Sorry, you need to enable JavaScript to visit this website.
Skip to main content

The Dangerous Detour

Josh Gibson, MD; David H. Taylor, MD | June 1, 2003
View more articles from the same authors.

The Case

Following an overdose of alcohol and Ativan, a 26-year-old woman was admitted to the Medicine service for observation after being placed on a 72-hour hold by Psychiatry. Per hospital policy, a 24-hour sitter was placed with the patient.

The patient was to be accompanied to Radiology for a chest film and asked to go to the bathroom first. The transport and sitter were in the room when the nurse left the room to get the chart, which would go down with the patient. The sitter and transporter began to chat. The nurse returned and became concerned that the patient was still in the bathroom. She opened the bathroom door to find the patient with her gown tied around her neck and the door hinge. The patient was standing on the upside-down garbage can and was seconds from stepping off and hanging herself. The patient was unharmed; she was stopped in time.

The Commentary

This case illustrates the risk to safety that can arise when patients with acute psychiatric symptoms are admitted to a medical/surgical unit. Fortunately, no harm came to this patient, but this “near miss” provides an ideal opportunity to discuss how best to provide for the safety of a suicidal patient in a non-psychiatric acute care setting.

From 1995 to early 2003, 324 inpatient suicides were reported in US institutions.(1) This number, drawn from a national “sentinel event” database, represents a significant under-estimate, compared with numbers drawn from the psychiatric literature. One article on the epidemiology of suicide suggests that among the roughly 30,000 suicides occurring each year in the US, approximately 1,500 suicides (5%) occur in hospitals.(2) A recent review of 76 cases in the US (2) revealed that only 40% of inpatients who committed suicide had been admitted because of suicidal ideation.

Since both the team and the patient admitted to medical or surgical services generally seek improved health, the systems guiding inpatient medical care have been built to support that goal. Such systems are typically geared toward identifying and responding to signs of distress through a number of means, including those that are automated (eg, electronic monitors or alarms), provider-initiated (eg, physical exam or interview), and patient-initiated (eg, calls for help or reports of complaints).

The suicidal patient is atypical in this system.(3) Such a patient may not choose to be helped toward health, or even to be passively monitored while on that path. Importantly, the suicidal patient did not share the goal of the treatment team prior to admission (having recently tried to end his or her life) and quite likely maintains significant ambivalence about living.

How does the general acute hospital system deal with such an atypical patient? No automated system or lab test can provide useful information about suicide risk. Provider-initiated exams are useful, but prediction of suicide in an individual patient remains difficult.(2) False positive assessments cause indirect harm to patients by squandering scarce resources, while false negative assessments pose direct threats, as patients may receive inadequate protections. Patient-initiated calls for help are unlikely with the truly suicidal. Clearly, the typical systems in place will not function adequately to protect the safety of a suicidal patient.

For patients admitted to a medicine service following a serious suicide attempt who are judged an ongoing suicide risk, the usual solution has been continuous observation by a “sitter,” as in this case. Typically, these sitters lack formal medical, nursing, or psychiatric training.(4) They function, in effect, as monitors able to call for help when an unsafe situation presents itself. As this case demonstrates, continuous observation by a sitter has shortcomings that jeopardize patient safety. In essence, we are witness to a serious human factors problem in which the most immediately responsible member of the team is unfamiliar with the unit, unknown to the rest of the team, and ill-prepared to manage a very intense psychiatric condition.

Little literature exists regarding a “best practice” for treating suicidal patients on the medical ward, but aspects have been addressed separately in a number of studies, often specific to the psychiatric ward. A recent survey suggested that continuous observation procedures on psychiatric units are more effective at keeping suicidal patients safe (6), but many patients still commit suicide.(6) The challenges of keeping suicidal inpatients safe were made clear in a Canadian review of 100 inpatient suicides, which found that most medical ward suicides occurred in the first 7 days of admission and that suicides on the medical ward were significantly less predictable than those on psychiatric wards.(7) Given systems concerns and existing evidence, components of a “best practice” should address: the physical environment, administrative procedures, continuous observation procedures, suicide risk assessment, and the relationship between the patient and the treating team.

The physical environment of the psychiatric ward typically includes safety design features such as unbreakable mirrors, unbreakable windows with limited ability to open, and protuberances (eg, shower rods) that give way at low weight loads. Nonetheless, no unit is “suicide proof.”(9) Suicidal patients can and do examine the layout of a psychiatric ward to assess the lethal means that may be available to them, and they often perceive environmental safeguards as evidence of caring on the part of the staff.(9) As new medical/surgical facilities are constructed, they should be designed as suicide resistant. We also propose that, just as signs such as “Fall Risk” or “Seizure Precautions” are posted in and outside patient rooms, signs saying “Continuous Behavioral Observation” should be posted for suicidal patients. These signs would alert any member of the treatment team (including ancillary staff, such as the transport person in this case), that the patient should never be unaccompanied. Of course, training for all staff (including nonclinical personnel) regarding what “Continuous Observation” means should be undertaken.

