We had been waiting to write a column about suicide risk in bathrooms on behavioral health units after 2 such cases in a recent batch of CDPH case reports. Now a third such case reminded us that this risk continues and there are lessons learned that might help prevent suicide in other hospitals.
The first case (CDPH 2016a) was a 17 y.o. patient admitted voicing suicidal intent. Because of continued suicidal ideation her observation was moved from every 15 minutes to every 5 minutes. When the patient was showering, the nurse who was monitoring her was called away for an emergency (unruly patient) to give medications in that other emergency. The patient was left unattended in the shower. There was a handoff to a MHW who responded “within minutes” but the patient was the found in the shower with the shower hose wrapped around her neck. CPR was unsuccessful and she died.
The bathroom had been equipped with a hand-held shower head and flexible metal hose that was intended to be compatible with the Americans with Disabilities Act (ADA). But this patient had no disabilities that would have merited use of that special shower apparatus and it should have been removed from that bathroom after use by someone who may have needed it. It was only intended for use under the supervision of facility staff and, when not in use, was supposed to be removed and stored in a secure location.
The hospital’s policy was that patients should be visually checked at least every 30 seconds while showering.
It was felt that the need to remove the special shower head had not been adequately conveyed to all staff. So in addition to staff training, they also developed a sign in/out log for that shower head.
The case also provides a good example of how another emergency may lead to distractions that allow an incident to occur. We’ve previously mentioned how distractions like fire alarms may lead to other incidents, such as patient elopements.
The second case (CDPH 2016b), at the same hospital as above, involved a male patient admitted after an overdose attempt. Because of suicidal ideation he was originally on 1:1 observation but was then changed to every 5 minute observation. On the day of the incident he was showering “for quite a while”. A mental health worker (MHW) heard the water stop and went to check on the patient. But the patient angrily yelled that he was getting dressed and needed more time. The MHW felt intimidated and left to give him more privacy. When she returned to check on him, she found him hanging with clothing from bathroom doorknob, pulseless and not breathing. He was resuscitated and transferred to another facility.
The plan of correction included an inservice on how to deal with intimidating patients, including role playing to demonstrate various scenarios. Staff were also educated on guidelines for shower use and what clothing can be taken into shower. The facility also installed anti-ligature door knobs on all patient bathrooms and showers. They also developed a reassessment for suicide risk tool.
We find it somewhat surprising that neither the facility nor the CDPH investigator mentioned the influence of gender issues in monitoring patients taking showers. We would have presumed most facilities would have a same-sex person monitor for patients in the shower. Perhaps that was one factor that allowed for the female MHW to be intimidated by the male patient in this case.
The third case (Fisher 2017) was described in an inquest in Alberta, Canada. The 49 y.o. male patient with a history of bipolar illness and major depressive illness was brought to the hospital by police after family stated he had wanted to cut his own head off. Though the patient himself denied any suicidal ideation, he was admitted involuntarily to a behavioral health unit. Though his mood and judgment were described as good, it was noted he was vague at times and jumped from one topic to another. No hallucinations were noted. Staff noted the discrepancy between the history provided by family and by the patient. Few or no concerns were noted during the patient’s stay and he was mostly on observation every 30 minutes. On the day following admission he was noted to be on the telephone at about 21:50 and then appeared to be upset following the phone call. At 22:30 he was noted by the nursing attendant (who was doing observation rounds while the nurse who would usually do the observation rounds was on break) to be sleeping and breathing normally (observation done from about 10 feet away). He was found at 23:00 hanging by a bedsheet from his bathroom door. He was temporarily revived but subsequently died.
He was not in one of the seclusion rooms on the behavioral health unit that had video cameras. The light in his room was off, though the light in the bathroom was on when he was found. The room did have a large window facing the nursing station but the curtains were closed.
The inquest noted that it had been recommended in a number of past Inquiries regarding psychiatric units, that all rooms in all psychiatric facilities be monitored by video cameras but this has not been undertaken by the hospital. The current inquest recommended that all units within the psychiatric unit in this hospital, whether the rooms are seclusion rooms or not, should be monitored by video cameras, with a rationale that the safety of the patients is more important than the privacy issues that might be argued.
The evidence also noted the patient had tied his bed sheet to a hanger on the inside of the bathroom door. It was recommended that all mechanisms attached to bathroom doors in the psychiatric units be removed.
There was also testimony that patients become aware of the specific times that patient observations are carried out and that the patients should not be able to predict exactly when the patient observations will be carried out so it was recommended that nursing staff and security staff who carry out patient observations not utilize exactly the same routine for checking the patients, while still being within reasonable compliance with the ordered frequency of observation.
Previous medical records were not readily available to the treating psychiatrist. These apparently would have shown a series of suicidal attempts and threats dating back 10 years, including two earlier stays at the hospital that year, including one only two weeks before his last admission (Lo 2017). Conceivably, those records might have led to assignment of a higher suicide risk category and potentially increasing the frequency of observation or moving the patient to one of the rooms that had video monitoring capability.
