An Evaluation Framework for Large Language Models in Clinical Nursing: A Scoping Review and Expert Consultation
TL;DR
The findings provide nursing managers with a structured reference for identifying core evaluation dimensions, interpreting evidence, and planning the evaluation and deployment of nursing‐specific LLM applications.
Abstract
Aim To examine the current state of large language model (LLM) evaluation in clinical nursing, synthesize the evaluation dimensions and metrics reported in the literature, and develop a preliminary evaluation framework for LLMs in clinical nursing through expert consultation. Background The advancement of artificial intelligence products, exemplified by LLMs, has generated excitement about their potential applications in clinical nursing practice, but their effectiveness remains uncertain. Methods A scoping review was conducted in accordance with Arksey and O’Malley’s framework and incorporated experts’ consultation. A literature search was conducted across Web of Science, PubMed, Embase, and the Cochrane Library, from their inception to June 21, 2025. Three expert meetings involving eight experts were conducted between August and October 2025 to synthesize evaluation frameworks and scenarios. Results A total of 42 studies were included, and the GPT family was the most frequently evaluated. Thirty‐seven evaluation metrics were extracted and refined through expert consultation into six primary domains: performance and accuracy, clinical validity and safety, workflow integration and efficiency, usability and user experience, model reliability and ethical considerations, and competency development. A scenario classification and a proposed minimum technical reporting checklist were also developed to support the transparent and comparable evaluation of LLMs in clinical nursing. Conclusion This study used a scoping review and expert consultation to summarize contemporary literature on LLM evaluations in clinical nursing practice. It provides a preliminary, structured basis for developing and refining a standardized evaluation framework. Implications for Nursing Management This study highlights the need for systematic and context‐sensitive evaluation of LLMs in clinical nursing. The findings provide nursing managers with a structured reference for identifying core evaluation dimensions, interpreting evidence, and planning the evaluation and deployment of nursing‐specific LLM applications.