Minor head injury—defined as a Glasgow Coma Scale rating of 13 to 15 and a history of loss of consciousness, amnesia or disorientation following head trauma—is the cause of more than 1 million visits to emergency departments every year. Most patients recover well with no sequelae, but a few develop intracranial hematoma. The use of computed tomographic (CT) scanning in cases of minor head injury has increased dramatically, although more than 96 percent of these scans do not contribute to the management of the patient. Guidelines for use of CT scanning in this situation are inconsistent and confusing. Stiell and colleagues used data from 10 Canadian hospitals to develop decision rules for the use of CT following minor head injury.

The authors gathered data on adult patients meeting the above definition who presented within 24 hours of head trauma. Patients with skull fractures, unstable vital signs, focal neurologic signs, bleeding disorders, or pregnancy were excluded from the study. All patient assessments were conducted by qualified emergency department physicians who had specific training in the 22 standardized clinical items abstracted from the literature as potential predictors. A subset of patients was independently assessed by a second physician to judge consistency of assessment. All patients were examined clinically, and CT scanning was ordered at the discretion of the examining physician. The radiologists interpreting these scans were not aware of the status of the patient regarding the study. The primary outcome assessed was the need for neurologic intervention. A secondary outcome was the identification of clinically significant brain injury on CT scan.

During the three years of the study, data were gathered on more than 3,000 patients. One third were discharged directly from the emergency department and reviewed by structured interview 14 days after assessment. CT scanning was performed on 2,078 patients (67 percent). Neurologic intervention was required in 44 patients (1.0 percent), including four patients who died as a result of the head injury. Clinically significant brain injury was assessed in 254 patients (8.0 percent). An additional 94 (4.0 percent) had detectable lesions of low clinical significance (isolated contusions smaller than 5 mm in diameter or localized subarachnoid hemorrhage).

Using logistic regression, the authors identified predictive variables and developed a statistical model to identify patients at high risk of neurologic intervention. The variables in the model were converted into a list of seven questions that could be applied clinically. The Canadian CT Head Rule is shown in the accompanying table. Retrospectively, the authors calculate that applying the rule would have identified 320 of the 348 patients found to have injury on CT scanning, including all the clinically unimportant conditions. This rule has the potential to standardize and improve the management of patients with minor head injury.

editor's note: This recommendation was developed by the same group that provided the Ottawa rules for radiologic studies in cases of ankle, knee, and cervical spine injuries. The implications for improved patient care and cost savings are obvious, in addition to the security afforded physicians who struggle with the paradox of exposing the patient unnecessarily to significant radiation by ordering CT versus the small chance of missing a clinically significant lesion.—a.d.w.