Identifying Trigger Cues for Hospital Blood Transfusions Based on Ensemble Learning Methods

Area of Science:

• Trauma resuscitation
• Emergency medicine
• Critical care

Background:

• Traumatic shock is a leading cause of preventable death, with most fatalities occurring within 6 hours.
• Prehospital interventions, including early blood transfusion, are crucial for improving survival in patients with hemorrhagic shock.
• Optimizing transfusion triggers in the prehospital setting is essential for managing occult shock.

Approach:

• Utilized logistic regression, Fast and Frugal Trees (FFTs), and Bayesian analysis.
• Identified factors associated with early in-hospital blood transfusion requirements.
• Developed a simple algorithm for prehospital transfusion decisions.

Key Points:

• A dataset of 2,157 trauma patients transported to a Level I trauma center was analyzed.
• Four key factors identified for the FFT algorithm: systolic blood pressure (SBP), prehospital lactate, Shock Index, and Abbreviated Injury Scale (AIS) of the chest.
• Prehospital lactate concentration was a significant predictor of early transfusion needs.

Conclusions:

• A novel, clinically relevant prehospital algorithm was developed using a combination of statistical and machine learning methods.
• The algorithm assists in identifying patients requiring transfusion within 4 hours of hospital arrival.
• This tool supports timely and appropriate blood product administration in the prehospital setting.