Abstract 16590: Incremental Value of Routine Preoperative Echocardiography in Patients Undergoing Coronary Artery Bypass Surgery: The Pre-Operative Surgical Stratification by Echocardiography (POSSE) Study
Background: Although echocardiography is commonly performed before coronary artery bypass surgery (CABG), there has yet to be a study examining the prognostic value of a comprehensive echocardiogram. As such, existing risk scores incorporate very few echocardiographic parameters and guidelines do not explicitly recommend routine echocardiography before CABG. We sought to determine the incremental value of echocardiography for predicting mortality or major morbidity in patients undergoing CABG.
Methods: We performed a cohort study of consecutive patients who underwent isolated CABG at the Massachusetts General Hospital in 2007-2009 (n=1,150). Patients who did not have an echocardiogram within 30 days of their surgery were excluded (n=483). A panel of quantitative echocardiographic parameters was measured in duplicate by two independent observers, and disagreements were resolved by a third. These parameters encompassed left and right-sided chamber size, geometry, systolic and diastolic function, and valvular function. Clinical variables were extracted from the STS database. The primary outcome was in-hospital postoperative mortality or major morbidity defined by the STS as stroke, renal failure, prolonged intubation, deep sternal wound infection, or reoperation.
Results: The cohort consisted of 667 patients with a mean age of 67 ± 11 years and 23% females. Excluded patients were similar to included patients. There were 103 observed events. After adjusting for the STS risk score, 3 echocardiographic parameters were found to be independently predictive of the primary outcome: biplane Simpson's left ventricular ejection fraction <30% (OR 2.47 95% CI 1.04, 5.84), indexed left atrial volume ≥34 mL/m2 (OR 1.88 95% CI 1.13, 3.12), and right ventricular myocardial performance index >0.40 (OR 1.91 95% CI 1.14, 3.21). These parameters resulted in a net improvement in model prediction as reflected by a change in c-statistic from 0.69 to 0.74 and an integrated discrimination index of 4% (95% CI 1%, 6%).
Conclusions: Three simple echocardiographic parameters reflecting systolic and diastolic function of the left and right ventricles add incremental value above existing risk scores for predicting mortality or major morbidity in patients undergoing CABG.
- © 2011 by American Heart Association, Inc.