A Statistical and Large Language Model to Define Oncology Nursing Advocacy Self‐Perceptions: An Explorative Study

Aim: To explore how Italian nurses working in oncology perceive patient advocacy within their clinical environments. Design: An observational, cross-sectional study examining self-perceptions of patient advocacy in oncology nursing practice. Methods: Demographic data were summarized as frequencies and percentages. Open-ended responses were analyzed using generative artificial intelligence (GAI). A freely accessible large language model (LLM) was employed to identify, cluster, and summarize the most relevant concepts and recurring terms provided by oncology nurses. The LLM enabled the extraction of key themes concerning experiences, strategies, and emotional outcomes related to cancer advocacy. Results: A total of 183 Italian oncology nurses participated in the study. Most participants were female (78.1%); 60.1% held a bachelor's degree, 55.2% had completed postbasic training, and 25.1% possessed a specific oncology certificate. The analysis revealed core emotional and professional themes, highlighting nurses' sense of moral responsibility, communication challenges, and ethical dilemmas when advocating for patients. Conclusions: The findings underscore the emotional and professional complexity inherent in oncology nursing advocacy. While many nurses reported empowerment and professional fulfillment through advocacy, others experienced frustration and isolation when encountering conflict or resistance. Implications for Nursing Practice: This study emphasizes the need for structured education, emotional support, and communication training to better prepare oncology nurses for the demands of patient advocacy.