Background: The Fordham Risk Screening Tool (FRST), a five-item clinical decision-making tool, was developed to streamline violence risk assessment and more efficiently screen patients for whom greater violence risk workup is warranted [1]. Because treating patients with mental health needs and working in the emergency department are both risk factors for experiencing healthcare violence, the utility of the FRST is especially promising in the emergency evaluation of patients in mental health crisis [2]. In 2023, the Maine Medical Center (MMC) FRST research team found that the FRST was 84% sensitive and 93% specific in screening psychiatric patients who later tested positive for risk of violence using the current gold standard violence risk assessment tool [3]. No violence risk assessment screening tool has ever been directly connected to actual violent outcomes. The MMC FRST research team recently sought to explore whether FRST results correlated with markers of violence during inpatient psychiatric hospitalization. In conducting this evaluation, the MMC FRST research team observed that roughly 10% of the study participants experienced multiple psychiatric hospitalizations during the study enrollment period. Objective: To analyze the differences in FRST performance and inpatient violence incidence between patients in the MMC FRST study who experienced multiple inpatient psychiatric admissions (multiple-admission patients or “MAP’s,” n=53) and patients in the study who experienced only one inpatient psychiatric admission (single-admission patients or “SAP’s,” n=366). Hypotheses: (1) MAP’s will show a higher risk of violence as tested by the FRST compared with SAP’s and (2) MAP’s will demonstrate violent inpatient behavior at a higher rate than SAP’s. Design/Methods: Cohort study with admission status as exposure and FRST performance and violence incidence as outcomes. MAP’s and SAP’s were defined as patients who experienced >1 or 1 inpatient psychiatric admission during the six-month study enrollment period, respectively. Because each MAP had 2-3 FRST test results and 2-3 violence records throughout the course of the study, their performances were consolidated into one equally weighted aggregate score to be compared with the single FRST result and single violence record of each SAP. Statistical analysis was completed using IBM SPSS software. Results: MAP’s appeared more likely than SAP’s to screen positive for violence risk using the FRST, with FRST positivity rates of 37.0% and 25.5%, respectively (RR: 1.45, 95% CI: 0.97, 2.17). Additionally, MAP’s tended to show higher rates of violence incidence across nine violence metrics compared to SAP’s (RR: 1.19, 95% CI: 0.89, 1.60). Conclusions: These results appeared to be consistent with initial hypotheses regarding FRST performance and violence incidence, but the data did not reach the threshold for statistical significance. Trends present in these data warrant follow-up work to increase statistical power and determine if MAP’s are in fact at greater risk of both testing positive with the FRST and demonstrating inpatient violence compared to SAP’s. Further research could improve the ability of clinicians to use the FRST to anticipate violent behavior in psychiatric patients who experience repeated hospitalizations in short timeframes.
Learning Objectives:
Understand the available research regarding violence risk screening tools.
Understand how violence risk screening tools, such as the FRST, can be applied in clinical settings.
Understand the potential utility and limitations of implementing use of a screening tool such as the FRST.
