Benchmark of Intraoperative Activity in Cardiac Surgery: A Comparison between Pre- and Post-Operative Prognostic...

Area of Science:

• Cardiovascular Surgery
• Medical Informatics
• Health Services Research

Background:

• Cardiac surgery carries significant complication risks despite safety advancements.
• Existing predictive scores (e.g., EUROSCORE, STS, ACEF) rely solely on pre-operative data, neglecting the intra-operative period.
• There is a need for models that capture dynamic risk changes during surgery.

Purpose of the Study:

• To develop and validate a novel double-fold predictive model for cardiac surgery outcomes.
• To evaluate patient risk evolution by integrating pre-operative and post-operative data.
• To assess the direct impact of the intra-operative period on patient mortality.

Main Methods:

• Analysis of 15,882 cardiac surgery patients from the Margherita-Prosafe cohort study.
• Development of two logistic regression models: one pre-operative and one including post-operative data.
• Testing model calibration and discrimination using Area Under the Curve (AUC).

Main Results:

• The pre-operative model achieved an AUC of 0.81, while the post-operative model showed improved discrimination with an AUC of 0.87.
• Significant variations in pre- and post-operative mortality differences were observed across centers.
• The double-fold model demonstrated utility in benchmarking medical care within hospitals.

Conclusions:

• Post-operative mortality predictions differ significantly from pre-operative ones, quantifying surgical impact.
• The double-fold model effectively assesses intra-operative team performance and patient risk trajectory.
• This model can facilitate hospital benchmarking and drive improvements in cardiac surgery care.