Amidst the quest to protect patients from healthcare-associated harm, acts of patient self-harm may challenge even the best-laid safety plans in unanticipated and potentially devastating ways. A query of the Pennsylvania Patient Safety Reporting System database for reports submitted from July 1, 2015, through June 30, 2018, using the event type "Patient Self-Harm" identified 624 reports describing a spectrum of events involving intentional bodily harm outside of psychiatric settings—from possession of means to self-harm to completion of a self-harm act. Emergency departments and medical-surgical units were the most common event locations, and ingestion accounted for more than 40% of the reports describing self-harm acts. A list of risk reduction strategies and resources, including the Pennsylvania Patient Safety Authority's ligature-risk assessment initiative, is provided.
Patients in hospitals sometimes harm themselves intentionally. These acts, defined broadly as self-harm, may have varying goals and outcomes—the most devastating being suicide, a "never event" in the eyes of the healthcare community at large.1 This safety risk has been compounded by the rise in mental illness and substance abuse and the need to care for many of these vulnerable individuals in nonpsychiatric settings.2-4 To explore this topic further, an analysis of self-harm events in the nonpsychiatric setting was conducted using the Pennsylvania Patient Safety Reporting System (PA-PSRS) database.
Pennsylvania Patient Safety Authority analysts queried the PA-PSRS database for event reports submitted from July 1, 2015, through June 30, 2018, using the taxonomy's event type "Patient Self-Harm." This time period represents the first three complete academic years (i.e., July through June) since patient self-harm was added as a discrete PA-PSRS event type in April 2015.5 Reports from psychiatric facilities, psychiatric units, chemical dependency units, and outpatient psychiatric clinics were excluded due to differences in staff training, expertise, and the care environment.
Analysts manually reviewed 890 report narratives and applied the following exclusion criteria:
- Insufficient detail to suggest that the act was intentional (e.g., delirious patient pulled out catheter), n = 117
- Incidental harm to self (e.g., patient walked into a table), n = 103
- Not a direct act of damage to one's body (e.g., elopement, discharge against medical advice), n = 40
- Insufficient detail in report to analyze event, n = 4
- Violent threat or act toward others, n = 1
- Description of an event occurring in a psychiatric care area, n = 1
The remaining reports were analyzed by harm score,6 facility type, patient age, and patient gender. Reports submitted by hospitals were further analyzed to determine the care area and whether the event occurred off the unit or outside of the facility (i.e., after discharge, during elopements, or between outpatient encounters).
Analysts categorized each report using four self-harm categories, as described in Table 1.
|Means||Possession of an item that could be used for a self-harm act, but was identified before use||Razor discovered and confiscated|
|Threat||Verbal statement or gesture expressing a desire to engage in a self-harm act||"I'm going to kill myself"|
|Attempt||An effort to initiate a self-harm act that was stopped by staff or patients themselves||Patient attempted to ingest battery and was stopped before it was swallowed |
|Act||Initiation of a self-harm behavior that may be partially or completely executed, and may or may not require medical attention||Patient cut wrist with razor; patient found with phone cord wrapped around neck|
Reports in the self-harm "act" category were further characterized by mechanisms of injury, as described in Table 2. A single report could involve multiple mechanisms of injury.
MECHANISM OF INJURY||
|Ingestion||Includes oral, nasal, injection (not via peripheral or central catheter). Includes intentional ingestion of illicit substances, regardless of ability to determine intended outcome based on the report.|
|Laceration, puncture, scratch||Includes cutting and stabbing.|
|Blunt injury||Impact or collision with a surface, object, or part of one's body.|
Strangulation—divided into three subtypes:
Applied the definition provided by Sauvageau and Boghossian:* "Asphyxia by closure of the blood vessels and/or air passages of the neck as a result of external pressure on the neck."
"Pressure on the neck is applied by a constricting band tightened by a force other than the body weight."
