Case Comparisons: An Efficient Way of Learning Radiology

Abstract

Radiologists commonly use comparison films to improve their differential diagnosis. Educational literature suggests that this technique might also be used to bolster the process of learning to interpret radiographs. We investigated the effectiveness of three comparison techniques in medical students, whom we invited to compare cases of the same disease (same-disease comparison), cases of different diseases (different-disease comparison), disease images with normal images (disease/normal comparison), and identical images (no comparison/control condition). Furthermore, we used eye-tracking technology to investigate which elements of the two cases were compared by the students. We randomly assigned 84 medical students to one of four conditions and had them study different diseases on chest radiographs, while their eye movements were being measured. Thereafter, participants took two tests that measured diagnostic performance and their ability to locate diseases, respectively. Students studied most efficiently in the same-disease and different-disease comparison conditions: test 1, F(3, 68)=3.31, P=.025, ηp(2)=0.128; test 2, F(3, 65)=2.88, P=.043, ηp(2)=0.117. We found that comparisons were effected in 91% of all trials (except for the control condition). Comparisons between normal anatomy were particularly common (45.8%) in all conditions. Comparing cases can be an efficient way of learning to interpret radiographs, especially when the comparison technique used is specifically tailored to the learning goal. Eye tracking provided insight into the comparison process, by showing that few comparisons were made between abnormalities, for example.