In the year that has gone by since our January 6, 2009 Patient Safety Tip of the Week “Preventing Inpatient Suicides” there have been several very informative articles pertaining to inpatient suicides. Keep in mind that suicides may occur in general hospitals as well as psychiatric hospitals so you need to be cognizant of the risks regardless of what type of facility you have.
An excellent review of inpatient suicides (Tishler 2009) emphasized that individual risk assessment is critical and that one of the biggest mistakes made is that decisions about managing suicide risk are often made based upon staffing levels rather than the individual patient’s risk assessment. In fact, Tishler et al caution against relying too heavily on risk factors or predictors taken from previous studies because those often don’t provide a good assessment of current risk. Rather they make the case for determining the presence of warning signs or immediate “red flags”. They make the distinction between chronic and acute risk factors, the latter tending to be more predictive of suicide and including symptoms such as severe anxiety, agitation, and severe anhedonia.
A study from Taiwan (Cheng et al 2009) compared psychiatric inpatients to “nonpsychiatric” inpatients who attempted or completed suicide in general hospitals and found some very important differences. Patients who attempted or committed suicide in general hospitals were older, more likely male, had more chronic physical conditions, and were more likely to attempt suicide in the first week of admission and use more violent means than those patients who were admitted to psychiatric units. They were also more likely to attempt suicide at night or while absent without leave and less likely to have communicated suicidal intent. Delirium and substance abuse were also more common in nonpsychiatric patients who committed suicide. Tishler et al (Tishler 2009) also noted that patients with delirium or dementia that is associated with agitation or impulsivity are at increased risk for suicide. Like the Cheng study, Tishler et al also noted that patients who attempt or commit suicide on general hospital units are more likely to be male and older, have agitation or delirium, have pain or other physical distress related to their medical or surgical condition, and often have factors such as poor family relationships, divorce, unemployment, bereavement, or a poor prognosis for longevity.
And when we are talking about “nonpsychiatric” patients above, we are not talking about the overdose patient who is temporarily admitted to an ICU until medically stable enough to be transferred to an inpatient psychiatry unit. Those patients usually get sufficient attention to suicide risk and close observation. Rather, we are talking about patients admitted to a med/surg unit with primarily a medical problem who go on to attempt or commit suicide, usually to everyone’s surprise. We don’t do a very good job of assessing suicide risk in the medical/surgical patient nor have we really come across any good tools for assessing that risk. Note that in some cases we may even enable it. We’ve talked before about the value of moving the delirious patient to a room that has more natural daylight. Well, some of these rooms also have windows from which someone can jump! So the choice of rooms needs to consider whether the windows can be opened from the inside or whether there are protective screens. Also keep in mind that patients on med/surg floors are more likely to have access to things like plastic trash can liners and cleaning materials (see below) that can also be used for suicide attempts. And whereas we may do a good job of removing suicide hazards from the environment on a psychiatric inpatient unit, the patient on the med/surg unit has access to sharps, tubing, loopable items and more easy egress for elopement (Bostwick et al 2009).
And one other good caveat: when we put a patient on 1:1 observation, whether for suicidal risk or because of delirium, we also need to remember it is not practical for any one individual to remain continuously vigilant for long periods of time nor should they be engaged in other activities. Tishler et al (Tishler 2009) recommend changing the observation person every two hours to avoid burnout.
Elopement/absconding is an issue that appears in many articles on inpatient suicide. That is these are inpatients who abscond/elope from the unit (either psychiatric unit or med/surg unit) and then commit suicide. A large study done in the UK (Hunt et al 2010) notes that such patients tend to be young, unemployed, and homeless with high rates of schizophrenia, previous violence, and substance abuse. They were also more likely to be involuntarily admitted to psychiatric units and likely to be noncompliant with treatment.
One of the more comprehensive studies on suicide focused on avoidable deaths (University of Manchester 2006). They confirmed that absconding from inpatient wards was a significant risk factor, particularly in the first 7 days. They note that wards can reduce absconding by:
• understanding the factors that trigger it, such as a disturbed ward environment or an incident affecting the patient
• making greater use of technology, such as CCTV or swipe cards, to observe and control ward entry and exit
The time immediately following discharge from an acute inpatient service is also a vulnerable period. 22% of the suicides in the Manchester study occurred between discharge and first followup appointment in community. They recommend regular assessment of risk during the period of discharge planning (or temporary leave) include:
The Manchester study was quite informative. Despite a substantial number of deaths, in only 28% of in-patient suicides did clinicians retrospectively view these deaths as preventable yet all such deaths should be regarded as potentially preventable. 22% of the in-patient deaths occurred in people who were (or were supposed to be) under observation, with 3% said to be under one-to-one observation. Two conclusions are clear from this. Firstly, intermittent observation regimens provide long gaps in observation and they are unsuitable for the care of high risk patients unless additional measures are taken, such as the observation of ward exits. Secondly, close observation must be strictly carried out. There should be no gaps in one-to-one observation and if a patient is to be observed every ten minutes, this time gap must be carefully adhered to. They also stressed that clinical staff need be diligent in removing non-collapsible curtain rails and eliminate other ligature (loopable) points, or at least make them inaccessible, with particular attention to hooks and handles on windows and doors. They also stressed risk factors for suicide, noting that those patients with dual diagnoses were especially at risk and that there are different antecedents of suicide in older patients.
