The German Aortic Valve Score II.

ABSTRACT

OBJECTIVES: The German Aortic Valve Score (GAVS) was developed for national quality assurance regarding the in-hospital mortality rate of patients following isolated aortic valve replacement. The goal of this work was the recalibration of the GAVS in the context of increased numbers of transcatheter aortic valve implantations. METHODS: In 2011 and 2012, 36 183 cases were documented who had either surgical aortic valve replacement or transcatheter aortic valve implantation (45%). All cases were randomly assigned to the study or to the validation group. All items of the data set were checked for significance by developing a multiregression risk model using iterative backward elimination. Calibration was ascertained using the Hosmer-Lemeshow method. To define the quality of discrimination, the area under the receiver operating characteristic curve (C-statistic) was calculated. RESULTS: The randomized study cohort comprised 18 054 patients. After modelling with multiple regression algorithms, 18 of the initial 28 risk factors entered the risk model. When applied to the validation group, the newly developed GAVS II showed good calibration with a P-value of 0.411 in the Hosmer-Lemeshow test and good discrimination with a C-statistic of 0.741. CONCLUSIONS: The GAVS II is a new risk model that is applicable to cohorts having surgical aortic valve replacement or transcatheter aortic valve implantation procedures.