Examination of nursing data elements from evidence-based recommendations for clinical decision support

Area of Science:

• Nursing Informatics
• Clinical Data Management
• Healthcare Systems Engineering

Background:

• Delirium in intensive care units (ICUs) poses a significant patient risk.
• Effective prevention strategies for ICU delirium require robust data capture.
• Electronic health records (EHRs) are crucial for collecting patient data.

Purpose of the Study:

• To assess the current use, metadata, and availability of delirium data elements within EHRs and clinical data repositories.
• To compare nursing practice data elements for ICU delirium risk with evidence-based recommendations.
• To identify data-related challenges for developing clinical decision support systems to prevent ICU delirium.

Main Methods:

• A descriptive, exploratory study using content analysis and descriptive statistics.
• Analysis of instance-level data from a convenience sample of 1714 patients.
• Matching 160 EHR data elements against 41 derived data element categories based on nursing process components.

Main Results:

• A high number of potential data element matches were identified.
• Matched data elements were primarily text-based, entered by registered nurses (RNs).
• Considerable variability in data availability was observed due to clinical, conceptual, and technical factors.

Conclusions:

• Current EHR data elements for delirium are often text-based with variable availability.
• Significant challenges exist in data quality and accessibility for ICU delirium prevention.
• Development of valid, reliable data capturing nurse-patient-family interactions is essential for future clinical decision support systems.