Background: A specific risk model concerning mortality of patients undergoing isolated coronary artery bypass grafting (CABG) is developed based on the national quality benchmarking mandatory by law in Germany.
Methods: On the basis of the national data pool from 2004, a risk score model for patients undergoing isolated CABG was developed and finally adjusted with the data of 43,145 patients of the year 2008. Modeling was performed by logistic regression analysis. This risk model was validated with the 2007 data pool which comprised 45,569 patients.
Results: Observed in-hospital mortality after isolated CABG procedures was 3.0% in 2008. Hosmer-Lemeshow test p value was 0.189 and area under receiver operating characteristic curve was 0.826. Applying the German CABG score for 2007 resulted in an observed-to-expected mortality ratio of 1.01.
Conclusion: The German CABG score for in-hospital mortality is a risk score with proven validity for isolated CABG, developed by means of the patient population in Germany. It can be used for the assessment of patient risk groups and for interhospital benchmarking. We encourage other researchers to apply and validate this score in comparable health care systems.
Georg Thieme Verlag KG Stuttgart · New York.