Introduction: The skeletal survey is widely used as the principal radiological investigation in suspected physical abuse of infants and young children. However, the evidence on which current guidelines are based is limited, especially for siblings of index cases. We conducted a retrospective study to describe the characteristics of children who underwent skeletal surveys for suspected non-accidental injury (NAI) in the Edinburgh area; to evaluate the diagnostic efficacy of skeletal surveys; and to identify any predictive factors that might guide clinical practice.
Methods: All skeletal surveys performed at the Royal Hospital for Sick Children in Edinburgh for suspected non-accidental injury between 1/1/99 and 31/12/03 were reviewed.
Results: Seventy-seven children underwent skeletal surveys for suspected NAI. Data were available for 76 (70 index cases and 6 siblings). Of the index cases, 17 (24%) skeletal surveys were positive, with a mean of 2.5 fractures per child (range 0-9). The age of positive cases ranged from 2 weeks to 36 months with the majority (14, 82%) being under 12 months of age. Nine (53%) were boys. In index children with positive results, indications for skeletal survey were head injury (skull fracture in 2, 12%; intracranial injury in 4,24%); skeletal fracture (excluding skull fracture) in 6 (35%) and bruising in 4 (24%). chi(2) analysis revealed no significant differences in age, sex or primary reason for skeletal survey between those with a negative and positive skeletal survey. Six skeletal surveys were performed on siblings of children with known or suspected NAI. The siblings' ages ranged from 1 to 36 months although 3 (50%) were under 12 months of age. One (17%) of the siblings' surveys (a twin) was positive.
Conclusion: The age of a child (<12 months) and type of presenting injury (overt fracture or head injury) are factors which can help the clinician to decide whether or not to obtain a skeletal survey. There remains a lack of evidence for the effectiveness of skeletal survey in the siblings of index children. Guidelines for the process of repeating uncertain skeletal surveys are needed.