Behavioral Health
Patient Safety Risk Assessment

by James M. Hunt, AIA, NCARB
and David M. Sine, DrBE, CSP, ARM, CPHRM

Distributed by the National Association of Psychiatric Health Systems

The level of concern for the safety of patients and staff due to the design of the built environment is not the same in all parts of a behavioral health unit or facility.  The level of precautions necessary depends on the staff’s knowledge of the patient (i.e. the patient’s intentions regarding self-harm) and the amount of supervision the patient will have while using that part of the facility.

Previous editions of this Design Guide have proposed that the level of concern for patient safety in the behavior health built environment could be stratified into five categories (with five being the highest level of concern).  The lowest level (Level I) was described as spaces having no patient access or under constant supervision such as staff and service areas and correlated to an area in which some latitude was available regarding design, construction, and materials used.  The highest level (Level IV) was described as an area within the built environment where patients were present with unknown or unassessed risks and in which patients could be cared for that were in a highly agitated condition.  The Level V areas present special considerations that need to be addressed individually.

This approach of risk levels based on a functional statement of intended usage has been adopted by many others with varying numbers of levels but all share a similarity of describing the level of risk for a room or space that is similar to spaces with a similar occupant function (i.e. admissions rooms, examination rooms, etc.).   However, some rooms or room functions can comfortably fit into more than one category or sit on a blurry boundary between two categories.   In addition, the categories do not anticipate every use of every room.  Thus, facility clinical staff and facility designer may be making assumptions when a room is described as an “activity room” and a level of concern to drive design choices is made that does not meet the actual needs of the stakeholders in an operating environment.  For example, a Day Room may be located so that it is within line of sight of a nurse station that “always has staff present”.  However, if there is a patient who can’t sleep and he or she is in the Day Room watching television at 2 AM and the only staff on duty is making rounds, the patient may actually be “completely alone” for a period of time in a space that may contain hazards.

For this edition of the Design Guide the authors propose that the conversation between clinical staff and designers regarding patient safety could be facilitated by the use of a “patient safety risk assessment” (PSRA, see page 14) that, in a Cartesian matrix, considers the opportunity for a patient to be alone in a particular space (of any name) on one axis and a level of risk of self-harm on the other axis.  The greater the opportunity for a patient to be alone the greater the opportunity for self-harm and the greater the caution that should be taken regarding design choices and materials.   The authors acknowledge that patient intent for self-harm is often opaque and difficult to assess but place “actively suicidal” on the far end of this scale and describe the opposite end as “self-harm not anticipated”.  Privacy is arranged with close observation (such as “1:1 observation”) on one end of the opportunity scale and the patient “completely alone” on the opposite end of that scale.  This risk matrix is partly informed by longitudinal studies done by the Veterans Health Administration of the frequent locations of acts of self-harm by inpatients and supported by Joint Commission data and is further influenced by the seminal works of Richard Prouty on risk maps.  Designers and clinicians, rather than seeking agreement on what is meant by a particular room name, may now seek to agree on the actual or anticipated degree of aloneness or privacy a patient will experience in a room or space independently of room name and it is that agreement that will drive design choices for that room or space.

For example, a room such as a patient bathroom in which the patient is anticipated to be alone and have privacy would be far along the privacy axis.  If that assessment intersects far along the patient intent for self-harm axis then the space should be designed with the attributes of a Level IV space as described below.  In sum, no matter the name of the room, a high level of privacy warrants a high level of concern if it is anticipated that patients who are actively suicidal (or patients with an unknown or unassessed intent for self-harm) are to be treated or housed in that space.  Some may also note that spaces with risk assessments located in the upper right (Level IV) of the risk map will have a more “institutional” look than spaces with a risk assessment located in the lower left (Level I) corner of the risk map (which may look more residential than the institutional spaces).

Although the authors believe that the use of such a tool will facilitate the necessary conversation regarding patient safety and design between operators, clinicians, and designers the tool is not an absolute and not intended to predict risk levels in a particular facility (which the authors believe to be dynamic and non-static).  The tool is only intended to encourage a dialog and promote a common understanding of for whom a designed space is intended and the risks of an anticipated patient population.  Neither should this proposal be interpreted as a suggestion that patient privacy is to be avoided or a risk to be avoided.  Quite the contrary, privacy is generally considered a good thing and desirous in the built environment, but privacy has associated with it a risk that should be considered and mitigated though good design when possible.

bhfcllc-patient-safety-chartEdited to reflect Recommendations in The Joint Commission’s “Perspectives”  November 2017 – Volume 37 – Number 11

  • Level I:  Areas where patients are not allowed or are under constant supervision such as staff and service areas.
  • Level II: Areas behind self-locking doors where patients are highly supervised and not left alone for periods of time such as counseling rooms, activity rooms, interview rooms and corridors that are fully visible.
  • Level III: Areas that are not behind self-locking doors where patients may spend time with minimal supervision such as lounges, day rooms and corridors that are not fully visible.
  • Level IV:  Areas where patients spend a great deal of time alone with minimal or no supervision such as patient rooms (semi-private and private) and patient toilets
  • Level V:  Areas that require special consideration where staff interacts with newly admitted patients that present potential unknown risks or where patients may be in a highly agitated condition. Due to the unknowns, these areas fall outside of the risk map and require special considerations for patient safety. Such areas include seclusion rooms, examination rooms and admission rooms.