Workplace violence against healthcare staff is a major concern in clinical environments. Incidents often involve patients or visitors and arise in settings where staff manage distress, behavioural health conditions and complex care needs. Monitoring has often relied on surveys, focus groups and interviews, which can be hard to sustain at scale.
Patient safety event (PSE) reports provide an additional lens. Routinely collected in many health systems, they combine structured fields with narrative descriptions of harms and safety incidents. Systematic analysis can support classification of workplace violence and clarify patterns across care settings.
Using Safety Reports to Detect Workplace Violence
A structured classification approach was applied in a 10-hospital health system in the mid-Atlantic region of the United States. The system includes 20 entities and more than 300 outpatient services, spanning urgent care, emergency services and primary and specialist care. Staff submit safety incidents through a centralised reporting platform intended to capture workplace violence alongside other events.
Must Read: Workforce in Transition: Addressing the Pressures on Nurses
Reports filed between 1 March and 20 September 2023 formed the dataset. There were 15,429 reports submitted and 15,426 contained an incident description. Using three structured fields, 975 reports were flagged as potential workplace violence incidents (6.3% of eligible reports). Flagging used event-type categories such as workplace violence, verbal abuse, physical assault and disorderly conduct, supplemented by two questions on whether the event was related to workplace violence and whether anyone was physically harmed due to workplace violence.
Two investigators manually reviewed free-text descriptions for 300 of the 975 flagged reports. Using definitions aligned with the Joint Commission, they refined criteria for classifying incidents as workplace violence and sought expert input when classification was uncertain. The agreed approach was then applied to the remaining flagged reports.
Confirmed Incidents and Forms of Harm
Of the 975 flagged reports, 831 met at least one condition in the Joint Commission definition of workplace violence (85.2%). All 831 also had at least one workplace violence related event type coded in structured data. Narratives supported further coding in 809 incidents (97.4% of confirmed cases), including incident type, harm, exposed individual and perpetrator.
Patient- or visitor-on-staff incidents were most frequent, accounting for 637 of 831 events (76.7%). Staff-on-staff workplace violence accounted for 190 incidents (22.9%), with a small number that could not be clearly categorised. Verbal violence was the leading form of harm in 331 incidents (39.8%). Multiple forms of violence were also common. Physical violence as a distinct category appeared in 14.8% of reports. Sexual harassment was less common. Other harms included aggression and behaviours such as recording staff without consent, with suspected or threatened weapon use described in 18 incidents (2.2%).
Patients were the most frequently named perpetrators (581 incidents, 69.9%). Visitors and family members were identified in 55 incidents (6.6%). Staff or other individuals were linked to 190 incidents (22.9%), alongside a small group where the perpetrator was unclear.
Who Was Exposed, Where Incidents Occurred and What Triggered Them
Nurses were most often named as the primary exposed person (277 incidents, 33.3%). Other incidents involved multiple staff members, licensed independent clinicians, security officers or patient sitters. A large “other” category accounted for 377 incidents (45.4%) and included roles such as radiology technicians and registration staff, indicating that workplace violence affects a wide range of clinical and non-clinical personnel.
Patient sitters, who provide one-to-one care for patients with conditions such as delirium or dementia, were highlighted as a group with notable exposure. Reports raised concerns that some may see violence as an expected part of the role and be reluctant to report incidents or seek support.
Most confirmed incidents occurred in hospital settings. Of the 831 cases, 767 incidents (92.3%) were reported from hospitals. Smaller proportions arose in urgent care, primary and ambulatory care, physical therapy or rehabilitation medicine and other locations including specialist offices, infusion centres and radiology departments.
Registered nurses and licensed practical nurses submitted 533 workplace violence reports (64.1%), while other patient care staff contributed 10.5%. Additional reports came from physicians, imaging personnel, rehabilitation professionals, pharmacy staff and general staff.
Security officers were contacted in 391 incidents (47.1%), not contacted in 201 (24.2%) and unknown or not recorded in 239 (28.8%). Law enforcement was contacted in 70 incidents (8.4%) and not contacted in 488 (58.7%), with the remainder unknown. Narratives described security officers being drawn into situations that had become unmanageable, including severe substance use problems, withdrawal symptoms, use of restraints or administration of medications, increasing the risk of exposure to workplace violence.
Precipitating factors were documented for about half of the confirmed incidents. Agitation was recorded in 193 incidents (23.2%) and aggression in 179 incidents (21.5%). Behavioural health diagnoses were present in 79 incidents (9.5%). Narratives also described cognitive impairment, confusion or disorientation, substance use or intoxication, psychiatric crisis, delirium and pain. Other precipitating circumstances included dissatisfaction with searches, emotionally intense events such as the death of a loved one and disagreements between staff. For 414 incidents (49.8%), precipitating factors were unknown or not recorded.
A structured approach to classifying workplace violence in routine PSE reports identified 831 incidents over six months, predominantly involving patient- or visitor-on-staff aggression and most often presenting as verbal abuse or multiple forms of harm. Exposure extended across nursing, patient sitters, security officers and many other roles, mainly in hospital settings. Security involvement was common, and agitation, aggression and behavioural health factors were often recorded, although information on triggers was frequently incomplete.
Source: JAMA Network Open
Image Credit: iStock
