Fast-track failure after cardiac surgery: external model validation and implications to ICU bed utilization

Area of Science:

• Cardiothoracic Surgery
• Critical Care Medicine
• Health Services Research

Background:

• Perioperative care pathways aim to optimize patient recovery after surgery.
• Fast-track protocols expedite recovery for select cardiac surgery patients.
• Predicting fast-track failure is crucial for efficient intensive care unit (ICU) bed allocation.

Purpose of the Study:

• To externally validate a fast-track failure risk prediction model in adult cardiac surgery patients.
• To assess the clinical impact of this model on ICU bed utilization.

Main Methods:

• Prospective cohort study design.
• Application of the St. Mary's Hospital fast-track failure risk model to a database of 1,597 adult cardiac surgery patients (2006-2011).
• Model performance evaluated using discrimination (AUC) and calibration (Hosmer-Lemeshow), with decision curve analysis for ICU bed utilization impact.

Main Results:

• 175 out of 1,597 patients (11%) failed fast-track management.
• The validated model demonstrated excellent discrimination (AUC = 0.82) and adequate calibration.
• Using the model with a 5-20% failure probability threshold could increase effective ICU bed utilization by 23-67%.

Conclusions:

• The validated fast-track failure prediction model performs well and can reliably estimate individual patient risk.
• Implementing this model offers substantial benefits for effective ICU bed management in cardiac surgery settings.