Ten guidelines for integrating generative AI into nursing research proposed

Researchers from the Fisabio Foundation and Universitat Jaume I have developed 10 recommendations for the responsible integration of generative artificial intelligence (GenAI) into nursing research. Published in the journal Enfermería Clínica, the study identifies how AI can streamline literature reviews and data analysis while warning against risks like hallucinations and the oversimplification of complex patient experiences. This framework is significant for the AI sector as it establishes a precedent for domain-specific governance and ethical standards in healthcare-related scientific inquiry.

The study, conducted by the GIENF-241 and eNURSYS research groups within the NURSIA Joint Research Unit, explores the transformative potential of GenAI across various research stages. While these tools can significantly improve efficiency in formulating research questions and disseminating results, the authors emphasize that they also introduce substantial methodological and ethical challenges. Specifically, the researchers warn of hallucinations, where AI systems generate seemingly rigorous but factually incorrect texts or nonexistent references, necessitating a mandatory verification process for all AI-generated content before it is incorporated into scientific work.

A primary concern highlighted in the report is the potential for AI to compromise the quality of nursing research, which often relies on qualitative methodologies and highly contextualized human experiences. The authors argue that GenAI may overlook fundamental dimensions of care, such as the therapeutic relationship and social determinants of health, by providing plausible but oversimplified interpretations. Furthermore, the study warns that AI models often reproduce biases from mainstream scientific literature, frequently prioritizing biomedical perspectives and English-language sources over the specific cultural contexts inherent to the nursing discipline.

To mitigate these risks, the researchers propose 10 practical guidelines aligned with international data protection regulations and the new European Artificial Intelligence Act. These recommendations include using AI only when it provides clear added value to human work, avoiding the input of sensitive data, and maintaining transparent documentation of how AI tools are utilized. By focusing on the preservation of critical thinking and methodological rigor, the study aims to guide researchers in adopting AI without sacrificing the ethical standards required for high-quality scientific output in the healthcare sector.