Nurses' Perspectives on Evidence Dissemination Barriers and Large Language Model-Based Support: Qualitative Study...

Area of Science:

• Nursing Informatics
• Artificial Intelligence in Healthcare
• Evidence-Based Practice

Background:

• Current evidence dissemination methods are inadequate for clinical nurses, impeding evidence-based nursing practice.
• General-purpose large language models (LLMs) lack the domain-specific knowledge required for effective clinical evidence dissemination.
• Developing tailored AI tools is crucial to meet nurses' needs and enhance evidence dissemination.

Purpose of the Study:

• Identify challenges and barriers clinical nurses encounter in evidence dissemination.
• Examine nurses' perspectives on utilizing existing LLMs for evidence dissemination.
• Explore nurses' needs and preferences for an LLM-based nursing evidence question-answering system.

Main Methods:

• Qualitative study combining focus group discussions and the nominal group technique (NGT).
• 22 clinical nurses participated, recruited via purposive sampling across diverse specialties and experience levels.
• Online focus groups and NGT sessions were conducted to explore barriers, LLM perspectives, and desired system functionalities.

Main Results:

• Focus groups identified themes of evidence dissemination pathways, barriers, and LLM advantages/limitations.
• Nominal group technique identified 9 desired functions for an LLM-based nursing evidence system.
• Top-ranked desired functions include evidence-based, high-quality question-answering, evidence source provision, and personalized evidence recommendation.

Conclusions:

• Significant barriers exist in current nursing evidence dissemination processes.
• Large language models (LLMs) show potential for supporting evidence dissemination but require refinement.
• Clinical nurses' identified functional needs provide a roadmap for developing specialized LLMs for nursing practice.