Background. Stress electrocardiography (ECG)-gated single photon emission computed tomography (SPECT) for assessment of left ventricular perfusion and function improves the confidence of interpretation and enhances specificity for detection of coronary artery disease. The reproducibility of visual interpretation of ECG-gated SPECT images and the significance of training and experience have not been reported previously in a large series of consecutive patients. We evaluated both intraobserver and interobserver agreement of interpretation of ECG-gated SPECT images among 3 cardiology trainees and 3 experienced nuclear cardiologists from 3 institutions. Methods and Results. Three nuclear cardiologists and 3 cardiology trainees who had fulfilled American College of Cardiology/American Society of Nuclear Cardiology Core Cardiology Training Symposium (ACC/ASNC COCATS) guidelines for level II training in nuclear cardiology independently evaluated 106 consecutive technetium 99m sestamibi SPECT images with ECG gating of either the stress or rest images. All cases were interpreted blindly, twice in random sequence, without clinical data. We assessed intraobserver and interobserver agreement for myocardial perfusion, left ventricular regional and global systolic function, and overall clinical impression, by means of percent agreement and Cohen's κ statistic. Intraobserver agreement was good (82%–92%, κ = 0.54–0.84) for assessment of myocardial perfusion, systolic function, and overall impression. Interobserver agreement was also good, ranging from 65% to 90% (κ = 0.32–0.76), with better agreement found for assessment of function (77%–85%, κ = 0.52–0.7) than for perfusion (65%–80%, κ = 0.32–0.6). For all measures, there were no significant differences in reproducibility between nuclear cardiologists and cardiology trainees. Conclusions. Interpretation of ECG-gated SPECT images has high reproducibility and agreement among both nuclear cardiologists and cardiology trainees. (J Nucl Cardiol 2002;9:263-70.)
Read full abstract