Integrating a Large Language Model to Streamline Nursing Handover Documentation Across Multiple Hospitals in Taiwan:...

Area of Science:

• Healthcare Informatics
• Artificial Intelligence in Nursing

Background:

• The global nursing shortage is worsened by high workloads and turnover, with digital health tools increasing documentation burden.
• Generative artificial intelligence (GenAI) and large language models (LLMs) offer potential solutions for streamlining nursing documentation.

Purpose of the Study:

• To evaluate the integration of an LLM into a nursing information system (NIS) to decrease handover documentation time.
• To assess the operational and economic impacts of GenAI-assisted nursing workflows.

Main Methods:

• A multidisciplinary team integrated an LLM into an existing NIS across three Taiwanese hospitals.
• Prompt-based interfaces were developed for automated generation of nursing handover document content.
• NIS log data was analyzed to compare documentation times before and after LLM implementation.

Main Results:

• LLM integration led to significantly shorter per-patient handover documentation times.
• Average documentation time decreased from 3.45-4.32 minutes to 1.17-2.54 minutes per patient.
• Monthly time savings ranged from 474 to 981 hours across the three hospitals.

Conclusions:

• LLM integration into NIS streamlines nursing handover documentation without altering existing workflows.
• GenAI assistance resulted in shorter documentation times and positive net labor value.
• Future work includes refining prompts and assessing generalizability to other documentation tasks.