"External pressure on the structures of the neck by hands, forearms, or other limbs."
"Pressure on the neck is applied by a constricting band tightened by the gravitational weight of the body or part of the body." Includes partial or complete suspension.
|Intravenous (IV) or medical device tampering||Includes injection into peripheral and central lines and manipulation of devices such as IV pumps. Includes intentional ingestion of illicit substances, regardless of ability to determine intended outcome based on the report.|
|Insertion of foreign object||Includes non-oral orifices, such as rectum or urethra.|
|Other||Mechanism of injury that falls outside of the categories above, such as anorexia, bulimia, drowning, wound interference, biting.|
Unable to determine
|Insufficient detail in report to further categorize self-harm act mechanism of injury.|
* Sauvageau A, Boghossian E. Classification of asphyxia: the need for standardization. J Forensic Sci. 2010 Sep;55(5):1259-67. Also available:
http://dx.doi.org/10.1111/j.1556-4029.2010.01459.x. PMID: 20561144 |
Analysts conducted a review of the literature to collate relevant background and risk reduction strategies related to self-harm in nonpsychiatric settings.
The query identified 890 reports, of which 266 were excluded, leaving 624 reports for further analysis.
More than 80% of reports were submitted as Incidents (80.9%; n = 505 of 624; Figure 1). The most frequent harm score was D, an event that required monitoring to confirm that it resulted in no harm and/or required intervention to prevent harm (49.2%; n = 307).6 Of the 119 Serious Events, events that contributed to or resulted in temporary harm and required treatment or intervention were most common (i.e., harm score E).6 Ten (1.6%) of the reported events resulted in death.
Facility Type and Care Area
Hospitals submitted more than 99% of the reports (99.5%; n = 621 of 624). Within hospitals, the highest number of reports occurred in the emergency department (37.5%; n = 233 of 621), followed closely by medical-surgical units (34.0%; n = 211; Figure 2). Twenty reports described events occurring outside of the hospital, such as after discharge, during elopements, or between outpatient encounters. These reports were submitted under several different care areas including emergency department and other, and accounted for 4 of the 10 reports that resulted in death (i.e., harm score I).
Gender and Age Cohort
Figure 3 shows the distribution of reports by age cohort and gender. The age cohort 25 to 34 years had the highest number of reports for both males and females (27.4%; n = 171 of 624). The overall number of reports was nearly evenly split between males and females (49.4% and 50.6%, respectively).
The distribution of reports by the 4 self-harm categories assigned by analysts, described in Table 1, is as follows: 3.4% means (n = 21 of 624), 5.4% threat (n = 34), 2.7% attempt (n = 17) and 88.5% act (n = 552).
Figure 4 displays the mechanism of injury for the 552 reports involving self-harm acts.
More than 40% of the self-harm acts involved ingestion (41.8%, n = 231 of 552); examples of specific substances include illicit drugs, hand sanitizer, and batteries (e.g., from hearing aids, telemetry monitors). Ingestion of illicit drugs accounted for 4 of the 10 reports that resulted in death; injection of unknown substances into IV devices, categorized as "IV or medical device tampering," accounted for an additional 2 reports resulting in death.
Laceration, punctures, and scratches occurred in 22.1% of the self-harm acts (n = 122); examples of specific instruments used include plasticware, razor blades, and caps (e.g., toothpaste, marker, lotion).
Blunt injury, such as head banging, hitting self, and punching walls, occurred in 19.7% of the self-harm acts (n = 109).
Strangulation occurred in 11.6% of self-harm acts (n = 64). Examples of materials used for the most common strangulation type, ligature, include linens (e.g., sheets, blankets, pillow cases), cords (e.g., telemetry, call light, phone), and patient garments (e.g., belt, pants, paper scrubs). Four out of seven strangulations by hanging involved patient garments, and the most common ligature point was a door, including frames and hardware.
