Is spinal computed tomography necessary in pediatric trauma patients?

Background To investigate the prevalence of findings of spinal injury on computed tomography (CT) images of pediatric trauma patients and to define indicators for the possible presence of spinal injuries. Methods Spinal CT for pediatric trauma patients (age <= 12 years) over a 2-year period was retrospectively evaluated for the presence of findings suggestive of spinal injury. Results Of the 773 patients reviewed, 19 (2.4%) showed traumatic spinal lesions on their spinal CT images. These patients were significantly older than those without spinal lesions (mean age 7.9 +/- 3.3 years vs 6.1 +/- 3.3 years; P = 0.02). The prevalence of spinal trauma was significantly lower in patients aged 0-8 years than in those aged 9-12 years (P = 0.025). Spinal injury was significantly higher in female patients (P = 0.014). Most of the spinal injuries were located at the lumbar and sacral vertebral levels, and most did not cause neurological complications or require surgical treatment. Important indicators of the possible presence of spinal injuries were pain, tenderness, or ecchymosis over the spine, a low Glasgow Coma Scale score (<= 12), head injury (for cervical injuries), or intrathoracic injuries and pelvic fractures (for lumbar and sacral injuries). Conclusions Most spinal CT examinations for the diagnosis of spinal injuries in children did not show positive findings. Thus, many children were exposed to an unnecessary high dose of radiation. New clinical evaluation criteria and indicators should be defined to diagnose spinal injuries and avoid unnecessary radiological examinations.