Sharma Aditi, Medapalli Tejas, Alexandrou Micaella, Brilakis Emmanouil, Prasad Anand
Department of Medicine, Division of Cardiology, University of Texas (UT) Health San Antonio, San Antonio, USA.
Department of Cardiology, Minneapolis Heart Institute, Minneapolis, USA.
Cureus. 2024 Apr 24;16(4):e58936. doi: 10.7759/cureus.58936. eCollection 2024 Apr.
Chat Generative Pre-Trained Transformer (ChatGPT) is a chatbot based on a large language model that has gained public interest since its release in November 2022. This systematic review examines the current literature on the potential applications of ChatGPT in cardiology. A systematic literature search was conducted to retrieve all publications on ChatGPT in PubMed, Scopus, MedRxiv, and the Cochrane Library published on or before September 30, 2023. Search terms relating to ChatGPT and cardiology were used. Publications without relevance to ChatGPT and cardiology were excluded. The included publications were divided into cohorts. Cohort A examined ChatGPT's role in improving patient health literacy. Cohort B explored ChatGPT's role in clinical care. Cohort C examined ChatGPT's role in future literature and research. Cohort D included case reports that used ChatGPT. A total of 115 publications were found across all databases. Twenty-four publications met the inclusion criteria and were included in the review. Cohort A-C included a total of 14 records comprised of editorials/letters to the editor (29%), research letters/correspondence (21%), review papers (21%), observational studies (7%), research studies (7%), and short reports (7%). Cohort D included 10 case reports. No relevant systematic literature reviews, meta-analyses, or randomized controlled trials were identified in the search. Based on this review of the literature, ChatGPT has the potential to enhance patient education, support clinicians providing clinical care, and enhance the development of future literature. However, further studies are needed to understand the potential applications of ChatGPT in cardiology and to address ethical concerns regarding the delivery of medical advice and the authoring of manuscripts.
聊天生成预训练变换器(ChatGPT)是一种基于大语言模型的聊天机器人,自2022年11月发布以来受到了公众的关注。本系统评价研究了关于ChatGPT在心脏病学中潜在应用的当前文献。进行了系统的文献检索,以检索在PubMed、Scopus、MedRxiv和Cochrane图书馆上发表的截至2023年9月30日或之前的所有关于ChatGPT的出版物。使用了与ChatGPT和心脏病学相关的检索词。排除与ChatGPT和心脏病学无关的出版物。纳入的出版物被分为不同队列。队列A研究了ChatGPT在提高患者健康素养方面的作用。队列B探讨了ChatGPT在临床护理中的作用。队列C研究了ChatGPT在未来文献和研究中的作用。队列D包括使用ChatGPT的病例报告。在所有数据库中总共发现了115篇出版物。24篇出版物符合纳入标准并被纳入本评价。队列A - C总共包括14条记录,由社论/致编辑的信(29%)、研究信/通信(21%)、综述论文(21%)、观察性研究(7%)、研究性研究(7%)和简短报告(7%)组成。队列D包括10篇病例报告。在检索中未发现相关的系统文献综述、荟萃分析或随机对照试验。基于对文献的这一综述,ChatGPT有潜力加强患者教育、支持临床医生提供临床护理以及促进未来文献的发展。然而,需要进一步研究以了解ChatGPT在心脏病学中的潜在应用,并解决有关提供医疗建议和撰写稿件的伦理问题。