Crystal balls

Knowing the likelihood of survival after cardiac surgery is useful. If we know the likely outcome, we can compare it with actual outcome and thus gain some insight into the overall performance of the cardiac surgical unit. Knowledge of who is likely to develop major morbidity also has an impact on the use of valuable resources and may allow for sensible planning of operating lists. In addition, some believe that being able to predict mortality with some certitude may help clinicians to determine when further efforts are futile. Unfortunately, the perfect predictor – a crystal ball to foresee the future – has not yet been fully developed.

Risk models or scoring systems

Scoring systems allow reasonable prediction of outcome after cardiac surgery. Many models have been devised to work out the likelihood of survival, and these and others have also been shown to predict major morbidity, long-term survival and resource use with some accuracy. Models can be broadly divided into two groups:

1. • preoperative models, applied before the operation, with no knowledge of intraoperative events; and

2. • postoperative models, applied immediately after the operation on admission into the critical care unit, taking some account of what the operation did to the patient.