Applying artificial intelligence technology to support decision-making in nursing: A case study in Taiwan

Area of Science:

• Nursing Informatics
• Artificial Intelligence in Healthcare
• Clinical Decision Support Systems

Background:

• Nurses' diagnostic accuracy can be limited by reliance on professional knowledge, clinical experience, and instinct.
• Incomplete knowledge or delayed thinking can compromise diagnostic accuracy and decision-making.
• Nursing information systems offer a technological approach to support clinical judgment.

Purpose of the Study:

• To investigate the application of artificial intelligence (AI) in generating nursing diagnoses.
• To evaluate the accuracy and impact of an AI-powered nursing information system.
• To compare AI-driven diagnostic suggestions with nurse-led diagnoses.

Main Methods:

• Utilized data mining and statistical analysis with a back-propagation neural network.
• Implemented an adaptive-network-based fuzzy inference system (ANFIS) for comparison.
• Assessed diagnostic agreement, question differences, and nurse satisfaction before and after system implementation.

Main Results:

• The AI system achieved up to an 87% agreement rate with diagnoses made by nurses.
• Calculated the probability of various nursing diagnoses based on patient characteristics.
• Identified differences in questioning and reported increased nurse satisfaction post-implementation.

Conclusions:

• AI can significantly enhance the accuracy and efficiency of nursing diagnoses.
• The developed nursing information system demonstrates a high level of agreement with expert clinical judgment.
• Integrating AI into nursing workflows holds promise for improving patient care and supporting healthcare professionals.