ARRS 2022 Abstracts


1350. The Impact of Virtual Radiology Education in Medical School
Authors * Denotes Presenting Author
  1. Alexandra Jordan *; New York Medical College
  2. Faariah Shakil; New York Medical College
  3. Susan Rachlin; New York Medical College
  4. Daniel Katz; The University of Chicago Booth School of Business
Due to the COVID-19 pandemic and the need for social distancing, medical schools were required to adopt a hybrid or fully virtual curriculum for the 2020-2021 school year (AAMC 2020). At our home institution, the Radiology and Diagnostic Medicine (RDM) clerkship is designed as a fourth-year medical student elective in which students learn the basics of radiology and its application in all medical specialties. In Spring 2021, this course adopted a hybrid model of learning composed of both virtual and in-person components. However, the content remained consistent with those clerkships that were fully in-person prior to the pandemic (2020). In this article, we aimed to determine the efficacy of virtual learning in this radiology clerkship at the undergraduate medical education level by analyzing the final examination scores.

Materials and Methods:
RDM is offered in the spring semester to students entering medical specialties other than radiology. The educational content for this clerkship is offered solely through in-class participation, as content is not available from outside resources. The class concluded with a cumulative final examination that consisted of 40 multiple choice questions. To compare in-person and hybrid learning, we selected the classes of 2019 and 2021, excluding 2020, as the clerkship was not offered in its entirety that year. We compared final examination scores from the 2019 Spring semester, fully in person, to the 2021 Spring semester, offered in the hybrid format. We tested for differences in the magnitude of the scores as well as differences in the distributions.

The mean score on the final dropped from 88.88% in the 2019 class to 82.16% in the 2021 class, which was statistically significant (t(267.36) = 6.33, p < .001, Cohen’s D = 0.75). The distribution of grades in 2021 was not only lower than in 2019, it also had greater dispersion. An F-test for difference in variances showed the variance of scores was significantly higher in 2021 (20192 = 52.65, 20212 = 73.97, F(151,136) = 0.71, p = 0.042). This suggests that low-scoring students were especially disadvantaged by the hybrid format, while high-scoring students, although still negatively impacted, show slightly less variation.

Our results demonstrate the hybrid learning format was not equivalent to in-person learning for this particular clerkship. A likely explanation for these results is the unique educational format, where the acquired knowledge necessary is fully embedded within the clerkship. There are limited outside resources available and therefore, in-class participation and engagement are essential to the adequate understanding of the material. Additionally, students who need the most help may be negatively affected more by the hybrid learning format. We believe these results have important pedagogical implications for those seeking to understand the effect that hybrid learning has on medical student education.