Has This Child Experienced Physical Abuse?: The Rational Clinical Examination Systematic Review

Importance: Nearly 100 000 children experience physical abuse each year in the US. Among approximately 2000 annual deaths related to child maltreatment, more than 40% resulted from physical abuse, and half of those children were younger than 1 year. Many of these young children had unidentified abusive injuries before the fatal event.

Objective: To determine the accuracy of clinical and radiologic findings for identifying physical abuse among children who have sustained an injury.

Data sources and study selection: MEDLINE, PubMed Central, and Embase were searched for articles published from 1970 to September 2024. Three authors identified studies describing clinical and radiologic characteristics in children and adolescents undergoing assessment for physical abuse.

Data extraction and synthesis: The number of children with and without each clinical or radiologic finding, and the presence or absence of physical abuse, which had been determined by expert panels, predefined criteria, or standardized scales that quantify the level of concern for abuse, were recorded.

Main outcomes and measures: The sensitivity, specificity, and likelihood ratios (LRs) of each finding for the presence of physical abuse were calculated and the range or calculated summary measures were reported when the finding was evaluated in more than 1 study.

Results: Of 7378 unique articles, 18 studies met inclusion criteria. The prevalence of physical abuse in these 18 studies ranged from 5% to 79%. Studies that were focused on skin findings in children evaluated for trauma showed that the presence of oral injury such as a torn frenulum (positive LR, 6.6 [95% CI, 3.2-14.0]), bruising on the buttocks (positive LR range, 15-83) or neck (positive LR range, 2.2-84), patterned bruises (positive LR range, 2.0-66), and subconjunctival hemorrhage (positive LR range, 5.4-130) were associated with increased likelihood of physical abuse. In studies of hospitalized children with head injury, the presence of retinal hemorrhages (positive LR, 11.0 [95% CI, 4.0-32.0]), seizures (positive LR, 3.9 [95% CI, 2.4-6.5]), hypoxic ischemic injury (positive LR, 3.4 [95% CI, 1.8-6.4]), or a subdural hematoma (positive LR, 3.2 [95% CI, 2.6-3.8]) increased the likelihood of physical abuse. In studies examining children who underwent skeletal surveys, a single fracture (positive LR, 5.9 [95% CI, 2.9-12.0]) or multiple fractures (positive LR, 3.8 [95% CI, 2.4-6.0]) increased the likelihood of physical abuse.

Conclusions and relevance: A detailed physical examination that reveals oral injury, bruises on the buttocks or neck, patterned bruises, and subconjunctival hemorrhage in young children should alert clinicians to the possibility of physical abuse. Findings on neuroimaging and ophthalmologic evaluation in infants and young children with head trauma can help clinicians determine the likelihood of physical abuse.