Early Warning Signs and Indicators of Potentially Violent Behavior

A number or reputable sources, including: OSHA’s Guidelines for Preventing Workplace Violence for Healthcare and Social Service Workers, the Emergency Nurses Association’s Workplace Violence Toolkit, and the American Organization of Nursing Executives, report a laundry list of signs which might indicate someone is on the verge of committing a violent act. These signs can be categorized from possible to imminent, and in stages of behavior from 1 through 3:

Stage 1: Representing early indicators of violence (i.e. challenging authority, defensiveness, excessive swearing, angry outbursts, and frequent signs of frustration.

Stage 2: Representing increased risk for hands-on violence (i.e. frequently argumentative, blatant disregard and disrespect, disruptive behavior in interactions with staff, suicidal threats, verbal threats towards other, displays of force through objects, offensive body posturing).

Stage 3: Representing extreme violent behavior, hands-on (i.e. physical assaults, display and use of weapons, prior criminal record of assaultive behavior, threats toward staff and erratic behavior). By identifying characteristic behaviors, staff can better prepare themselves should violence occur.8

Sharp (2015), suggested a brief overview of static factors common to the elevated risk of violence, including the following:

  • Male
  • Late adulthood
  • Low Socioeconomic Status
  • Instability in housing or gainful employment
  • History of violence or destruction of property
  • Mental or personality disorder
  • History of alcohol and/or substance abuse

A brief overview of dynamic factors indicating elevated risk of violence included:

  • Intoxication
  • Withdrawal from alcohol, opioids or other substances
  • Delirium
  • Psychosis, or paranoid delusions
  • Physical agitation, verbal aggression
  • Anger
  • Unmet pain management needs

Tools and Measures Used to Predict Violence

The Emergency Nurses Association’s 2011 Emergency Department Violence Surveillance Study reported that the greatest risk of violence in healthcare occurs in patient rooms, then corridors, halls, stairwells and elevators. The riskiest period is during the triage process, closely followed by patient restraint and/or invasive procedures.

In SEA #45, human resources-related factors, such as the increased need for staff education and competency of assessment processes, were noted in 60 percent of the causal factors of criminal events such as assaults in the workplace. Communication failures and deficiencies in safety and security procedures and practices were also cited. Assessments were noted in 58 percent of the events, particularly in the areas of flawed patient protocols, inadequate assessment tools, and a lack of behavioral health assessments.9 While many factors contributing to violence in healthcare may be beyond our control, assessments tools are not. There are a variety of tools and instruments available for clinicians to assess risks of violence from patients. The Clinical, Risk Management 20, as well as the Short-Term Assessment of Risk and Treatability tool, provide criteria for assessing risks. There is also a broadly used assessment known as the Bröset Violence Checklist, which uses key criteria to determine the level of risk that a patient may act violently within a relatively brief period. This one will be discussed in more detail. Many studies indicate that clinicians often overlook assessment tools, relying instead on gut feelings and personal intuition, which have been shown to result in grossly inaccurate assessments due to cognitive bias, confirmation bias, and human error.

  • Cognitive bias is a mistake in reasoning, evaluating, remembering, or other cognitive process, often occurring as a result of holding onto one’s preferences and beliefs regardless of contrary information
  • Confirmation bias is the tendency to seek only information that matches what one already believes

Cognitive biases can greatly hamper a healthcare worker’s ability to more accurately assess the risk of violence from patients. Confirmation bias is one such example, whereby there is a tendency to focus heavily on details aligned with what a person already expects or perceives to be a known threat, while ignoring the elephant in the room. A second kind of cognitive bias is when one assigns more weight to a particular event (i.e. multiple casualty events), while assigning a lesser value to another event (i.e. falls during inclement weather), which can lead to a marked oversight in judgement and proper assessment of risks. In other words, when you hear hoofbeats, think horses not zebras.10

The Bröset Violence Checklist (BVC) is an instrument that has been used to assess the short-term risk of violent attacks in behavioral health areas. In its application, the following variables are noted in patients:

  • Confused
  • Irritable
  • Boisterous
  • Physically or verbally threatening
  • Attacking inanimate objects

Using a measure to record observed behavior over time, a risk of violence score is assigned from 0 to >2 equating to risks from small to moderate to high:

  •  Score of 0 indicates the risk of violence is small
  •  Score of 1-2 indicates the risk of violence is moderate; preventative measures should be taken
  •  Score of >2 indicates the risk of violence is high. Preventative measures should be taken and plans should be developed to manage the potential violence

Equipped with a record of information and logs of a patient’s exhibited behaviors, staff are able to put appropriate safety measures, staffing, and other resources in place to mitigate the effects of patient violence over a 24-hour period following the previous assessment.11, 12

The following data has been reported on the use of the BVC toolkit:

  • Introducing twice-daily staff measures on risk assessment (BVC Swiss version)
    • 41 percent reduction in severe violent incidents
    • 27 percent reduction in the use of coercive measures
  • Implementing regular risk and violence assessment
    • 68 percent reduction in aggressive incidents
    • 45 percent reduction in time in seclusion
    • 48 percent reduction in violent incidents