IDENTIFICATION OF PREOPERATIVE VARIABLES NEEDED FOR RISK ADJUSTMENT OF SHORT-TERM MORTALITY AFTER CORONARY-ARTERY BYPASS GRAFT-SURGERY

Objectives. The purpose of this consensus effort was to define and pri oritize the importance of a set of clinical variables useful for monit oring and improving the short-term mortality of patients undergoing co ronary artery bypass graft surgery (CABG). Background. Despite widespr ead use of data bases to monitor the outcome of patients undergoing CA BG, no consistent set of clinical variables has been defined fur risk adjustment of observed outcomes for baseline differences in disease se verity among patients. Methods. Experts with a background in epidemiol ogy biostatistics and clinical care with an interest in assessing outc omes of CABG derived from previous work with professional societies, g overnment or academic institutions volunteered to participate in this unsponsored consensus process, Two meetings of this ad hoc working gro up were required to define and prioritize clinical variables into core , level 1 or level 2 groupings to reflect their importance for relatin g to short-term mortality after CABG. Definitions of these 44 variable s were simple and specific to enhance objectivity of the 7 care, 13 le vel 1 and 24 level 2 variables, Core and level 1 variables were evalua ted using data from five existing data bases, and core variables only were examined in an additional two data bases to confirm the consensus opinion of the relative prognostic power of each variable, Results, M ultivariable logistic regression models of the seven core variables sh owed all to be predictive of bypass surgery mortality in some of the s even existing data sets, Variables relating to acuteness, age and prev ious operation proved to be the most important in all data sets tested , Variables describing coronary anatomy appeared to be least significa nt. Models including both the 7 core and 13 level I variables in five of the seven data sets showed the core variables to reflect 45% to 83% of the predictive information. However, some level I variables were s tronger than some core variables in some data sets, Conclusions. A rel atively small number of clinical variables provide a large amount of p rognostic information in patients undergoing CABG.