CHPR: Please start by telling us a little about yourself.
Dr. Mesbah: I’m an assistant professor of emergency medicine at Baylor College of Medicine in Houston, Texas. My research focuses on behavioral health patients—specifically, predicting aggression and improving safety for patients in the emergency department (ED). I recently published a systematic review evaluating all the validated tools used to predict violent behavior in EDs (Mesbah H et al, Am J Emerg Med 2024;80:44–50), and my team is currently working on developing and validating new tools for predicting aggression.

CHPR: What prompted you to study this area?
Dr. Mesbah: Two reasons. First, the ED is a unique, fast-paced environment with high patient volume. It’s critical for ED staff to identify patients who are at high risk for aggression, not only for the patient’s safety but also for the safety of other patients and staff. We know that mental health presentations now make up more than 12% of all adult ED visits in the US, which highlights how common and pressing this issue is (Peters ZJ et al, National Health Statistics Reports 2023;181). Second, there’s no consensus on the most reliable screening tool for predicting aggressive behavior, so my team wanted to explore this question.

CHPR: Can you tell us about the tools that are available?
Dr. Mesbah: In our review, we identified 10 tools. The Brøset Violence Checklist (BVC) was the most promising tool we found for EDs (Woods P and Almvik R, Acta Psychiatr Scand Suppl 2002;412:103–105). It’s a six-item checklist of observable behaviors, like confusion, irritability, and verbal or physical threats, and it was designed for 24-hour observation. Its specificity is extremely high at over 99%. But on the downside, its sensitivity is under 50%, meaning it misses a substantial number of patients who later become aggressive. Still, when the BVC does flag someone, it’s very meaningful, as patients who score 3 or more are over 70 times more likely to become violent within 24 hours (Partridge B and Affleck J, Australas Emerg Care 2018;21(1):31–35). In Australia, the BVC has been incorporated into a structured decision-making tool called the Occupational Violence Assessment (OVA) score. It uses the same items but adds action steps (eg, “If score is 1, do X; if 2, notify the doctor; if 3, call security”) (Senz A et al, Emerg Med Australas 2021;33(4):665–671).

CHPR: Is the OVA used in the US also?
Dr. Mesbah: While the OVA itself is not routinely used in the US, it illustrates how a brief risk score can be paired with clear response protocols—an approach we could incorporate into our own agitation or workplace violence policies.

CHPR: And can you tell us about some of the other tools you reviewed?
Dr. Mesbah: Another tool is the Dynamic Appraisal of Situational Aggression (DASA). It’s essentially the BVC’s more sophisticated cousin. While the BVC gives you six observable behaviors and a straightforward score, the DASA adds a seventh item—impulsivity—and has been validated across more settings, including forensic units, general inpatient psych, and even medical-surgical floors. Research from the original validation study found that for each one-point increase in DASA score, the odds of aggression within 24 hours increased roughly threefold. What makes the DASA clinically useful is that it doesn’t just predict violence; it gives you a dynamic score that changes shift to shift, which means you can track whether de-escalation efforts are actually working. That said, in a busy ED where you have five minutes to triage a patient, the BVC’s simplicity often makes it more practical as the DASA requires more clinical judgment on some items, like “sensitivity to perceived provocation,” which can be difficult to assess during rapid turnover. There’s also the Behavioral Activity Rating Scale (BARS), which measures a patient’s current agitation level, from difficult-to-rouse to violent (Swift RH et al J Psychiatr Res 2002;36(2):87–95). It’s not predictive, but it is useful for assessing real-time agitation.

CHPR: Are there tools that incorporate more historical risk factors?
Dr. Mesbah: Yes. The Aggressive Behavior Risk Assessment Tool (ABRAT) is a brief checklist that combines behavioral items plus historical factors like prior incidents of aggression as well as psychiatric diagnoses. It was originally developed to predict violence in medical–surgical units, but it’s also been tested in EDs, where it had a sensitivity of 84% and specificity of 95% (Kim SC et al, J Am Coll Emerg Physicians Open 2022;3:e12693).

From the Expert Q&A
“Violence Prediction and Prevention Tools”
With Heba Mesbah, MD, PhD
The Carlat Hospital Psychiatry Report
Volume 6, numbers 3 and 4
April/May/June 2026

CHPR: Do the tools predict violent behavior across different populations like adolescents and older adults?
Dr. Mesbah: We didn’t find major differences based on the population studied, but we did see differences across settings. Some tools perform best in inpatient units, while others are more useful in busy EDs. For example, some of the ABRAT items require information that’s hard to obtain in an ED, such as detailed psychiatric history, so it generally ends up being a better fit for use in inpatient psychiatric or medical–surgical units than in general EDs. The setting matters more than the patient’s age in terms of how well a tool performs.

CHPR: It sounds like some of these tools could be useful for inpatient psychiatric units.
Dr. Mesbah: While my clinical focus is in emergency medicine rather than inpatient psychiatry, many of these tools, including the DASA, ABRAT, and BVC, were originally studied in inpatient settings. And they fit well where teams can reassess behavior over time. The same caveat applies: They support, but don’t replace, clinical assessment and good team communication.

