Jörgen Borg ^A1, Lena Holm ^A2, J. David Cassidy ^A2, Paul M. Peloso ^A5, Linda J. Carroll ^A3, Hans von Holst ^A6, Kaj Ericson ^A7
A1 Department of Neuroscience, Rehabilitation Medicine Uppsala University Hospital Uppsala Sweden
^A2 Section for Personal Injury Prevention, Department of Clinical Neurosciences Karolinska Institutet Stockholm Sweden
^A3 Alberta Centre for Injury Control and Research, Department of Public Health Sciences University of Alberta Edmonton Alberta Canada
^A4 Department of Medicine University of Alberta Edmonton Alberta Canada
^A5 Department of Internal Medicine University of Iowa Health Center Iowa City, Iowa USA
^A6 Department of Neurosurgery Karolinska Institutet Stockholm Sweden
^A7 Department of Neuroradiology Karolinska Institutet Stockholm Sweden

DOI: 10.1080/16501960410023822

We examined diagnostic procedures in mild traumatic brain injury by a systematic literature search. After screening 38,806 abstracts, we critically reviewed 228 diagnostic studies and accepted 73 (32%). The estimated prevalence of intracranial CT scan abnormalities is 5% in patients presenting to hospital with a Glasgow Coma Scale score of 15 and 30% or higher in patients presenting with a score of 13. About 1% of all treated patients with mild traumatic brain injury require neurosurgical intervention. There is strong evidence that clinical factors can predict computerized tomography scan abnormalities and the need for intervention in adults, but no such evidence for mild traumatic brain injury in children. We found evidence that skull fracture is a risk factor for intracranial lesions, but the diagnostic accuracy of radiologically diagnosed skull fracture as an indication of intracranial lesions is poor. There is only a little evidence for the diagnostic validity of cognitive testing and other diagnostic tools for mild traumatic brain injury.