CARDIAC COMPLICATIONS IN NONCARDIAC SURGERY - RELATIVE VALUE OF RESTING 2-DIMENSIONAL ECHOCARDIOGRAPHY AND DIPYRIDAMOLE-THALLIUM IMAGING

Although perfusion imaging studies are extensively used as a preoperat ive screening test for risk stratification of patients undergoing nonc ardiac surgery, no single cardiac noninvasive test has been shown to b e ideal for risk stratification. We investigated the relative impact o f transthoracic two-dimensional echocardiography (ECHO) compared with dipyridamole thallium scintigraphy (DT) in predicting major cardiac co mplications in patients undergoing noncardiac surgery. Eighty-seven co nsecutive patients undergoing 96 procedures (56 vascular, 40 general) underwent preoperative evaluation first with DT and then with ECHO bef ore surgery. Complications were prospectively defined as myocardial in farction (MI), cardiac death (of MI, heart failure, or arrhythmia), or need of revascularization before surgery. DT showed one or more rever sible defects in 44 (51%) patients, whereas ECHO demonstrated a reduce d left ventricular ejection fraction (LVEF) in 25 (29%) patients. Majo r postoperative cardiac complications occurred in 10 patients: 5 death s (2 ventricular fibrillation, 3 fatal MIs) and 5 nonfatal MIs. Four a dditional patients required urgent revascularization (coronary bypass graft surgery in 3 and percutaneous transluminal coronary angioplasty in 1). Of the 20 patients with both abnormal DT and ECHO, 11 (55%) had major complications, compared with none of the 26 (0%; p < 0.01) with an abnormal DT but normal LVEF. The sensitivity of DT and ECHO were n ot significantly different (79% [95% CI, 52% to 93%] vs 86% [60% to 96 %], respectively), although the specificity of DT was lower (51% [40% to 62%] vs 81% [70% to 88%]; p < 0.05). The positive predictive value of DT was significantly improved from 22% (12% to 35%) to 52% (32% to 72%) when both DT and ECHO were abnormal. The results were not signifi cantly different when the 4 patients who underwent revascularization w ere excluded. in conclusion, (1) in spite of similar sensitivity of EC HO and DT, ECHO appears to be relatively more specific in predicting m ajor CC, and (2) when ECHO and DT are both abnormal, the risk of CC re lated to noncardiac surgery is significantly increased. Use of the com bination of UT and ECHO before major noncardiac surgery can improve th e identification of patients at risk for complications.