future design should include consumers in the codesign process. Patients with lived experience of inpatient psychiatry bring a perspective that architects, managers, and clinicians often miss. They know which spaces feel threatening, which layouts increase surveillance, which rooms support withdrawal without isolation, and which settings help preserve dignity during crisis. Their input is not decorative; it is likely to shape whether a facility truly reduces coercion or simply repackages it in a more modern interior.
That approach also aligns with the broader logic of trauma-informed care. If coercive episodes are partly triggered by environmental loss of control, then giving service users a role in shaping the space may itself be therapeutic. Co-design can identify practical features that matter at the bedside: access to daylight, predictable circulation, places for calm retreat, and areas where patients do not feel trapped or constantly observed. These details may seem modest, but they are precisely the kinds of details that can influence whether agitation is settled early or escalates into forced intervention.
The review leaves the field with a clear caution. The evidence points in a promising direction, but it remains preliminary, heterogeneous, and methodologically limited. Future work will need stronger designs, better reporting, and outcome measures that can isolate which architectural changes are truly associated with lower seclusion and restraint rates. Until then, the most defensible conclusion is narrower: the physical environment likely contributes to safer mental health care, and it deserves a place in efforts to reduce coercion.