Administrative procedures can also migrate from the psychiatric ward to the medical inpatient setting. A personal search of the patient and the patient’s belongings should always take place to remove dangerous articles of clothing. Visitors should be supervised to ensure no transfer of potentially lethal means including prescriptions from home. Standard order sets would minimize risk to patients and serve as the basis for care guidelines. Such orders must be clear and explicitly spell out what the treatment team needs to do (eg, “Patient must be observed continuously. Provide same gender observer for toileting/bathing.”).(10) A suicidal patient unattended in a bathroom is as risky as a person with a recent hip fracture unattended in a bathroom: both are accidents waiting to happen.

Continuous observation procedures for psychiatric wards lack standardization and use inconsistent terminology.(11) Often these focus on clerical fulfillment of checklists rather than true care or patient safety.(12) Each hospital should have a set of clearly defined protocols for continuous observation, and all members of the treatment team should be aware of their roles. While costs would be higher if mental health technicians were employed as sitters on medical wards, patient safety would be improved.

Risk assessment remains a critical part of caring for suicidal medical inpatients, but it lacks precision. Suicidality should be assessed regularly, as a vital sign, in such patients, just as pain assessment is now managed. Psychiatric consultation should also be utilized regularly in decision-making by the treatment team, which retains primary responsibility for patient safety.

The suicidal patient’s relationship with the treatment team plays a critical role in safety. All too often, there is palpable contempt for the recently suicidal individual who is perceived as creating unnecessary work for the team, using valuable resources that could be devoted to the “truly” ill, and trying to die when the team’s goal is to help people live. These perceptions only sharpen patients’ shame and rage at their own predicament and are not only traumatic but probably pro-suicidal. The team’s attitudes and interactions with the patient therefore influence the patient’s decision-making about suicide. Patients who feel the team cares for and understands them are more likely to report a sense of personal value and a decrease in suicidal ideation.(13)

Josh Gibson, MD Student Health Services University of California, San Francisco

David H. Taylor, MD Medical Services Director Clinical Services Development University of California Office of the President


1. Sentinel Event Statistics. Vol. 2003: Joint Commission on Accreditation of Healthcare Organizations; 2003.

2. Busch KA, Fawcett J, Jacobs DG. Clinical correlates of inpatient suicide. J Clin Psychiatry. 2003;64:14-9.[ go to PubMed ]

3. Kelly MJ, Mufson MJ, Rogers MP. Medical settings and suicide. In: Jacobs DG, ed. The Harvard Medical School Guide to Suicide Assessment and Intervention. San Francisco, Calif: Jossey-Bass: 1998; 498-99.

4. Kelly MJ, Mufson MJ, Rogers MP. Medical settings and suicide. In: Jacobs DG, ed. The Harvard Medical School Guide to Suicide Assessment and Intervention. San Francisco, Calif: Jossey-Bass: 1998; 514.

5. Green JS, Grindel CG. Supervision of suicidal patients in adult inpatient units in general hospitals. Psychiatr Serv. 1996;47:859-863.[ go to PubMed ]

6. Appleby L, Shaw J, Amos T, et al. Suicides within 12 months of contact with mental health services: national clinical survey. BMJ. 1999;318:1235-9.[ go to PubMed ]

7. Proulx F, Lesage AD, Grunberg F. One hundred in-patient suicides. Br J Psychiatry. 1997;171:247-50.[ go to PubMed ]

8. Jacobson G. The inpatient management of suicidality. In: Jacobs DG, ed. The Harvard Medical School Guide to Suicide Assessment and Intervention. San Francisco, Calif: Jossey-Bass: 1998; 394-397.

9. Benensohn HS, Resnik HLP. Guidelines for “suicide-proofing” a psychiatric unit. 1973.

10. Motto JA. Identifying and Treating Suicidal Patients in a General Medical Setting. Resid Staff Physician. 1983;79-87.[ go to PubMed ]

11. Bowers L, Gournay K, Duffy D. Suicide and self-harm in inpatient psychiatric units: a national survey of observation policies. J Adv Nurs. 2000;32:437-444.[ go to PubMed ]

12. Horsfall J, Cleary M. Discourse Analysis of an ‘observation levels’ nursing policy. J Adv Nurs. 2000;32:1291-7.[ go to PubMed ]

13. Cardell R, Horton-Deutsch S. A model for assessment of inpatient suicide potential. Arch Psychiatr Nurs. 1994;8:366-372.[ go to PubMed ]

This project was funded under contract number 75Q80119C00004 from the Agency for Healthcare Research and Quality (AHRQ), U.S. Department of Health and Human Services. The authors are solely responsible for this report’s contents, findings, and conclusions, which do not necessarily represent the views of AHRQ. Readers should not interpret any statement in this report as an official position of AHRQ or of the U.S. Department of Health and Human Services. None of the authors has any affiliation or financial involvement that conflicts with the material presented in this report. View AHRQ Disclaimers
Related Resources From the Same Author(s)
Related Resources