• Enda O'Connor and James Walsham.
• The Prince Charles Hospital, Brisbane, Queensland, Australia. endamed@yahoo.com.au
• Emerg Med Australas. 2009 Apr 1;21(2):94-101.

AbstractThoracolumbar spine injury is a common complication of blunt multitrauma and up to one third of fractures are associated with spinal cord dysfunction. Delayed fracture diagnosis increases the risk of neurological complications. While validated screening guidelines exist for traumatic c-spine injury equivalent guidelines for thoracolumbar screening are lacking. We conducted a literature review evaluating studies of thoracolumbar injury in trauma patients to generate indications for thoracolumbar imaging. We performed MEDLINE and Pubmed searches using MeSH terms "Wounds, Nonpenetrating", "Spinal Fractures", "Spinal Injuries" and "Diagnostic Errors", MeSH/subheading terms "Thoracic Vertebrae/injuries" and "Lumbar Vertebrae/injuries" and keyword search terms "thoracolumbar fractures", "thoracolumbar injuries", "thoracolumbar trauma", "missed diagnoses" and "delayed diagnoses". Limits and inclusion criteria were defined prior to searching. We evaluated 16 articles; 5 prospective observational studies (1 cohort study) and 11 retrospective observational studies. Predictors of TL injury in prospective studies - high-risk injury mechanism, distracting injury, impaired cognition, symptoms/signs of vertebral fracture and known cervical fracture--were defined and used to construct a decision algorithm, which in a total of 14189 trauma patients from all eligible studies recommended TL screening in 856(99.1%) of 864 patients with TL fractures and would probably have directed TL imaging in the remaining 8 patients. There is limited low level evidence guiding surveillance TL imaging in adult blunt trauma patients. Despite this, we propose and evaluate an algorithm with a high negative predictive value for TL fractures. This should be incorporated into spinal injury assessment protocols.

14 keywords...

hide…