Large language model-supported interactive case-based learning: a pilot study

Haelynn Gim; Benjamin Cook; Jasmin Le; Brandon Stretton; Christina Gao; Aashray Gupta; Joshua Kovoor; Christina Guo; Matthew Arnold; Galina Gheihman; Stephen Bacchi

doi:10.1111/imj.70030

Large language model-supported interactive case-based learning: a pilot study

Haelynn Gim
, Benjamin Cook
, Jasmin Le
, Brandon Stretton
, Christina Gao
, Aashray Gupta
, Joshua Kovoor
, Christina Guo
, Matthew Arnold
, Galina Gheihman
, Stephen Bacchi

College of Medicine and Public Health

Research output: Contribution to journal › Article › peer-review

4 Citations (Scopus)

Abstract

Large language models (LLMs) have been proposed as a means to augment case-based learning but are prone to generating factually incorrect content. In this study, an LLM-based tool was developed, and its performance evaluated. In response to student-generated questions, the LLM adhered to the provided screenplay in 832/857 (97.1%) instances, and in the remaining instances, it was medically appropriate in 24/25 (96.0%) cases. Use of LLM appears to be feasible for this purpose, and further studies are required to examine their educational impact.

Original language	English
Pages (from-to)	852-855
Number of pages	4
Journal	Internal Medicine Journal
Volume	55
Issue number	5
DOIs	https://doi.org/10.1111/imj.70030
Publication status	Published - May 2025

Keywords

artificial intelligence
clinical reasoning
machine learning
medical education
natural language processing

Access to Document

10.1111/imj.70030Licence: In Copyright

Cite this

@article{0cdf994b9bd34a9e8ddc9f3d9bd8e5c6,

title = "Large language model-supported interactive case-based learning: a pilot study",

abstract = "Large language models (LLMs) have been proposed as a means to augment case-based learning but are prone to generating factually incorrect content. In this study, an LLM-based tool was developed, and its performance evaluated. In response to student-generated questions, the LLM adhered to the provided screenplay in 832/857 (97.1\%) instances, and in the remaining instances, it was medically appropriate in 24/25 (96.0\%) cases. Use of LLM appears to be feasible for this purpose, and further studies are required to examine their educational impact.",

keywords = "artificial intelligence, clinical reasoning, machine learning, medical education, natural language processing",

author = "Haelynn Gim and Benjamin Cook and Jasmin Le and Brandon Stretton and Christina Gao and Aashray Gupta and Joshua Kovoor and Christina Guo and Matthew Arnold and Galina Gheihman and Stephen Bacchi",

year = "2025",

month = may,

doi = "10.1111/imj.70030",

language = "English",

volume = "55",

pages = "852--855",

journal = "Internal Medicine Journal",

issn = "1444-0903",

publisher = "John Wiley \& Sons, Inc",

number = "5",

}

TY - JOUR

T1 - Large language model-supported interactive case-based learning

T2 - a pilot study

AU - Gim, Haelynn

AU - Cook, Benjamin

AU - Le, Jasmin

AU - Stretton, Brandon

AU - Gao, Christina

AU - Gupta, Aashray

AU - Kovoor, Joshua

AU - Guo, Christina

AU - Arnold, Matthew

AU - Gheihman, Galina

AU - Bacchi, Stephen

PY - 2025/5

Y1 - 2025/5

N2 - Large language models (LLMs) have been proposed as a means to augment case-based learning but are prone to generating factually incorrect content. In this study, an LLM-based tool was developed, and its performance evaluated. In response to student-generated questions, the LLM adhered to the provided screenplay in 832/857 (97.1%) instances, and in the remaining instances, it was medically appropriate in 24/25 (96.0%) cases. Use of LLM appears to be feasible for this purpose, and further studies are required to examine their educational impact.

AB - Large language models (LLMs) have been proposed as a means to augment case-based learning but are prone to generating factually incorrect content. In this study, an LLM-based tool was developed, and its performance evaluated. In response to student-generated questions, the LLM adhered to the provided screenplay in 832/857 (97.1%) instances, and in the remaining instances, it was medically appropriate in 24/25 (96.0%) cases. Use of LLM appears to be feasible for this purpose, and further studies are required to examine their educational impact.

KW - artificial intelligence

KW - clinical reasoning

KW - machine learning

KW - medical education

KW - natural language processing

UR - https://www.scopus.com/pages/publications/105000740177

U2 - 10.1111/imj.70030

DO - 10.1111/imj.70030

M3 - Article

C2 - 40125598

AN - SCOPUS:105000740177

SN - 1444-0903

VL - 55

SP - 852

EP - 855

JO - Internal Medicine Journal

JF - Internal Medicine Journal

IS - 5

ER -

Large language model-supported interactive case-based learning: a pilot study

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this