Predictive Analytics in Cardiothoracic Care: Enhancing Outcomes with the Healthcare Enabled by Artificial...

Area of Science:

• Cardiovascular Surgery
• Artificial Intelligence in Healthcare
• Predictive Analytics

Background:

• Postoperative complications after cardiac surgery significantly impact patient survival and increase healthcare expenditures.
• Effective prevention of adverse events is crucial for improving patient outcomes, reducing morbidity, and lowering costs.

Purpose of the Study:

• To develop a real-time predictive analytics model to support clinical decision-making for patients undergoing cardiac surgery.
• The Healthcare Enabled by Artificial Intelligence in Real Time (HEART) project aims to predict various adverse outcomes.

Main Methods:

• A supervised, closed-loop machine learning design is employed to train the predictive model.
• Static and dynamic patient variables are collected from preoperative, intraoperative, and postoperative periods.
• Data, including blood product transfusions and medications, are transmitted from EHR to a data warehouse.

Main Results:

• The HEART project has successfully established a data-collecting infrastructure.
• Data collection and validation are ongoing, prioritizing accuracy and completeness.

Conclusions:

• The HEART project is developing a real-time decision support tool for cardiothoracic intensive care units.
• The ultimate goal is to integrate this AI tool into routine patient care to enhance clinical decision-making and improve postoperative outcomes.