PROCESSES PREDICTIVE OF CABG COMPLICATIONS

University Of Colorado Denver, Aurora CO

Investigators

Abstract

BACKGROUND: According to the American Heart Association, cardiovascular disease ranks number on the united States as the leading cause of death. In 1989, the cost of coronary artery bypass graft (CABG) surgery represented 1.5% of the nation's total health care expenditures. In 1995, approximately 573,000 CABG surgeries were done in the United States at an average cost of $44,820 per procedure. For this high cost and high volume procedure, it is critically important to identify the processes of care that clinicians or administrators may change to improve patient outcomes. PURPOSE: The purpose of this research is to provide a scientific basis for modifying complications (morbidity) for CABG procedures. SPECIFIC AIMS: The specific aims for this study are to determine the independent effect of the pre-operative, intraoperative, and supervisory processes of care upon patient risk-adjusted outcomes after CABGF surgery. METHODS: The study population will include all first time, CABG-only patients in the processes, Structures, and Outcomes of Care in Cardiac Surgery (PSOCS) database. Using over 22,000 records from the Continuous Improvement in Cardiac Surgery Program (CICSP) database, a logistic regression risk model will be build to predict the presence of major complications or mortality after CABG surgery. To conserve degrees of freedom, this CICSP risk model will then be applied to the 3,605 PSOCS CABG records. By sequentially adding process variables to this baseline model, the specific actions that care providers records. By sequentially adding process variables to this baseline model, the specific actions that care providers may take to improve risk-adjusted outcomes will be identified. Specific processes of care to be evaluated include: 1) the completeness and quality of preoperative evaluations (167 variables), 2) the completeness and quality of the surgical procedure (182 variables), and 3) the degree of supervision by senior physicians (25 variables). As part of the PSOCS data reduction, these 374 process of care variables have been reduced to a total of 53 cores (13, 33, and 7 scores for these processes, respectively). Final logistic models that include statistically significant process scores will be reported. For the final study models, performance will be assessed using traditional indicators such as c-indices and Hosmer-Lemeshow statistics. IMPACT: In summary, this dissertation research project will identify the processes of care that impact risk-adjusted CABG surgery outcomes. Thus, the specific actions that care providers may take to improve cardiac surgical patient outcomes in the future will be identified.