Variation in ED use of computed tomography for patients with minor head injury

Study objective: To determine the frequency of utilization, yield for brain injury, incidence of missed injury, and variation in the use of computed tomography (CT) for ED patients with minor head injury. Methods: This retrospective health records survey was conducted over a 12-month period in the EDs at seven Canadian teaching institutions. Included in this review were adult patients who sustained acute minor head injury, defined as witnessed loss of consciousness or amnesia and a Glasgow Coma Scale score of 13 or greater. Data were collected by research assistants who were trained to select cases and abstract data in a standardized fashion according to a resource manual. Subsequently, patient eligibility was reviewed by the study coordinator and principal investigator. Results: Of the 1,699 patients seen, 521 (30.7%) were referred for CT, and 418 (79.8%) of these scans were negative for any type of brain injury. Overall, 105 (6.2%) of these patients sustained acute brain injury, including 9 (.5%) with an epidural hematoma. Cochran's Q test for homogeneity demonstrated significant variation between the seven centers for rate of ordering CT (P<.0001), from a low of 15.9% to a high of 70.4%. All five cases of ''missed'' hematoma occurred at the institutions with the highest acid third highest rates of CT use. After controlling for possible differences in case severity and patient characteristics at each hospital, logistic regression analysis revealed that five of seven hospitals were significantly associated with use of CT (respective odds ratios [OR], .4, .5, .5, 3.2, and 4.7). Three of the centers (two with the highest ordering rates) showed significant heterogeneity in the ordering of CT among their attending staff physicians, from a low of 6.5% to a high of 80.0%. Conclusion: There was considerable variation among institutions and individual physicians in the ordering of CT for patients with minor head injury. Although emergency physicians were selective when ordering CT, the yield of radiography was very low at all hospitals. None of the cases of ''missed'' intracranial hematoma came from the lowest ordering institutions, indicating that patients may be managed safely with a selective approach to CT use, These findings suggest great potential for more standardized and efficient use of CT of the head, possibly through the use of a clinical decision rule.