VOJVODINASCORE - LOCAL SYSTEM FOR CARDIAC OPERATIVE RISK EVALUATION VOJVODINASKOR - LOKALNI SISTEM ZA EVALUACIJU KARDIOHIRURŠKOG OPERATIVNOG RIZIKA

Summary Introduction. The aim of the study was to investigate the prog- nostic value, sensitivity and specificity of both the logistic and ad- ditive European System for Cardiac Operative Risk Evaluation (as well as the European System for Cardiac Operative Risk Eval- uation II and to assess the necessity for developing a local out- come prediction model in cardiac surgery. Material and Meth- ods. The research included 406 consecutive patients who had un- dergone cardiac surgical procedures at Institute of Cardiovascular Diseases of Vojvodina from January 2012 to July 2012. The au- thors compared the predicted mortality according to the additive and logistic European Systems for Cardiac Operative Risk Evalu- ation, the new European System for Cardiac Operative Risk Eval- uation II and the observed mortality (30 days after surgery). Re- sults. The difference between the predicted and observed mortal- ity regarding the whole group of 406 operated cardiac patients was not statistically significant for the additive European System for Cardiac Operative Risk Evaluation (p=0.081) and the Europe- an System for Cardiac Operative Risk Evaluation II (p=0.164), but it was statistically significant for the logistic European System for Cardiac Operative Risk Evaluation (p=0.031). The areas under the receiver operating characteristic curves are statistically differ- ent from 0.5 for both models (additive and logistic European Sys- tem for Cardiac Operative Risk Evaluation), as well as for the Eu- ropean System for Cardiac Operative Risk Evaluation II. Howev- er, the proper classification of the patients has not been observed since their sensitivity and specificity are not satisfactory. Conclu- sion. The additive and logistic European Systems for Cardiac Op- erative Risk Evaluation overestimate while the European System for Cardiac Operative Risk Evaluation II underestimates the risk in cardiac surgery. We believe that a locally derived model would be of great use in the everyday clinical practice since it would faithfully illustrate the actual state of patient population of the re- gion where it was developed. At the same time it would provide the accurate prediction of surgical outcome.