The Perceptions of US Medical Students on Artificial Intelligence in Medicine: Mixed-Methods National Survey (Preprint)

Abstract

BACKGROUND Given the rapidity with which artificial intelligence (AI) is gaining momentum in clinical medicine, current physician leaders have called for more incorporation of AI topics into undergraduate medical education. This is to prepare future physicians to better work together with AI technology. However, the first step to curriculum development is to survey the needs of the end-users. There has not been a study to determine which mediums and which topics are most preferred by US medical students to learn about the topic of AI in medicine. OBJECTIVE We aim to survey US medical students on the need and means to incorporate AI in undergraduate medical education to assist with future education initiatives. METHODS A mixed-methods survey was sent through Qualtrics to US medical students in May 2021. Likert scale questions first assessed various perceptions regarding AI in medicine. We also asked how many hours they would like to spend per week to learn about AI. Then, we asked respondents to choose which learning format and which AI topics they would be most interested in. Finally, we used a free-response section to capture any remaining thoughts. RESULTS A total of 390 US medical students (average age: 26±3) from 17 different medical programs were surveyed (estimated response rate: 3.5%). A majority (92%) of respondents agreed that training in AI concepts during medical school is useful for their future career, but 91% reported of receiving no formal education related to AI. While 79% are excited to use AI technologies, 91% reported that their medical schools did not offer resources. Short lectures (68%), formal electives (48%), and Q&A panels (44%) were identified as preferred formats, while fundamental concepts of AI (65%), when to use AI in medicine (60%) and pros and cons of using AI (59%) were the most preferred topics for enhancing their training. Responses for the preferred formats and topics significantly differed between respondents who answered they wanted to spend ≤2 hours vs. ≥3 hours per month to learn about AI. CONCLUSIONS The results of this study indicate that current US medical students recognize the importance of AI in medicine, current formal education, and resources to learn AI-related topics are limited in most U.S. medical schools. Respondents indicated that a hybrid formal/ flexible format would be most appropriate for incorporating AI as a topic in US medical schools. Furthermore, multiple learning objectives for different groups of learners according to their future goals with AI (users or innovators) might be necessary.