PII: S0003-4975(99)01511-8

Abstract

Historically, cardiothoracic surgeons have been at the forefront of demonstrating the effectiveness of their surgical procedures. This article by Ferguson and colleagues is yet another example of how the cardiothoracic surgical community has united to measure and report their outcomes. A decade ago, visionaries developed and implemented the concept of a national database that would measure and risk stratify outcomes from cardiac surgical procedures. Along the way they learned much about element definitions, database interfaces, data quality assurance, model development, and a fickle computer industry. That initial vision has grown to almost 450 participants and 1.3 million procedures, and the task has become so great that responsibilities now have been distributed. Eight software vendors will provide entry and local analysis software. The Duke Clinical Research Institute (DCRI) was chosen as the data warehouse and analytic site, and a committee has been formed to aid in prioritizing ad hoc queries to the database from outside individuals. This efficient structure will allow easier growth to 100% participation of surgical sites, routine regional reporting, developing and answering research questions, and developing economic outcome measures. However, data collection and simple analyses are insufficient endpoints for such a lofty task. Mortality and morbidity rates are low, and further reductions may be difficult. What else should a national database be doing? First, the database technology always must be cutting edge and must be “light on its feet” in responding to new techniques such as minimally invasive surgery, new conduits, and even robotics. Second, the data must be analyzed in new and innovative ways. Because of advances in angioplasty and stenting—competing procedures to coronary artery bypass grafting operations—we must be able to determine if patients are receiving the optimal therapy for their individual conditions. The STS Database could help define “quality,” perhaps not in terms of morbidity and mortality alone, but by considering the process of care. Formal measures of appropriateness and benefit could clearly demonstrate that a particular procedure performed was indicated—something that has not been done with angioplasty. A link to a national cardiology database could be synergistic in defining optimal care in patients with coronary disease. Finally, the STS must decide whether this marvelous system should generate data for local use only, or whether there should be formal “triggers” that would initiate STS processes to aid centers with adverse outcomes. The cardiothoracic surgical community in general, and the STS specifically, is to be congratulated for their outstanding development of a solid data collection and analytic foundation. We anxiously await the construction of the next level of this important edifice.