Maximizing efficiency in emergency care: triple interventions to minimize left without being seen: An observational...

Area of Science:

• Health Informatics
• Artificial Intelligence in Healthcare
• Emergency Medicine Operations

Background:

• Left Without Being Seen (LWBS) is a critical quality metric in emergency departments (EDs) impacting patient outcomes.
• Identifying and mitigating factors contributing to LWBS is complex, requiring advanced analytical methods.
• Previous interventions have shown varied success, necessitating a data-driven approach.

Purpose of the Study:

• To leverage Artificial Intelligence and Machine Learning (AI/ML) algorithms to identify key risk factors for LWBS.
• To implement targeted triple interventions addressing identified risks.
• To compare daily LWBS rates before and after the implementation of these interventions.

Main Methods:

• Retrospective observational study analyzing daily ED throughput data from March 2019 to May 2024.
• Utilized Extreme Gradient Boosting (XGBoost) and Random Forest AI/ML algorithms for LWBS prediction and risk factor identification.
• Implemented triple interventions: rapid triage, direct bedding, and boarding reduction, followed by a comparative analysis of LWBS rates.

Main Results:

• AI/ML models demonstrated favorable performance in predicting LWBS.
• Key factors identified influencing LWBS were: triage-to-bed wait time, boarding time, and door-to-triage time.
• The average daily LWBS rate decreased from 4.82% before interventions to 1.93% after interventions (P < .001), a reduction of approximately 60%.

Conclusions:

• AI/ML approaches are effective in identifying critical factors associated with LWBS in EDs.
• Targeted triple interventions addressing triage, bedding, and boarding times significantly reduce daily LWBS rates.
• This study highlights the efficiency of AI/ML-driven operational management in improving ED patient flow and care quality.