Long bone fracture detection in suspected child abuse: contribution of lateral views

ACR guidelines for routine skeletal survey for child abuse recommend only AP radiographs of the long bones; however, many institutions add lateral radiographs. To evaluate whether adding lateral radiographs for long bones changes the frequency and confidence of fracture detection in skeletal survey radiographs for suspected abuse. We identified 100 children younger than 2 years of age who underwent skeletal survey for child abuse; 56 with multiple long bone fractures, 22 with a single fracture and 22 with no fractures. Four radiologists (two pediatric radiologists, one pediatric radiology fellow and one general radiologist) evaluated two randomized series (one series included only frontal, and the other series, frontal and lateral radiographs). Likert scale of 1–5 was used to score for detection of metaphyseal and diaphyseal fractures. For combined readers, significantly more metaphyseal fractures (P = 0.01) were detected with the two-views series of radiographs compared with the frontal-only view; there was no significant difference for diaphyseal fractures. Confidence was also significantly higher for the two-views series. Kappa improved (from 0.32 to 0.48) when the lateral view was added only for the metaphyseal fractures. Adding lateral radiographs resulted in increased detection and confidence levels of metaphyseal fractures.