[A quality improvement program in cardiac surgery. Four-year experience from the Ospedali Riuniti of Trieste].

Background: Dissatisfaction with clinical outcomes prompted an intervention to assess and improve processes and outcomes in a cardiac surgery unit.

Methods: Starting on September 1st, 1998, 1836 consecutive patients requiring a heart operation in our unit were prospectively enrolled by recording a series of anamnestic, clinical and procedural descriptors in a dedicated database. Expected mortality rates were estimated by means of nine different stratification models, one of which also allowed the prediction of excess intensive care unit and total hospital length of stay.