Predictors of Normal Single Photon Emission Computed Tomography Myocardial Perfusion Imaging (original) (raw)

2010, Journal of the American College of Cardiology

Background: The growth of cardiac imaging has led to recent concerns over the effective use of imaging in the evaluation of patients for cardiac disease. The aim of this study was to evaluate the results of inappropriate referrals for SPECT MPI based on the 2009 ACC/ASNC published appropriateness criteria (AC). Methods: Consecutive patients from February 2009 to April 2009 were prospectively studied. Based on AC, SPECT MPI studies were classified into appropriate (A), inappropriate (I), uncertain (U) or unclassified (X). Results: 1121 patients (61% males) with mean age 61 years (SD 12) were studied. Bruce treadmill SPECT MPI was used in 56%, dipyridamole stress in 43% and dobutamine stress in 1%. The most common indications for SPECT MPI were chest pain evaluation for coronary artery disease (32%), evaluation of chest pain syndrome post-revascularization (17.4%), preoperative evaluation for non-cardiac surgery (11%) and assessment of asymptomatic patients with low risk (4%). Applying the AC, 859 (77.5%) were A, 130 (11.7%) were I, 56 (5.0%) were U, 64 (5.8%) were X. The agreement in characterizing appropriateness between two cardiology consultants was moderate (kappa=0.64). A grading of inappropriate was more likely in young patients (< 60 years) (OR 2.42 95%CI 1.69-3.45 p<0.001), asymptomatic patients at low risk (OR 66.91 95%CI 25.60-174.84 p<0.001) and preoperative evaluation of patients for non-cardiac surgery (OR 7.51 95%CI 4.86-11.61 p<0.001). In particular, inappropriate referrals for asymptomatic low risk individuals, females and preoperative evaluation for non-cardiac surgery were associated with normal MPI results (OR 3.85 95%CI 1.53-9.64 p=0.01, OR 2.01 95%CI 1.14-3.56 p=0.018, OR 4.76 95%CI 2.56-8.86 p<0.001). Conclusions: Normal MPI results in patients inappropriately referred for SPECT MPI suggests that the appropriateness criteria is fairly robust in predicting the presence of cardiac disease particularly in asymptomatic low risk patients, females and patients who require preoperative evaluation. These findings may be used to support measures to promote effective use of this imaging modality.