A four-item screening tool predicts 12-month interpersonal firearm violence risk among young adults presenting to emergency departments, with rates increasing from 2% to 25% across score categories in a multisite prospective study, researchers reported.
In a prospective longitudinal study of 1,506 patients aged 18 to 24 years, the SaFETy score identified increasing risk of firearm violence during 12 months of follow-up among patients seen in four level 1 emergency departments (EDs) across three US cities, wrote lead study author Jason E. Goldstick, PhD, of the University of Michigan in Ann Arbor, and colleagues.
Among 1,122 patients with follow-up data, representing 75% of the baseline sample, 73 (7%) experienced firearm violence within 12 months, defined as being shot or shooting someone. Firearm violence rates were 2% among patients with a score of 0, 12% among those with scores of 1 to 5, and 25% among those with scores of 6 or greater.
The score demonstrated moderate discrimination, with an area under the receiver-operating characteristic curve of 0.78. A threshold above 0 yielded 83% sensitivity and 62% specificity, while a threshold above 4 produced the highest positive predictive value at 32%.
Adding the SaFETy score to routinely available triage variables increased the area under the curve from 0.75 to 0.82. Adding the score to a model that also included baseline firearm violence, mental health screening results, and drug misuse increased the area under the curve from 0.81 to 0.84. Models without the score underestimated risk among patients with higher scores and overestimated risk among patients with a score of 0.
Performance was similar among male and female patients. By site, the area under the curve was 0.82 in Seattle, 0.81 in Philadelphia, and 0.66 in Flint. The score performed less well among patients with assault-related injuries at baseline, although that subgroup was small.
Methods and Findings
The Screening to Predict Young Adults at Risk for Firearm Violence study enrolled patients presenting to EDs for any reason and followed them for 12 months using self-report surveys and medical record review. The SaFETy score, which ranges from 0 to 10, includes four self-reported domains: serious fighting, friend weapon carrying, community exposure to gunshots, and firearm threats. Scores were grouped as 0, 1 to 5, or 6 or greater.
Limitations
The results relied primarily on self-reported measures, and some patients at highest risk were missing SaFETy scores or were lost to follow-up. The findings were based on urban ED settings and may not generalize to other clinical settings or populations.
Conclusion
“The SaFETy score predicts [firearm violence] risk in general samples of young adults in the ED,” the researchers wrote, adding that the score provides “distinct predictive information.”
Disclosures
The study was funded by the Centers for Disease Control and Prevention. Disclosure forms are available with the article online.
Source: American College of Physicians
