Development of a Competency-Based Radiology Clerkship Using Categorical and Statistical Analysis of Assessment

The purpose was to create and analyze a competency-based model of educating medical students in a radiology clerkship that can be used to guide curricular reform.

During the 2019 to 2020 academic year, 326 fourth-year medical students were enrolled in a 2-week required clerkship. An online testing platform, ExamSoft (Dallas, Texas), was used to test pre- and postinstruction knowledge on "must see" diagnoses, as outlined in the National Medical Student Curriculum in Radiology. Assessment analysis was used to compare the frequency with which the correct diagnosis was identified on the pretest to that on the posttest. At the end of the academic year, in addition to statistical analysis, categorical analysis was used to classify the degree of this change to uncover topics that students found most challenging.

For 23 of the 27 topics (85%), there was a significant improvement in diagnostic accuracy after instruction in the test curriculum. Categorical analysis further demonstrated that the clerkship had a high impact in teaching 13 of the 27 topics (48%), had a lower impact for 6 topics (22%), and identified the remaining 8 topics (30%) as gaps in teaching and learning.

For medical students, our instructional program significantly increased competency for most critical radiologic diagnoses. Categorical analysis adds value beyond statistical analysis and allows dynamic tailoring of teaching to address gaps in student learning.