A Clinical Risk Prediction Model for Identifying Patient Candidates for Same-day Discharge After Transcatheter Aortic...

Area of Science:

• Cardiology
• Medical Technology
• Health Informatics

Background:

• Same-day discharge after transcatheter aortic valve replacement (TAVR) is possible for select patients.
• Defining a low risk for adverse clinical events is crucial for same-day discharge planning.

Purpose of the Study:

• To develop a clinical risk prediction model for same-day TAVR discharge.
• To identify patients with a low probability of in-hospital adverse events post-TAVR.

Main Methods:

• A random forest machine learning algorithm was employed to build the prediction model.
• Patients were stratified into low, moderate, and high-risk groups based on calculated scores.
• The model utilized 9 clinical parameters to predict adverse events.

Main Results:

• The model identified 23.6% of patients as low-risk for same-day discharge (event rate <3% within 6 hours).
• The low-risk group experienced no in-hospital events within 6 hours and no 30-day mortality.
• External validation demonstrated 94% sensitivity in predicting overall adverse events.

Conclusions:

• Approximately 1 in 4 TAVR patients may be candidates for same-day discharge.
• This prediction model can effectively identify suitable candidates for early discharge.
• Findings may inform hospital resource allocation for TAVR procedures.