We investigated if gated and/or attenuation-corrected SPECT improved performance of experienced or in-experienced operators in the diagnosis of coronary artery disease (CAD) or stenosed arteries in patients having intermediate or high pretest likelihood.

Rest/dipyridamole stress gated attenuation-corrected SPECT was performed in 81 patients (M:F=48:33, 62±8.6 years old, of high (n=38) or intermediate (n=43) pre-test likelihood for CAD, 1 vessel; 14, 2 vessel; 19, 3 vessel disease; 31, normal; 17). Two experienced and one novice physicians graded 1 (normal) to 5 (definitely abnormal) for each artery 1) by conventional rest Tl-201/stress Tc-99m-MIBI SPECT, 2) 1) plus viewing gated SPECT (+gated) and 3) 2) plus attenuation-corrected SPECT (+AC gated). Areas under curves (AUC) of receiver operating characteristic (ROC) curves were compared.

AUC was greater in patients with high likelihood than intermediate likelihood. Novice physician performed better in diagnosing RCA stenosis (AUC: 0.53 for usual, 0.58 for +gated, and 0.66 for +AC gated) in patients with intermediate likelihood. AUC was not different for the diagnosis of CAD and LAD or LCx. Performance of experienced physicians was not different regardless of pre-test likelihood. Novice operator's specificity increased from 53% to 73% for CAD, and 62% to 89% (p<0.05) for RCA in patients with intermediate likelihood.

We conclude that gated attenuation-corrected SPECT was helpful only for novice physician to diagnose RCA stenosis by increasing specificity in patients with intermediate likelihood. Even for inexperienced physician, these methods were not helpful in patients with high pre-test likelihood or for LAD or LCX.