Diagnostic accuracy of seismocardiography compared with electrocardiography for the anatomic and physiologic diagnosis of coronary artery disease during exercise testing.

A multicenter study was performed to compare the diagnostic accuracy of a new technique, seismocardiography, with that of electrocardiography for physiologically and anatomically significant ischemic coronary artery disease (CAD) during exercise stress testing. Five participating centers enrolled 129 patients who had simultaneous seismocardiograms and 12-lead electrocardiograms at the time of their exercise treadmill stress tests. Two different definitions of CAD were used: anatomic and physiologically significant disease. The presence of anatomically significant CAD (> or = 50% diameter stenosis) was documented by coronary angiography. Physiologically significant CAD was defined as present in the same 129 patients when coronary arteriography (> or = 50% diameter stenosis) and thallium-201 scintigraphy (defect on initial postexercise images) were both abnormal. Seismocardiography had a significantly better sensitivity for detecting anatomic CAD than did electrocardiography (73 vs 48%; p < 0.001), without loss of specificity (78 vs 80%; p = NS). Exercise seismocardiography added significant incremental diagnostic information beyond that provided by exercise electrocardiography. Seismocardiography was more sensitive (without less specificity) in women and in patients who did not achieve maximal predicted heart rate. In patients with physiologically significant CAD, the seismocardiogram was also significantly more sensitive (78%) than was the electrocardiogram (55%) (p < 0.02), without loss of specificity (84 vs 74%). Seismocardiography significantly improved sensitivity for the detection of anatomic and physiologic CAD. It is easy to perform and may be a clinically useful adjunct in exercise stress testing.