Sequential logistic models for 30 days mortality after CABG: pre-operative, intra-operative and post-operative...

Area of Science:

• Cardiovascular Surgery
• Medical Informatics
• Health Services Research

Background:

• Coronary artery bypass grafting (CABG) is a major surgical procedure with inherent risks.
• Accurate prediction of 30-day mortality is crucial for patient management and resource allocation.
• Existing risk models often focus on pre-operative patient characteristics (case-mix).

Purpose of the Study:

• To evaluate the incremental predictive value of operative and post-operative variables on 30-day mortality.
• To assess how these peri-operative factors enhance existing case-mix models.
• To provide clinical insights into the impact of hospital experience on surgical outcomes.

Main Methods:

• A prospective study involving 4835 patients undergoing CABG in Israel in 1994.
• Sequential logistic regression models were used, incorporating risk factors chronologically: case-mix, operative, and post-operative.
• Data were collected from admission through the immediate post-operative period, with independent ascertainment of deaths.

Main Results:

• The case-mix model (pre-operative factors) achieved a c-statistic of 78.8%, identifying factors like left ventricular dysfunction and emergency admission.
• The operative model, including factors like operation duration and bleeding, improved prediction to a c-statistic of 85.3%.
• The post-operative model, incorporating factors such as low hemoglobin and prolonged respirator use, further enhanced prediction to a c-statistic of 92.4%.

Conclusions:

• Sequential analysis effectively updates patient risk stratification, especially with limited events relative to risk factors.
• Peri-operative factors significantly increase the predictive power of mortality models.
• These findings highlight the importance of intra-hospital events and hospital experience in influencing 30-day mortality after CABG.