INTRODUCTION: Aggression and violence are major concerns in emergency departments (EDs), and have negative consequences for patient and staff health and safety. Few validated tools exist for identifying patients at risk of agitation. This study conducted a systematic literature review to identify and summarize the scores that predict aggressive behavior in EDs.
METHODS: The search included articles published between Jan 1st, 1987, and Dec 31st, 2022, using the terms "aggress*," "violent*," "emergency," "acute," "score," or "scale."
RESULTS: Ten scores were found to be relevant, with eight of the developed scores intended for use in EDs. The Aggressive Behavior Risk Assessment Tool (ABRAT) was found to be sensitive (84.3%) and specific (95.3%). The Brøset Violence Checklist (BVC) was highly specific (99.4%), whereas the Violence Screening Checklist (VSC) was less sensitive (57.2%) and specific (45.7%). The violence and aggression (OVA)/BVC checklist was found to significantly decrease the number of security call activations (P < 0.001). The Behavioral Activity Rating Scale (BARS) and OVA/BVC scores were the shortest, with seven and six items, respectively.
CONCLUSION: The OVA/BVC checklist is a valuable tool for predicting and preventing violence in the EDs. Future prospective studies should investigate its effectiveness.
| Discipline Area | Score |
|---|---|
| Physician |  |
Physician rater
This systematic review of tools for predicting violent behavior in emergency department patients had limited searches that were done only in PubMed, PsycINFO, and MEDLINE. The search may have missed studies done outside the USA. Ten studies were found of which only 3 reported sensitivity and specificity values. Although the authors do not present likelihood ratios (LRs), they are easily calculated. For the ABRAT tool, the positive (LR+) and negative (LR-) LRs are 18 and 0.2, respectively; for the BVC tool, they are 76 and 0.5, respectively; and for the VSC tool, 1 and 0.9 respectively. Only one (the (OVA)/BVC checklist) was found to reduce security call activations.
, McMaster University