Garcia Sanchez Carolina, Kharko Anna, Hägglund Maria, Riggare Sara, Blease Charlotte
Participatory eHealth and Health Data Research Group, Department of Women's and Children's Health, Uppsala University, Uppsala, Sweden.
Centre for Primary Care and Health Services Research, University of Manchester, Manchester, United Kingdom.
JMIR Res Protoc. 2025 Aug 8;14:e73602. doi: 10.2196/73602.
Generative artificial intelligence (GenAI) leverages large language models (LLMs) that are transforming health care. Specialized ambient GenAI tools, like Nuance Dax, Speke, and Tandem Health, "listen" to consultations and generate clinical notes. Medical-focused models, like Med-PaLM, provide tailored health care insights. GenAI's capability to summarize complex data and generate responses in various conversational styles or literacy levels makes it particularly valuable since it has the potential to alleviate the burden of clinical documentation on health care professionals (HCPs). While GenAI may prove to be helpful, offering novel benefits, it comes with its own set of challenges. The quality of the source data can introduce biases, leading to skewed recommendations or outright false information (so-called hallucinations). In addition, due to the conversational nature of chatbot responses, users may be susceptible to misinformation, posing risks to both safety and privacy. Therefore, careful implementation and rigorous oversight are essential to ensure accuracy, ethical integrity, and alignment with clinical standards. Despite these advances, currently, no review has investigated HCPs' experiences and opinions about GenAI in clinical documentation. Yet, such a perspective is crucial to better understand how these technologies can be safely and ethically adopted and implemented in clinical practice.
We aim to present the protocol for a scoping review exploring HCPs' experiences and opinions about GenAI and ambient scribes in clinical documentation.
This scoping review will be carried out following the methodological framework of Arksey and O'Malley and the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Scoping Reviews) checklist. Relevant papers will be searched for in PubMed, IEEE Xplore, APA PsycInfo, CINAHL, and Web of Science. The review will include studies published between January 2023 and September 2025. Studies will be included that represent original peer-reviewed work that explores HCPs' experiences and opinions about the use of GenAI or ambient scribes for clinical documentation. Data extraction will include publication type, country, sample characteristics, clinical setting, study aim, study design, research question, and key findings. Study quality will be assessed using the Mixed Methods Appraisal Tool.
The results will be presented as a narrative synthesis structured along the key themes of the evidence mapped. Data will be collated and presented in charts and tabular format. Findings will be reported in a peer-reviewed scoping review.
This will be the first scoping review that considers HCPs' experiences and opinions about GenAI and ambient scribes in clinical documentation. The results will clarify how HCPs use-or avoid using-GenAI in daily health care work. This insight will help address perceived benefits, risks, expectations, and uncertainties. It may also reveal key research gaps in the field.
INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): PRR1-10.2196/73602.
生成式人工智能（GenAI）利用正在改变医疗保健行业的大语言模型（LLMs）。像Nuance Dax、Speke和Tandem Health这样的专业环境生成式人工智能工具“聆听”会诊并生成临床记录。像Med-PaLM这样以医学为重点的模型提供量身定制的医疗保健见解。GenAI能够总结复杂数据并以各种对话风格或读写水平生成回复，这使其特别有价值，因为它有可能减轻医疗保健专业人员（HCPs）的临床文档负担。虽然GenAI可能被证明是有帮助的，能带来新的益处，但它也带来了一系列自身的挑战。源数据的质量可能会引入偏差，导致有偏差的建议或完全错误的信息（即所谓的幻觉）。此外，由于聊天机器人回复的对话性质，用户可能容易受到错误信息的影响，对安全和隐私都构成风险。因此，谨慎实施和严格监督对于确保准确性、道德完整性以及与临床标准保持一致至关重要。尽管有这些进展，但目前尚无综述研究医疗保健专业人员对GenAI在临床文档方面的体验和看法。然而，这样的观点对于更好地理解这些技术如何在临床实践中安全、合乎道德地采用和实施至关重要。
我们旨在介绍一项范围综述的方案，该综述探讨医疗保健专业人员对GenAI和临床文档中的环境抄写员的体验和看法。
本范围综述将按照Arksey和O'Malley的方法框架以及PRISMA-ScR（范围综述的系统评价和Meta分析优先报告项目）清单进行。将在PubMed、IEEE Xplore、APA PsycInfo、CINAHL和Web of Science中搜索相关论文。该综述将纳入2023年1月至2025年9月期间发表的研究。纳入的研究将代表经过同行评审的原创作品，探讨医疗保健专业人员对使用GenAI或环境抄写员进行临床文档记录的体验和看法。数据提取将包括出版物类型、国家、样本特征、临床环境、研究目的、研究设计、研究问题和关键发现。将使用混合方法评估工具评估研究质量。
结果将以叙事性综述的形式呈现，按照所绘制证据的关键主题进行结构化。数据将进行整理并以图表和表格形式呈现。研究结果将在经过同行评审的范围综述中报告。
这将是第一项考虑医疗保健专业人员对GenAI和临床文档中的环境抄写员的体验和看法的范围综述。结果将阐明医疗保健专业人员在日常医疗工作中如何使用——或避免使用——GenAI。这一见解将有助于解决感知到的益处、风险、期望和不确定性。它还可能揭示该领域的关键研究空白。
国际注册报告识别号（IRRID）：PRR1-10.2196/73602