A number of other issues were also identified in the inquest (lack of prompt inquiry, failure to sequester video recordings from the nursing station, lack of prompt access to prior medical records, issues surrounding patient use of phones, and requirement that any informal notes made by security personnel doing observation rounding be made part of the formal patient record).
These 3 cases illustrate some dilemmas and tradeoffs. The tradeoff between privacy and safety is an obvious one. The Canadian case raises the question of video monitoring in patient rooms. But a real dilemma is the privacy/safety tradeoff when patients are showering or using the bathroom.
The other tradeoff was the issue of having an ADA-compliant hand-held showerhead vs. the threat such could be used for hanging. The intervention (a sign in/out log for that shower head/hose) implemented by the hospital in first case (CDPH 2016a) is a good step but it still requires someone to remember to return the shower head/hose promptly to its secure site. In this day and age where RFID and Bluetooth technologies are readily available one could envision sending timed alerts to prompt removal of that item from patient bathrooms.
Note also that in the first 2 cases the hospital’s policy was that patients should be visually checked at least every 30 seconds while showering. The first two cases demonstrate that may not be practical in many cases (in one case the unrelated emergency took the observer away from the patient and in the other case the patient’s vociferous complaints led to non-compliance with the 30 second policy).
Perhaps the more important lesson is the need to eliminate items in the environment that can be used for suicide. Many of those issues are addressed in the VA’s Mental Health Environment of Care Checklist (MHEOCC), which we’ve discussed now in several columns (our Patient Safety Tips of the Week for January 6, 2009 “Preventing Inpatient Suicides” and February 9, 2010 Patient Safety Tip of the Week “More on Preventing Inpatient Suicides” and our July 2012 What's New in the Patient Safety World column “VA Checklist Reduces Suicide Risk”). That checklist is available online on the VA Patient Safety website.
One very pertinent question asked in the MHEOCC is “Are doors that are within rooms and that open to other in-room areas such as bath/shower/toilet areas (i.e., not corridor doors) designed to eliminate anchor points?”. But keep in mind that almost any type of solid door might be used as an anchor even if it lacks latches, hooks, or other obvious loopable items. One could still conceivably loop bedsheets or clothing over the top of a solid door even if it has a “sloped” surface. Therefore, the MHEOCC recommends soft break-away doors for bathrooms and showers.
In our February 14, 2017 Patient Safety Tip of the Week “” we mentioned 2 publications (Watts 2016, Mills 2016) showing sustained results from implementation of the Mental Health Environment of Care Checklist (MHEOCC). The checklist and program became mandated at all VA hospitals in 2007. Inpatient suicide rates in VA hospitals dropped from 4.2 per 100,000 admissions to 0.74 per 100,000 admissions from 2000 to 2015. The reduction in suicides coincided with introduction of the MHEOCC and has been sustained since implementation in 2007. The authors stress that the physical changes brought about by the MHEOCC likely have a bigger impact on inpatient suicide reduction than the numerous other interventions used.
The MHEOCC is an excellent tool for identifying and abating environmental factors that might facilitate inpatient suicide. But don’t forget the risk of suicide in bathrooms not on behavioral health units. In several of our columns on suicide outside of behavioral health units we’ve noted that the potentially suicidal patient on an intrahospital transport, such as a trip to the radiology suite, may lock him/herself in a bathroom in that suite and there are a number of loopable items in those bathrooms. You’ll recall that in our March 16, 2010 Patient Safety Tip of the Week “A Patient Safety Scavenger Hunt” we included the items below as ones to search for in your patient safety scavenger hunt:
So what should facilities do to minimize the risk of suicide in the bathroom or shower? We’d recommend at least the following:
Some of our prior columns on preventing hospital suicides:
CDPH (California Department of Public Health). Complaint Intake Number CA00307558; 2016
CDPH (California Department of Public Health). Complaint Intake Number CA00313278; 2016
Fisher FC. Report to the Minister of Justice and Solicitor General. Public Fatality Inquiry. Province of Alberta (Canada); May 29, 2017
Lo T. Psych ward video cameras could save lives, suggests report after hospital bathroom suicide of Glenn Piche. CBC News 2017; August 21, 2017
Mental Health Environment of Care Checklist (VA)
Watts BV, Shiner B, Young-Xu Y, Mills PD. Sustained Effectiveness of the Mental Health Environment of Care Checklist to Decrease Inpatient Suicide. Psychiatric Services 2016; Published Online Ahead of Print: November 15, 2016
Mills PD. Use of the Mental Health Environment of Care Checklist to Reduce the Rate of Inpatient Suicide in VHA. TIPS (Topics in Patient Safety) 2016; 16(3): 3-4 July/August/September 2016