The fact that many psychiatric treatment services are now administered as outpatient programs has not diminished the need for inpatient behavioral healthcare beds. In fact, there appears to be a mini-boom in design projects for both new construction and the renovation of existing behavioral health hospitals and units. Many existing facilities have become costly to maintain, are energy inefficient, are technologically outdated, and do not provide a built environment that has kept pace with the current standard of care for the safe delivery of mental health services.
As Stewart Brand notes in How Buildings Learn, almost no buildings adapt well.1 Healthcare facilities are arguably more adaptable than most, but the rate of change in healthcare treatment modalities and technologies demands a higher rate of change than is needed for other structures. Louis Sullivan may be right that form follows function, but in the case of healthcare design the form is forever playing catch up with technologic innovation and treatment theories that are in a constant state of flux. In short, often by the time the behavior health treatment unit is built, codes have changed and clinicians have moved on, both literally and figuratively, to a different way of thinking.
Add to this a lack of consensus on so many of what would appear to be basic assumptions about the “architectural standard of care” in behavioral health architectural design. Unresolved a priori notions such as cottage/campus design vs. consolidated/coherted (i.e., grouped together based on a similar diagnosis, trait, or characteristic) services or semi-private vs. single-bed patient rooms do not provide design teams and hospital staffs with clear direction as to just what needs to be built. The only current agreement seems to be that the more residential and less institutional the look of the place the better.
Unfortunately, the focus on a residential look and feel for behavioral health units, often based on weak or poorly designed research, has eclipsed the lessons learned regarding acts of self-harm by behavioral health patients. This is not to say that some of the evidence-based design choices are not legitimate (But it wasn't too long ago that the “evidence” told us to paint all seclusion rooms pink2). Yet an oft-cited post-occupancy survey of a new behavioral health facility, which found increased patient satisfaction in a new space while staff expressed feeling isolated, cannot be taken as a ringing endorsement of the design.3 Awards for architectural design often are based on how things look and not how things work, their functionality, or their context. Thus, we run the risk of emulating “magazine architecture”: Artistic architecture work that looks good on paper and in print but does little to improve patient outcomes.4
Patient acts of self-harm and elopement remain the two events that drive many of the safety-related design choices in the behavioral health unit. Although the Joint Commission requires that all patients be assessed for suicide risk, suicide is the second most frequently identified Joint Commission Sentinel Event.5,6 In 75% of the inpatient suicides reported to the Joint Commission, the method was a hanging in a bathroom, bedroom, or closet, and 20% resulted from patients jumping from a building.7 A 2008 study found that doors and wardrobe cabinets accounted for 41% of the anchor points when hanging was the method of self-harm.8 Reliance on suicide risk assessments or no-suicide contracts (i.e., a patient essentially says he won't hurt himself and will notify staff if he thinks about hurting himself) as a reliable tool for suicide risk management is fraught with peril.9 Designers, architects, and their clients should proceed as if there is no reliable means to determine a patient's true intentions.10-12
Emerging successful patterns
If patients should be presumed to be at risk of self-harm and the unit should be designed accordingly, how can the conflicts in the evidence-based design literature between those who call for a “homelike” environment and those who call for a “secure” unit be resolved? Although there is no consensus regarding the best practices and many studies are not well controlled, a confluence of patterns is beginning to emerge.
The designer should be aware that experience and research in a medical/surgical setting do not translate well to the behavioral health environment. A 2007 survey by the National Association of Psychiatric Health Systems communicated three important differences between the psychiatric treatment environment and the medical/surgical environment to the AIA/FGI Healthcare Guidelines Revisions Committee.
First, patient treatment does not take place in sleeping rooms on psychiatric units. Patients receive group and individual therapy in group rooms or interview rooms. Patients are encouraged to be in the dayroom during their waking hours for therapeutic reasons.
Second, infection control is handled differently in psychiatric treatment environments. Isolation is rarely attempted on psychiatric units because of the ambulatory nature of the unit population. Thus, the studies that show benefits to single-bed patient rooms in reducing airborne and presumably contact infections are rendered largely moot in a behavioral health environment.13