Missed Recognition of Potential Abuse in Young Children Presenting to the Emergency Department: A Retrospective Study

Aim To determine the prevalence of abuse‐related red flags and the frequency of subsequent evaluations among young children presenting to the emergency department (ED) with trauma‐related complaints, and to quantify the association between red flags and further evaluation. Methods We conducted a retrospective cohort study of children < 3 years presenting to a tertiary paediatric ED in 2019 with trauma‐related chief complaints identified from structured electronic health record (EHR) data. Red flags were categorised as historical, physical, diagnostic, or caretaker/household. Two trained abstractors independently coded red flags from explicit provider documentation; discrepancies were resolved by consensus (inter‐rater reliability on a 20% sample; kappa = 0.82). Outcomes were the performance of skeletal surveys, laboratory testing, and social work/child protection consultations. Logistic regression estimated the association between red flags and any subsequent evaluation, adjusting for age. Results Among 962 eligible trauma‐related visits, 317 children (33%) had ≥ 1 documented red flag. Across the cohort, 8.6% received a skeletal survey, 4.3% had laboratory testing, and 1.5% had a social work consultation. Children with any red flag were more likely to undergo further evaluation (adjusted OR [aOR] 3.7; 95% CI: 2.5–5.4). Historical red flags were associated with skeletal survey (OR 3.7, 95% CI: 1.3–11) and laboratory tests (OR 4.1, 95% CI: 1.3–13.6). Physical red flags were associated with skeletal survey (OR 6.4, 95% CI: 2.1–19.6) and laboratories (OR 5.3, 95% CI: 1.5–19). Diagnostic red flags were associated with skeletal survey (OR 7.0, 95% CI: 2.3–21.6) and neuroimaging (OR 2.4, 95% CI: 1.3–4.5). Conclusion(s) One in three children under age 3 with trauma‐related ED visits had documented red flags suggestive of possible abuse, yet comprehensive evaluations were infrequent. Findings support the need for nuanced, standardised approaches that prioritise high‐yield flags while avoiding extensive evaluations.