Development of a Mapping Table for Nursing Notes Based on Nurses' Concerns in ICU Patients

Area of Science:

• Nursing Informatics
• Clinical Data Analysis

Background:

• Unstructured nursing notes contain valuable information on patient status.
• Standardizing this data is crucial for advanced clinical analysis and prediction.
• Identifying nurses' concerns is key to understanding patient well-being in ICUs.

Purpose of the Study:

• To develop a mapping table linking nurses' concerns from ICU nursing notes to standard terminology.
• To facilitate the use of nursing notes for clinical deterioration prediction.
• To leverage natural language processing for identifying critical patient information.

Main Methods:

• Literature review on nurses' concerns in critical care.
• Extraction and deduplication of nursing notes from the MIMIC IV database.
• Definition, coding, and mapping of identified nurses' concerns using the CCC (Clinical Classification for Healthcare) framework.

Main Results:

• 265 unique nursing notes were identified from over 11 million records.
• 208 notes were categorized as reflecting nurses' concerns.
• These concerns were grouped into 15 categories aligned with the CCC framework.

Conclusions:

• A novel mapping table effectively links nurses' concerns to standard terminology.
• This table serves as a foundational resource for predicting clinical deterioration in ICU patients.
• Natural language processing can identify significant lexicons within nursing notes for improved patient monitoring.