Study objective: Previously published research (phase I) demonstrated a concerning misinterpretation rate of cranial computed tomography (CT) scans by emergency physicians. This study (phase II) determined whether an abbreviated educational session would improve emergency physician interpretation skills of cranial CT scans.
Methods: Participants in this prospective, interventional study in a county hospital ED were patients undergoing cranial CT scanning during ED evaluations and attending level emergency physicians. An abbreviated educational session on cranial CT interpretation skills was given to the same attending emergency physicians who participated in phase I. The educational session included basic CT interpretation skills and misinterpreted CT scans from phase I. We determined the postsession accuracy rate of the emergency physicians on 324 ED patient CT scans. The CT interpretation accuracy rates were then compared between phase I and phase II to determine the effectiveness of the educational session.
Results: The radiology/ED CT scan concordance rate improved from 61.3% (kappa = .22) to 88.6% (kappa = .70; P < .0001). Potentially clinically significant CT scan misinterpretations decreased from 24.1% to 4.0% (P < .0001). Most importantly, major missed findings on CT scans decreased from 11.4% to 2.8% (P < .0001). Continuous quality improvement monitoring found no instance of clinically significant patient mismanagement.
Conclusion: Within the limits of this study, we conclude that emergency physicians' interpretation skills of cranial CT scans may be improved using a 1-hour educational session.