CHPR: Is there any value in using more than one tool?
Dr. Mesbah: We didn’t find evidence supporting the use of multiple tools. Using more than one can be confusing and may produce conflicting results. A single tool that’s simple to use and accurate is preferable.

CHPR: How do these tools compare with clinical judgment?
Dr. Mesbah: A tool should be used together with clinical judgment, not in place of it. The tools offer a structured way to support clinical judgment and create a shared language among staff. Clinicians still need to use their judgment when interpreting scores and choosing interventions. And especially in EDs, where we often don’t have time to build rapport or gather a detailed history, having a brief, structured score at hand can make that judgment more efficient and more consistent across providers.

CHPR: What are the implications of false positives? Could staff overreact?
Dr. Mesbah: False positives aren’t a failure—they’re a form of prevention. If we believe a patient may become aggressive, we can take early steps: moving them to a quieter room, increasing observation, involving de-escalation teams, offering medication proactively, or alerting security. These measures help prevent escalation.

CHPR: Have these tools been shown to reduce adverse outcomes beyond violent incidents, like lowering the rates of staff injuries?
Dr. Mesbah: We found that when EDs used the OVA tool—which is just the BVC with some action steps attached—security callouts dropped significantly (Senz et al, 2021). That’s the kind of outcome that matters day-to-day in an ED: Having to hit the panic button fewer times means things are getting de-escalated earlier, which probably means fewer staff getting hurt, though the review didn’t track injuries directly. In a separate ED project that embedded the DASA into routine rounding protocols, staff injuries decreased by 68% after the DASA was introduced (Olshan-Perlmutter M et al, Standardizing Behavioral Health Emergency Department Nursing Rounds Incorporating Diagnostic Sensitivity of the Dynamic Appraisal of Situational Aggression to Improve Workplace Violence [Poster presentation]. Advocate Health Midwest Region Nursing Research & Professional Development Conference 2024; virtual.).

CHPR: When someone is identified as moderate or high risk, what steps should staff take?
Dr. Mesbah: The OVA is currently the only tool that builds in specific action steps. When a patient scores in the moderate-risk range, the OVA recommends moving the patient to a quieter or more observable room, and adjusting the environment by removing any objects that the patient could use as a weapon. It also calls for an increase in nursing observation and to notify the physician so the team can begin verbal de-escalation and offer medication if appropriate. Security should then be alerted to be on standby, and the psychiatric team should be contacted to expedite evaluation or disposition when possible. Taken together, these proactive steps can significantly reduce the likelihood of a violent incident occurring.

CHPR: Might flagging a patient as high risk bias how staff approach them?
Dr. Mesbah: Formalizing violence risk with tools like the BVC and DASA actually helps counteract bias by replacing subjective “gut feelings” with objective, observable behavioral criteria. These tools don’t create a permanent label—they are dynamic assessments that track a patient’s current state. As scores drop and the patient stabilizes, staff receive clear data to scale back precautions and reduce restrictive measures. Rather than labeling a “dangerous person,” the goal is to trigger specific clinical supports. In practice, studies show these approaches can lead to less restrictive care, including a 55%–75% reduction in restraints and a 68% decrease in staff injuries (Senz et al, 2021; Olshan-Perlmutter et al, 2024).

CHPR: Do staff need any special training to use these tools?
Dr. Mesbah: Training is helpful so that everyone can interpret behaviors consistently. Some checklist items are subjective, like “boisterous” or “verbally threatening,” so it’s important that all nurses, doctors, and other staff members are on the same page. Formal training programs that specifically focus on the use of violence prediction tools are not yet universal, but they are becoming increasingly common. In some hospitals, tools like the BVC are incorporated into nursing orientation and reinforced through periodic refreshers. Some institutions also incorporate simulation-based training in aggression recognition and de-escalation, allowing staff to practice applying risk assessments in realistic scenarios and receive structured feedback.

CHPR: Might more experienced staff interpret behaviors differently than newer clinicians?
Dr. Mesbah: We’re in fact studying how a person’s level of experience and background might affect how they interpret these scores, such as whether a staff member has personally experienced being assaulted by a patient. Some research is now using AI and machine learning to pull charts—like millions of charts—to give us more information about the factors that predict violent behaviors.

CHPR: You mentioned your team is developing a new tool. Can you say more?
Dr. Mesbah: We want to expand beyond just observable behaviors. We’re looking at factors like prior aggression, psychiatric diagnoses, substance use, and arrival circumstances (such as whether the police were involved). Patient characteristics like homelessness or insurance status may also contribute. We’re seeking funding to develop and validate a tool that incorporates these variables.

CHPR: So perhaps we’ll eventually see a Mesbah violence prevention tool?
Dr. Mesbah: Not necessarily with my name on it!

CHPR: Either way, we look forward to seeing what your team develops. Thank you for your time, Dr. Mesbah.