Validity and reliability of the novel three-item occupational violence patient risk assessment tool.

The content validity reviewers were anonymous, so the researchers could not tell how representative they were of emergency department staff. The predictive validity analysis used triage notes, so it depended on what was documented and may have missed risk factors that were not recorded. Risk factors after triage were not captured, because the design was retrospective. The tool was developed and tested in one emergency department, so generalisability is limited. The study did not test whether using the tool reduces violence, injuries or restraints; that evaluation is planned after implementation.

The authors declared no conflict of interest. Funding was from the Health Innovation, Investment and Research Office and the University of New South Wales.

Do not conclude that the tool can reliably predict all violent patients. Each single risk domain identified only 22%-55% of violent patients, and the study relied on triage notes, so missing documentation can make a patient look lower risk than they are.