訪問診療実習を通して得た，医学生の学びの解析―KH Coderによるテキストマイニングから―

抄録

In current medical education, there are a few university medical schools that systematically provide education on "home medical care" and "end-of-life care"; however, their methods have not yet been established. It is an important issue for medical education instructors to examine whether students can deepen their understanding and learning about home medical care characteristics and its target patients and their families through home medical care practice.

To this end, this study analyzed medical students' free-text reports using text mining. Text mining is a method of cutting text data word by word, analyzing it in a quantitative way, and visualizing the results. In this study, 69 fifth-year medical students and 11 sixth-year medical students participated from 2015 to 2019, and submitted free-text reports for analysis. The total number of extracted words was 76,976, and the top five most frequently extracted words were "patient" (1,044 times), "think" (436 times), "medical care" (431 times), "home medical care" (362 times), and "hospital" (335 times). In the co-occurrence network, fifth-year medical students were strongly associated with words such as "surprise" and "know," and sixth-year medical students with words such as "cooperation" and "background." In the correspondence analysis, fifth-year medical students in 2017 were strongly associated with such terms as "death" and "end-of-life care." By using text mining, we were able to easily obtain a holistic overview of the reports of 80 students and capture their characteristics. Although this method requires attention to data processing and interpretation, it is expected to be a useful tool for analyzing medical students' reports.