Emergency department screening tool predicts firearm violence risk

1. In this prospective longitudinal study, the SaFETy score effectively predicted firearm violence risk among young adults presenting to the emergency department.

2. A SaFTEy cutoff above 0 optimized sensitivity (83.1%) while maintaining moderate specificity (62.4%). 

Evidence Rating Level: 2 (Good)

Study Rundown: Emergency department (ED) visits represent a key opportunity for interpersonal firearm violence (FV) prevention. A non-intrusive, externally validated screening tool could help identify at-risk youth and guide targeted interventions. The 10-point SaFETy  (Serious fighting, Friend weapon carrying, community Environment, and firearm Threats) score has shown promise in predicting future FV risk following ED visits; however, its initial development in a single-site population with an overrepresentation of youth with violent injuries and substance use raises questions about its broader applicability.This study aimed to externally validate the SaFETy score in a larger, more diverse cohort of youth across multiple EDs. In this sample, most participants had low scores, while smaller proportions fell into moderate- and high-risk categories. Over the follow-up period, a minority of youth experienced FV; however, risk increased substantially with higher SaFETy scores, demonstrating a clear gradient across risk groups. The score showed good discrimination, with an area under the receiver operating characteristic curve (AUC) of approximately 0.79. When incorporated into a multivariable model with additional covariates, the SaFETy score improved overall predictive performance, although some miscalibration was observed, with underestimation of risk in low-score groups and overestimation in higher-score groups. Limitations include reliance on self-reported outcomes and greater missing data among higher-risk participants. Nonetheless, these findings support the SaFETy score as a useful tool for predicting FV risk among youth in ED settings, with further validation needed in community-based contexts.

Click to read this study in AIM

Relevant Reading: Development of the SaFETy Score: A Clinical Screening Tool for Predicting Future Firearm Violence Risk

In-Depth [prospective cohort]: This prospective longitudinal study aimed to externally validate the SaFETy score among general samples of youth in the emergency department (ED). Participants were recruited from level 1 trauma centers in Seattle, Washington; Philadelphia, Pennsylvania; and Flint, Michigan, and included patients aged 18 to 24 years who spoke English. Exclusion criteria included being in police custody, inability to consent, active suicidal ideation or psychosis, sexual assault, or droplet precautions early in the COVID-19 pandemic. The primary outcome was any firearm violence (FV) within 12 months, defined as reporting being shot or shooting someone else; self-reported FV within 6 months was also assessed. The 4-item SaFETy score (range 0–10) evaluates fighting frequency, number of friends who carry weapons, frequency of hearing gunshots, and frequency of receiving firearm threats in the past 6 months. Baseline surveys were completed by 1,506 participants (61% female; 42% Black, 34% White, 22% Hispanic). Most visits were for non-injury reasons, though 33% were injury-related, primarily motor vehicle crashes, and 3.8% were assault-related. Baseline SaFETy scores were 0 in 55.5% of participants, 1–5 in 37.3%, and ≥6 in 5.0%, with 2.2% missing scores. Among 1,122 participants with 12-month outcomes, 6.5% experienced FV. Twelve-month FV rates increased with baseline scores: 1.8% for 0, 12.1% for 1–5, and 25.0% for ≥6. The SaFETy score showed good predictive performance, with an AUC of 0.78. A cutoff above 0 optimized sensitivity (83.1%) and specificity (62.4%), while a cutoff of 4 optimized positive predictive value (31.6%). Site-specific performance was highest in Seattle and Philadelphia. Logistic regression models incorporating routine or full covariates improved discrimination when the SaFETy score was added, increasing AUCs significantly (from 0.75 to 0.82 and 0.81 to 0.84, respectively). Models underestimated risk for those with a score of 0 and overestimated risk for higher scores. Overall, these findings support the SaFETy score as a strong predictor of 12-month FV risk in youth presenting to the ED, with potential utility in guiding prevention efforts.

Image: PD

©2026 2 Minute Medicine, Inc. All rights reserved. No works may be reproduced without expressed written consent from 2 Minute Medicine, Inc. Inquire about licensing here. No article should be construed as medical advice and is not intended as such by the authors or by 2 Minute Medicine, Inc.