Themes from Event Narratives
Event reports described supervision of patients at high risk of self-harm by both hospital personnel, such as sitters, and external personnel, such as correctional officers. Reports demonstrate mixed effects of supervision, including prevention, early interception, and delayed detection.
An example of an event report demonstrating how supervision prevented a self-harm act is as follows*:
1:1 sitter observed the patient trying to wrap the monitor cord around her neck. Patient was stopped.
Maintaining a line of sight on patients, even while in the bathroom, is an important safety intervention. The following is an example of an event report describing deviation from this practice:
When 1:1 sitter checked on the patient in the bathroom, the patient was found to be cutting himself with a toothpaste cap.
Event reports also describe detection of self-harm acts by remote monitoring; examples are as follows:
Security officer saw patient on camera with gown tied around neck.
Telemetry alarming for high heart rate. Found patient with phone cord tied around his neck.
Event reports describe several roles played by visitors including as reporters, inhibiters, and enablers of self-harm means, attempts, and acts.
An example of an event report in which a visitor prevented an act of self-harm is as follows:
The patient attempted to cut herself with a soda can tab but husband stopped her.
An example of an event report in which a visitor enabled self-harm means is as follows:
Patient ingested unknown item. Oxycodone pills were found in bag brought in by the patient's friends.
Threats Inside and Outside of the Patient Room
Event reports describe a variety of facility locations, both inside and outside of the patient room, where self-harm attempts and acts occurred.
An example of an event occurring in a classic patient room location—the bathroom—is as follows:
Patient cut right arm with a plastic knife from her food tray while in the bathroom.
An example demonstrating the need to evaluate self-harm risks outside of patient rooms—such as hallways and workstations—is as follows:
Patient swallowed a push pin she obtained while walking in the hallway with staff.
* The details of the PA-PSRS event narratives in this article have been modified to preserve confidentiality.
Despite mandatory reporting laws, PA-PSRS data are subject to the limitations of self-reporting, including the complexities of selecting the appropriate harm score and care area. Analyst ability to categorize reports as self-harm means, threats, attempts, and acts was limited to the information provided in the event narratives. Although this analysis made every attempt to exclude unintentional harm to self, the inability to differentiate reports by suicidal and nonsuicidal intent is an acknowledged limitation. Illicit drug use, regardless of outcome (e.g., high, overdose), was included in this self-harm data set because of the intentional nature of engaging in this self-destructive behavior. This analysis does not include reports of suicide attempts that did not result in an injury requiring additional healthcare services, which are submitted through PA-PSRS as Infrastructure Failures.7
Risk Reduction Strategies
The following risk reduction strategies come from event report recommendations and from the literature.
- Assess nonpsychiatric staff members' comfort, knowledge, and competence caring for patients at risk of self-harm. Based on each role's scope of practice, education may focus on dialoguing with patients about self-harm behaviors, identifying environmental hazards, using de-escalation techniques, or implementing appropriate risk reduction strategies.8,9
- A typical emergency department bay or medical-surgical room, unless designed specifically to accommodate a patient at risk of self-harm, may be an unsafe environment. Use a checklist to inspect rooms of at-risk patients to identify and remediate environmental hazards, such as unsecured supplies and long tubing or cords. Conduct inspections prior to admission and on regular intervals throughout the patient's stay (e.g., every shift).10
- Establish screening and assessment guidelines that support early identification of patients at risk for self-harm.8,11 Consider implementing universal screening, followed by assessment if screening is positive, at all points of entry, including the emergency department.12
- Conduct a proactive ligature risk assessment. The Authority has numerous resources on this topic and is available to conduct proactive ligature risk assessment training for Pennsylvania facilities. For more information and a list of resources see
Patient Safety Topics: Behavioral Health.
- Incorporate attention to ligature points and hazardous materials in the design of patient rooms—when possible, eliminate unnecessary risks. For a short video tour demonstrating examples of specific design elements, see
Up Front: Facility and Fixture Design Help Protect Suicidal Patients.
