Comparison of emergency physicians' and radiologists' interpretation of brain tomography in the emergency department

The number of patients presenting with complaints such as trauma and headache, altered consciousness, acute neurological deficits, etc., plays an important role in the emergency department population. In these patient groups, brain tomography is the most widely used diagnostic method for identifying intracranial pathologies. The ability of an emergency medical specialist to recognize and correctly interpret major pathologies such as intracranial hemorrhage in patients with acute complaints on the CT can play an important role in preventing delays in patient management. Our aim in this study is to compare the interpretation of the radiologist with the interpretation of the emergency physician of the brain tomography images in the Akdeniz University Emergency Service. Between 01.01.2016 and 31.12.2016, the images of 6640 patients who applied to Akdeniz University Hospital Emergency Service and who ordered head CT were analyzed. Images of a total of 120 patients with and without acute pathology were selected for CT imaging. Twenty emergency medical assistants were included in the study to interpret the images. Each of the selected images was evaluated by the emergency medical assistants included in the study and the findings were recorded in the form of findings, giving the clinical information and examination findings of the patients. Emergency medical assistants were grouped according to their seniority, and each group's comments were compared with the radiologist's interpretation and their compliance analyzed. Intracranial hemorrhage was detected in 51.7% according to the radiology report of the patients included in the study. Epidural hemorrhage in 12.5%, subdural hemorrhage in 20%, subarachnoid hemorrhage in 20%, intraparenchymal hemorrhage in 14.2% and intraventricular hemorrhage in 8.3% of the patients were detected. Acute infarction in 12.5% of the patients, intracranial mass in 12.5%, cerebral edema in 18.3%, fracture in 24.2%, and shift in 15% were found in the patients included to the study for the pathologies other than intracranial hemorrhage. In general, it has been observed that the compliance with radiology (mean kappa value 0.755) is good in emergency medical assistants in hemorrhage detection. Moderate compliance (mean kappa value 0,566 and 0,440) in subdural and intraventricular hemorrhage was found, and good compliance was observed in other hemorrhage types. It was found to be very good fit (mean kappa value 0.830 and 0.925) in calvarial fracture and midline structures shift detection. Acute infarction, cerebral edema, and mass detection of emergency medical assistants were found to be good compliance. With the increase in practical and theoretical training in BT interpretation in emergency medical education, compliance rates can be further increased. A physician who has just started an emergency medicine assistant demonstrates that, at the end of the training period, he / she will become an expert with sensitivity and selectivity close to radiologist commentary on head CT interpretation.