Some lessons from our previous column “Preventing Inpatient Suicides” are also worth repeating. While suicide risk assessments are usually done on admission (though sometimes incomplete due to the patient’s inability or unwillingness to participate), the Joint Commission Sentinel Event Alert on suicides noted that reassessments are not well done.
We noted cases where patients have attempted suicide after locking themselves in the bathroom of the radiology suite (or other area in the hospital aside from the behavioral health units). We recommended use of a “Ticket to Ride” type communication tool for such hospital transports including special warnings and considerations for potentially suicidal patients so that all staff at the “receiving” end understand their responsibilities.
Also, the “sitters” commonly utilized to monitor the potentially suicidal patient on the non-psychiatric unit are often not specifically trained in assessment of the environment or management of the suicidal patient.
And standardized order sets for various behavioral health conditions, whether paper-based or computer-based, are now being developed and implemented in a more widespread manner. With use of computer order entry and clinical decision support tools, alerts and reminders (eg. to do a suicide reassessment) might be used to improve care.
There are also some good tools available that can help reduce the likelihood of patients committing suicide as inpatients. In our January 6, 2009 Patient Safety Tip of the Week “Preventing Inpatient Suicides” we noted that the VA has developed a mental health environment-of-care checklist that is available by e-mail request. Actually that checklist is now available online on the VA Patient safety website and there is a new article on use of that checklist in this month’s Joint Commission Journal on Quality and Patient Safety (Mills 2010). They implemented the checklist at 113 VA facilities and identified over 7000 potential hazards. A real value of the checklist is that it not only itemizes hazards but it is actually weighted by potential severity of the risk of each hazard (on a scale of 1 to 5). The commonest hazards they identified were anchor points that could be used for hanging. The second most common hazards were materials that could be used as a weapon against staff or other patients. Third most common were security issues that raised the risk for elopement. They also looked at the location of hazards and noted that bathrooms and bedrooms were a frequent site for hazards. Those two rooms obviously are potentially at greater risk for suicide because of patient isolation. Their discussion of the elopement risk is also quite good. They note certain areas (eg. physical therapy rooms, art rooms, group rooms, utility rooms, etc.) where it is important to identify that patients will not be left unsupervised and they discuss safeguards such as self-closing and locking doors. Though they discuss the use of video camera monitoring, they point out that it is unreasonable to expect staff to reliably monitor video screens for long periods of time. They also point out that, though they found materials for suffocation or poisoning less often, the high potential for lethality of those materials merits special attention. This would include items like plastic liners in trash cans and cleaning products. Those are especially important to look for on units other than psychiatric units. Overall, this is a very good checklist for conducting environmental rounds with a purpose of reducing potential risk for suicides.
The National Patient Safety Agency (NPSA) in the UK also has a toolkit on preventing suicide for mental health services. This includes the toolkit itself, a ward manager checklist and an audit tool. These tools address not only the environmental risks similar to the VA tool but also look at your systems for evaluation of patient suicide risk assessment, layout of your unit, observation policies, treatment plans, post-discharge plans and follow-ups, medication and compliance issues, and staff training and retraining. And don’t forget that you need to educate and train any agency staff that may be working temporarily on your units. And teaching family members or significant others what signs of symptoms to watch for is also very important. Very comprehensive tools.
Suicide on one of your inpatient services, whether psychiatric or med/surg, is a devastating event for families, your other patients, your staff, your community, and your reputation. You need to get a better understanding of your vulnerabilities and take action to mitigate the risks of potentially avoidable events.
Update: See our December 2010 What’s New in the Patient Safety World column “Joint Commission Sentinel Event Alert on Suicide Risk Outside Psych Units”
Tishler CL, Reiss NS. Inpatient suicide: preventing a common sentinel event. General Hospital Psychiatry 2009; 31: 103-109
Cheng I-C, Hu F-C, Tseng M-C M. Inpatient suicide in a
General Hospital Psychiatry 2009; 31: 110-115
Bostwick JM, Lineberry TW. Editorial on “Inpatient suicide: preventing a common sentinel event”. General Hospital Psychiatry 2009; 31: 101-102
Hunt IM, Windfuhr K, Swinson N, et al. and the National Confidential Inquiry into Suicide and Homicide by People with Mental Illness. Suicide amongst psychiatric in-patients who abscond from the ward: a national clinical survey. BMC Psychiatry 2010; 10: 14
University of Manchester. Five Year Report of the National Confidential Inquiry Into Suicide and Homicide By People With Mental Illness. Avoidable Deaths. December 2006
Mental Health Environment of Care Checklist
Mills PD, Watts BV, Miller S, Kemp J, Knox K. DeRosier JM, Bagian JP.
A Checklist to Identify Inpatient Suicide Hazards in
Veterans Affairs Hospitals
Joint Commission Journal on Quality and Patient Safety. Volume 36, Number 2, February 2010 pp. 87-93(7)
NPSA (UK). Preventing suicide: a toolkit for mental health services
ward manager checklist