Prognostic information in administrative co-morbidity data following coronary artery bypass grafting☆

Abstract

The aim of this study was to evaluate the prognostic information obtainable from administrative data with respect to 30-day mortality following coronary artery bypass grafting (CABG) and to compare it with the European System for Cardiac Operative Risk Evaluation (EuroSCORE) recorded in a clinical database. We used a co-morbidity index calculated from administrative data in the Danish National Patient Register by means of all admissions 1 year prior to CABG. In addition, each CABG was categorised as being isolated or not, and acute or not. The prognostic power of the co-morbidity index was compared to that achieved using EuroSCORE from a clinical database comprising information on all patients treated with CABG in Denmark. The outcome was all-cause mortality within 30 days after CABG and the prognostic power was evaluated using logistic regression analyses. We identified 20078 patients treated with CABG from 2000 to 2007 with a complete registration of the total additive EuroSCORE in the clinical database. The co-morbidity index carried significant prognostic information regarding 30-day mortality (c-statistic 0.81). The prognostic power of the co-morbidity index was equal to that of the EuroSCORE (c-statistic 0.79). A standard co-morbidity index based on administrative data as well as on clinical data has proven equally useful for prediction of mortality amongst CABG patients.