Assessment of Large Language Models (LLMs) in decision-making support for gynecologic oncology

• Published in: Computational and structural biotechnology journal Vol. 23; pp. 4019 - 4026
• Main Authors: Gumilar, Khanisyah Erza, Indraprasta, Birama R., Faridzi, Ach Salman, Wibowo, Bagus M., Herlambang, Aditya, Rahestyningtyas, Eccita, Irawan, Budi, Tambunan, Zulkarnain, Bustomi, Ahmad Fadhli, Brahmantara, Bagus Ngurah, Yu, Zih-Ying, Hsu, Yu-Cheng, Pramuditya, Herlangga, Putra, Very Great E., Nugroho, Hari, Mulawardhana, Pungky, Tjokroprawiro, Brahmana A., Hedianto, Tri, Ibrahim, Ibrahim H., Huang, Jingshan, Li, Dongqi, Lu, Chien-Hsing, Yang, Jer-Yen, Liao, Li-Na, Tan, Ming
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.12.2024; Research Network of Computational and Structural Biotechnology; Elsevier
• Subjects: Accuracy; Artificial intelligence; biotechnology; Consistency; decision making; Gynecologic cancer; Large Language Models; patients; Accuracy; Large Language Models; Artificial intelligence; Gynecologic cancer; Consistency
• ISSN: 2001-0370; 2001-0370
• DOI: 10.1016/j.csbj.2024.10.050

This study investigated the ability of Large Language Models (LLMs) to provide accurate and consistent answers by focusing on their performance in complex gynecologic cancer cases. LLMs are advancing rapidly and require a thorough evaluation to ensure that they can be safely and effectively used in clinical decision-making. Such evaluations are essential for confirming LLM reliability and accuracy in supporting medical professionals in casework. We assessed three prominent LLMs—ChatGPT-4 (CG-4), Gemini Advanced (GemAdv), and Copilot—evaluating their accuracy, consistency, and overall performance. Fifteen clinical vignettes of varying difficulty and five open-ended questions based on real patient cases were used. The responses were coded, randomized, and evaluated blindly by six expert gynecologic oncologists using a 5-point Likert scale for relevance, clarity, depth, focus, and coherence. GemAdv demonstrated superior accuracy (81.87 %) compared to both CG-4 (61.60 %) and Copilot (70.67 %) across all difficulty levels. GemAdv consistently provided correct answers more frequently (>60 % every day during the testing period). Although CG-4 showed a slight advantage in adhering to the National Comprehensive Cancer Network (NCCN) treatment guidelines, GemAdv excelled in the depth and focus of the answers provided, which are crucial aspects of clinical decision-making. LLMs, especially GemAdv, show potential in supporting clinical practice by providing accurate, consistent, and relevant information for gynecologic cancer. However, further refinement is needed for more complex scenarios. This study highlights the promise of LLMs in gynecologic oncology, emphasizing the need for ongoing development and rigorous evaluation to maximize their clinical utility and reliability. [Display omitted] •Large Language Models (LLMs) are valuable tools in clinical practice, aiding healthcare professionals in making evidence-based decisions and improving patient care.•Gemini Advanced achieved 81.87 % accuracy in clinical decision-making.•Gemini Advanced consistently provided correct answers > 60 % every day during the testing period.•ChatGPT-4 and Gemini Advanced outperformed Copilot in treatment recommendations.•Further improvements are necessary to ensure accurate and relevant responses across clinical scenarios.