Article Text
PDF
Statistics from Altmetric.com
L Luck
Correspondence to: Ms L Luck, James Cook University, Queensland, Australia; lauretta.luck{at}jcu.edu.au
QUESTION
Which components of observable behaviour in patients, their families, and friends indicate a potential for violence toward nurses in the emergency department (ED)?
DESIGN
Instrumental case study using a concurrent mixed-method approach.
SETTING
33-bed ED in a public hospital in Australia.
PARTICIPANTS
20 ED nurses (90% women).
METHODS
Phase 1 comprised thematic analysis of 50 hours of unstructured participant observation, an unstructured interview with 3 nurses, and researcher journaling. In Phase 2, these findings provided items for a structured observation tool to collect quantitative data and informed the content for the qualitative interview guide. Qualitative data collection comprised 290 hours of participant observation on 51 separate occasions over 5 months (16 violent events were observed); 16 recorded, semi-structured, 45–60 minute interviews with nurses; 13 recorded, informal, and unstructured 30–40 minute field interviews, some of which occurred after a violent event was witnessed; review of organisational documents; and research journaling. Violent behaviour was defined as physical or non-physical (eg, abusive or threatening language).
MAIN FINDINGS
5 interconnected elements of observable behaviours that indicated a potential for violence in patients and accompanying persons were identified. The elements—staring and eye contact, tone and volume of voice, anxiety, mumbling, and pacing—are represented by the acronym STAMP. (1) Staring and eye contact. Staring was observed frequently and occurred in 9 violent events. Staring and glaring were considered to be a way of intimidating nurses into providing immediate attention and were often accompanied by pacing and restlessness. When nurses responded to these cues by engaging in conversation with patients and family or friends, escalation to violence did not occur. Absence of eye contact, unless culturally appropriate, was also associated with anger and passive resistance. (2) Tone and volume of voice. Raised voices, sarcastic comments, and urgent or demeaning speech patterns were important cues for violence and occurred in 13 violent events. (3) Anxiety. Anxiety in patients and family or friends was usually attributed to psychosocial, situational, and contextual stressors; comorbid health issues associated with confusion and disorientation; and lack of understanding of ED processes. Signs of anxiety included hyperventilation, flushed appearance, rapid speech, and confusion. Nurses frequently assessed for anxiety and intervened when anxiety escalated. Disorientation in patients could exacerbate anxiety and affect the ability to manage anxiety. Nurses who cared for such patients in quiet areas were able to avert violence. 13 of 16 violent events involved disoriented patients who could not be cared for in seclusion. (4) Mumbling. Mumbling under one’s breath was perceived as an indication of increasing frustration. It was sometimes associated with slurred or incoherent speech, which could signal intoxication or mental illness. Mumbling often involved negative statements about service, particularly wait times. When accompanied by staring or physical signs such as pacing, mumbling was seen as a very salient cue for violence. Mumbling occurred in 11 violent events. (5) Pacing. Pacing was seen as a sign of increasing agitation, as were staggering, flailing, or pulling out intravenous lines. Agitation occurred in 9 violent events.
CONCLUSION
5 observable behaviours of patients, their families, and friends in the emergency department were identified as cues for potential violence: staring and eye contact, tone and volume of voice, anxiety, mumbling, and pacing.
ABSTRACTED FROM
Luck L, Jackson D, Usher K. STAMP: components of observable behaviour that indicate potential for patient violence in emergency departments. J Adv Nurs 2007;59:11–9.
Commentary
Violence in healthcare settings, and against nurses in particular, is a significant problem internationally.1–4 Although differences in definitions and methodologies make comparisons difficult, these and other studies indicate that most nurses experience some form of assault at some point, with evidence that incidents of violence are increasing.
Luck et al conducted a single case study of a hospital in Australia, and so its applicability to other settings is limited. However, the authors offer a heuristic (represented by the acronym STAMP) that is useful not only for assessing for violence but also for de-escalating violence in individual patients. Importantly, the study showed how nurses were able to respond to patients displaying certain cues in ways that averted violence. Possible next steps may include development and validation of STAMP as a clinical prediction guide.
The study also highlights the importance of the issue of violence in healthcare workplaces, with implications for managers and policy makers. The findings are congruent with what is known about healthcare workplace environments and point to the relation between such environments and violence. For example, the findings show how anxiety was exacerbated by issues common to EDs, such as long waits and triage systems; how “mumbling” and aggressive statements were often about long waits; and how staring was a means of getting more immediate attention. The findings also show how providing care under stressful conditions to people with substance use and mental health issues exacerbated violence. While it is important to help individual nurses deal with individual patients, these results highlight the importance of dealing with the contextual features of healthcare settings that contribute to escalations of patient violence.
Footnotes
Source of funding: not stated.