By Mary Cheffers, MD
Minor head injury (generally defined as no alteration in mental status, well appearing, with GCS >13, on no anticoagulation) makes up 89% of all blunt head trauma. Yet not a few of them have a serious intracranial injury that requires medical action (observation or surgery). The Rational Clinical Exam article published in the December 2015 issue of JAMA entitled “Will Neuroimaging Reveal a Severe Intracranial Injury in This Adult With Minor Head Trauma?” shows that among this seemingly low risk population, there is an average prevalence rate for serious injury at 7.1%, with severe debilitating or fatal injuries at 0.9%. What distinguished these 7%?
The biggest historical risk factors were mechanism of Auto v Ped (LR of 4), Age > 60 (LR+ of 2.2), vomiting 2 or more times after injury (LR 3.6), GCS of 13 (LR+ of 5), an acute drop in GCS (LR+ ranges from 3-16) or post-traumatic seizure (LR 2.5). On physical exam, as would be expected, evidence of skull fracture (palpable fracture, raccoon eyes, Battle’s sign, hemotympanum or CNS rhino/otorrhea) conferred an average LR of 16, making it the strongest predictor of intracranial injury. Closely behind it was the presence of a new neurological deficit that can be mapped to an anatomical lesion with an LR of up to 7. Notably and importantly for patient education, loss of consciousness and headache are among the least helpful symptoms or signs with an LR around 1. Additionally, the presence of intoxication did not seem to increase their likelihood of having a severe intracranial injury (presumably because severely intoxicated patients are altered and therefore de facto fall out of the “minor” head injury category).
To put numbers on this, with a presumed prevalence of 7%, the presence of any one of these risk factors in a patient with minor blunt head injury raises the likelihood of an intracranial injury to approximately 20%, without any consideration for anticoagulation.
Several clinical decision rules have been proposed, the two most validated being the Canadian Head CT rule and the New Orleans criteria. With a prevalence of 7%, a negative Canadian CT Head Rule or New Orleans Criteria makes the likelihood of intracranial injury 0.3% or 0.6%, respectively, in patients with GCS 13+ and the presence of LOC, amnesia, or disorientation. N.B. neither of these rules have been validated in populations on anti-platelet/coagulant agents, or those intoxicated.
Two other general thoughts:
- We cannot forget that there are some clinically significant injuries that will be missed by a CT without contrast, namely artery dissection, dural venous thrombosis, or diffuse axonal injury. Patients who have serious GCS depression or neurological deficits with a negative CT head should still be considered for these injuries.
- Neither of these clinical decision tools has been directly compared to physician gestalt.
One general caveat in terms of clinically significant injuries that we should always keep in mind is that a CT without contrast will miss any arterial dissection, dural venous thrombosis, or diffuse axonal injury.
Notably missing in the development of these clinical decision tools is any studied direct comparison of their use with physician gestalt.
See the original JAMA article here:
Easter JS, Haukoos JS, Meehan WP, Novack V, Edlow JA. Will Neuroimaging Reveal a Severe Intracranial Injury in This Adult With Minor Head Trauma? The Rational Clinical Examination Systematic Review. JAMA.2015;314(24):2672-2681. doi:10.1001/jama.2015.16316.