- Provide training for observers of high-risk patients. Encourage therapeutic interactions that go beyond just observing, such as having conversation or doing an activity with the patient. See the International Association for Healthcare Security and Safety's (IAHSS) resource,
Guidance for Training for Observers of High-Risk Patients, which addresses training needs of both direct and remote patient observers.
- Ensure facility protocols address particularly sensitive safety issues such as the role of external patient monitors (e.g., police officer, correctional officer), restrictions and terms of visitation, and searching and securement of patient belongings.11
- Review the Joint Commission's
National Patient Safety Goal 15.01.01 for suicide prevention, effective July 1, 2019, for psychiatric hospitals, psychiatric units, and patients with a primary psychiatric condition in general hospitals, to identify evidence-based best practices that could be applied in nonpsychiatric settings that may care for patients with psychiatric comorbidities.
The prevention, detection, and response to patient self-harm are serious safety concerns in nonpsychiatric settings. This analysis demonstrates vulnerabilities—in the environment of care, workflow, procedures, and staff training—that may be overlooked or unanticipated. The intent of this analysis was to share the lessons learned by healthcare facilities in Pennsylvania to raise awareness and encourage a proactive approach to risk assessment.
- Patient safety primer: never events. [internet]. Rockville (MD): Agency for Healthcare Research and Quality (AHRQ); 2018 Aug [accessed 2018 Oct 16]. [6 p]. Available:
- Managing mental health care at the hospital. Hospitalist. 2017 Dec 7; Also available:
- National Center for Health Statistics. Health, United States, 2017 - trend tables: Table 50. Use of selected substances in the past month among persons aged 12 and over, by age, sex, race, and Hispanic origin: United States, selected years 2002-2016. Atlanta (GA): Centers for Disease Control and Prevention; 2018. 2 p. Also available:
- Curtin SC, Warner M, Hedegaard H. Increase in suicide in the United States, 1999-2014. NCHS Data Brief. 2016 Apr;241 Also available:
- 2016 Pennsylvania Patient Safety Authority annual report. Harrisburg (PA): Pennsylvania Patient Safety Authority; 2017 Apr 28. 103 p. Also available:
- Pennsylvania Patient Safety Authority harm score taxonomy. Harrisburg (PA): Pennsylvania Patient Safety Authority; 2015. 1 p. Also available:
- Pennsylvania Patient Safety Authority. Frequently asked questions on: The Final Guidance for Acute Healthcare Facility Determinations of Reporting Requirements under the Medical Care Availability and Reduction of Error Act. Harrisburg (PA): Pennsylvania Patient Safety Authority; 2015 Sep. 44 p.
- The Joint Commission. Detecting and treating suicide ideation in all settings. Sentinel Event Alert. 2016 Feb 24;56:1-7. Also available:
- Manister NN, Murray S, Burke JM, Finegan M, McKiernan ME. Effectiveness of nursing education to prevent inpatient suicide. J Contin Educ Nurs. 2017 Sep 01;48(9):413-9. Also available:
http://dx.doi.org/10.3928/00220124-20170816-07. PMID: 28850658
- Use checklist to reduce self-harm risks in the ER. Brief Hosp Saf. 2018 Jul;:9-10. Also available:
- ECRI Institute PSO. ECRI Institute PSO deep dive: meeting patients' behavioral health needs in acute care: strategies. [internet]. Plymouth Meeting (PA): ECRI Institute; 2018 Oct 5 [accessed 2018 Nov 08]. Available:
- Roaten K, Johnson C, Genzel R, Khan F, North CS. Development and implementation of a universal suicide risk screening program in a safety-net hospital system. Jt Comm J Qual Patient Saf. 2018 Jan;44(1):4-11. Also available:
http://dx.doi.org/10.1016/j.jcjq.2017.07.006. PMID: 29290245