TY - JOUR
T1 - Evaluating Large Language Models in extracting cognitive exam dates and scores
AU - Zhang, Hao
AU - Jethani, Neil
AU - Jones, Simon
AU - Genes, Nicholas
AU - Major, Vincent J.
AU - Jaffe, Ian S.
AU - Cardillo, Anthony B.
AU - Heilenbach, Noah
AU - Ali, Nadia Fazal
AU - Bonanni, Luke J.
AU - Clayburn, Andrew J.
AU - Khera, Zain
AU - Sadler, Erica C.
AU - Prasad, Jaideep
AU - Schlacter, Jamie
AU - Liu, Kevin
AU - Silva, Benjamin
AU - Montgomery, Sophie
AU - Kim, Eric J.
AU - Lester, Jacob
AU - Hill, Theodore M.
AU - Avoricani, Alba
AU - Chervonski, Ethan
AU - Davydov, James
AU - Small, William
AU - Chakravartty, Eesha
AU - Grover, Himanshu
AU - Dodson, John A.
AU - Brody, Abraham A.
AU - Aphinyanaphongs, Yindalon
AU - Masurkar, Arjun
AU - Razavian, Narges
N1 - Publisher Copyright:
Copyright: © 2024 Zhang et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
PY - 2024/12
Y1 - 2024/12
N2 - Ensuring reliability of Large Language Models (LLMs) in clinical tasks is crucial. Our study assesses two state-of-the-art LLMs (ChatGPT and LlaMA-2) for extracting clinical information, focusing on cognitive tests like MMSE and CDR. Our data consisted of 135,307 clinical notes (Jan 12th, 2010 to May 24th, 2023) mentioning MMSE, CDR, or MoCA. After applying inclusion criteria 34,465 notes remained, of which 765 underwent ChatGPT (GPT-4) and LlaMA-2, and 22 experts reviewed the responses. ChatGPT successfully extracted MMSE and CDR instances with dates from 742 notes. We used 20 notes for fine-tuning and training the reviewers. The remaining 722 were assigned to reviewers, with 309 each assigned to two reviewers simultaneously. Inter-rater-agreement (Fleiss’ Kappa), precision, recall, true/ false negative rates, and accuracy were calculated. Our study follows TRIPOD reporting guidelines for model validation. For MMSE information extraction, ChatGPT (vs. LlaMA-2) achieved accuracy of 83% (vs. 66.4%), sensitivity of 89.7% (vs. 69.9%), true-negative rates of 96% (vs 60.0%), and precision of 82.7% (vs 62.2%). For CDR the results were lower overall, with accuracy of 87.1% (vs. 74.5%), sensitivity of 84.3% (vs. 39.7%), true-negative rates of 99.8% (98.4%), and precision of 48.3% (vs. 16.1%). We qualitatively evaluated the MMSE errors of ChatGPT and LlaMA-2 on double-reviewed notes. LlaMA-2 errors included 27 cases of total hallucination, 19 cases of reporting other scores instead of MMSE, 25 missed scores, and 23 cases of reporting only the wrong date. In comparison, ChatGPT’s errors included only 3 cases of total hallucination, 17 cases of wrong test reported instead of MMSE, and 19 cases of reporting a wrong date. In this diagnostic/prognostic study of ChatGPT and LlaMA-2 for extracting cognitive exam dates and scores from clinical notes, ChatGPT exhibited high accuracy, with better performance compared to LlaMA-2. The use of LLMs could benefit dementia research and clinical care, by identifying eligible patients for treatments initialization or clinical trial enrollments. Rigorous evaluation of LLMs is crucial to understanding their capabilities and limitations.
AB - Ensuring reliability of Large Language Models (LLMs) in clinical tasks is crucial. Our study assesses two state-of-the-art LLMs (ChatGPT and LlaMA-2) for extracting clinical information, focusing on cognitive tests like MMSE and CDR. Our data consisted of 135,307 clinical notes (Jan 12th, 2010 to May 24th, 2023) mentioning MMSE, CDR, or MoCA. After applying inclusion criteria 34,465 notes remained, of which 765 underwent ChatGPT (GPT-4) and LlaMA-2, and 22 experts reviewed the responses. ChatGPT successfully extracted MMSE and CDR instances with dates from 742 notes. We used 20 notes for fine-tuning and training the reviewers. The remaining 722 were assigned to reviewers, with 309 each assigned to two reviewers simultaneously. Inter-rater-agreement (Fleiss’ Kappa), precision, recall, true/ false negative rates, and accuracy were calculated. Our study follows TRIPOD reporting guidelines for model validation. For MMSE information extraction, ChatGPT (vs. LlaMA-2) achieved accuracy of 83% (vs. 66.4%), sensitivity of 89.7% (vs. 69.9%), true-negative rates of 96% (vs 60.0%), and precision of 82.7% (vs 62.2%). For CDR the results were lower overall, with accuracy of 87.1% (vs. 74.5%), sensitivity of 84.3% (vs. 39.7%), true-negative rates of 99.8% (98.4%), and precision of 48.3% (vs. 16.1%). We qualitatively evaluated the MMSE errors of ChatGPT and LlaMA-2 on double-reviewed notes. LlaMA-2 errors included 27 cases of total hallucination, 19 cases of reporting other scores instead of MMSE, 25 missed scores, and 23 cases of reporting only the wrong date. In comparison, ChatGPT’s errors included only 3 cases of total hallucination, 17 cases of wrong test reported instead of MMSE, and 19 cases of reporting a wrong date. In this diagnostic/prognostic study of ChatGPT and LlaMA-2 for extracting cognitive exam dates and scores from clinical notes, ChatGPT exhibited high accuracy, with better performance compared to LlaMA-2. The use of LLMs could benefit dementia research and clinical care, by identifying eligible patients for treatments initialization or clinical trial enrollments. Rigorous evaluation of LLMs is crucial to understanding their capabilities and limitations.
UR - http://www.scopus.com/inward/record.url?scp=85212572536&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85212572536&partnerID=8YFLogxK
U2 - 10.1371/journal.pdig.0000685
DO - 10.1371/journal.pdig.0000685
M3 - Article
AN - SCOPUS:85212572536
SN - 2767-3170
VL - 3
JO - PLOS Digital Health
JF - PLOS Digital Health
IS - 12
M1 - e0000685
ER -