ACCURACY OF INTERPRETATION OF CRANIAL COMPUTED-TOMOGRAPHY SCANS IN ANEMERGENCY-MEDICINE RESIDENCY PROGRAM

Study objectives: To determine the concordance of emergency physicians and radiologists in interpreting cranial computed tomography (CT) sca ns. The study also sought to determine the clinical significance of mi sinterpretations of cranial CT scans by emergency physicians. Design: Prospective cohort study. Setting: A county hospital emergency medicin e residency program. Participants: Five hundred fifty-five patients un dergoing CT scanning during emergency department evaluation. Results: Forty-nine percent (272) of the indications for CT scanning were for t rauma, 14.2% (79) were for cerebrovascular accident, 25.1% (139) were for headache, 15.1% (84) were for seizure, and 13.7% (76) were for mis cellaneous reasons. The radiologists interpreted 46.1% (256) of the CT scans as abnormal. The most frequent abnormalities were scalp hematom a, 15.2% (39); infarction, 14.1% (36); calcification, 6.3% (16); contu sion, 6.3% (16); parenchymal hemorrhage, 5.1%(13); and mass, 5.1% (13) . Nonconcordance between radiologists and emergency physicians was fou nd in 38.7% (206) of the cases. Potentially clinically significant mis interpretations were found in 24.1% (131) of the total sample. These m isinterpretations included 62 missed major findings (11.4% of total sa mple): 25 new infarcts, 10 mass lesions, 8 cases of cerebral edema, 8 parenchymal hemorrhages, 5 contusions, 4 subarachnoid hemorrhages, 1 e pidural hematoma, and 1 subdural hematoma. However, on chart review, o nly three patients (0.6%) were found to have been managed inappropriat ely, and none had an adverse outcome. Conclusion: The misinterpretatio n rate of cranial CT scans by emergency physicians is of potential cli nical concern. However, clinical mismanagement is rare. We recommend t hat more formal education in CT interpretation be included in residenc y training and continuing medical education programs for emergency phy sicians.