Emergency Severity Index: accuracy in risk classification

Area of Science:

• Emergency Medicine
• Nursing Research
• Healthcare Resource Management

Background:

• Accurate prediction of patient resource needs in emergency departments is crucial for efficient healthcare delivery.
• The Emergency Severity Index (ESI) is a widely used triage tool, but its adapted versions require validation.
• Understanding factors influencing nurses' predictive accuracy is essential for optimizing triage processes.

Purpose of the Study:

• To assess the agreement between predicted and actual resources used by patients based on the adapted ESI.
• To investigate the correlation between nurses' experience (years since graduation, work experience, emergency service experience) and their accuracy in resource prediction.

Main Methods:

• A retrospective analytical study involving 538 patient records.
• Data collected from triage nurses between December 2012 and February 2013.
• Quantitative analysis of resource utilization and nurse experience variables.

Main Results:

• A low Kappa agreement coefficient (0.34) indicated poor concordance between predicted and actual resource use.
• No significant association was found between nurses' predictive accuracy and their years of graduation, work experience, or emergency service experience.
• Diagnostic exam predictions by nurses showed low agreement with those actually performed.

Conclusions:

• The adapted ESI demonstrated lower accuracy in predicting patient resources compared to the original scale in existing literature.
• Nurses' experience levels did not correlate with improved accuracy in predicting necessary patient care resources.
• Findings suggest a need for further refinement of triage tools and training to enhance resource prediction accuracy in emergency settings.