Abstract 1643: Does the ACCF/ASNC Appropriateness Criteria Predict Post Test Resource Utilization in Patients Without Known Coronary Artery Disease?
Background: The ACCF/ASNC Appropriateness Criteria for Single-Photon Emission Computed Tomography Myocardial Perfusion Imaging (SPECT MPI) project was initiated to ensure the effective use of advanced diagnostic imaging tools. However, the impact of these criteria on early post test resource utilization is not known.
Methods: We studied 90 day post SPECT-MPI resource utilization in 5943 consecutive patients (Mean age 61 years, 42% males) with no known coronary artery disease (CAD) that underwent SPECT-MPI for clinical indications between 3/02 and 10/06. SPECT-MPI was interpreted semi-quantitatively using a 17-segment heart model. Indications for SPECT-MPI were evaluated.Patients were scored using their pre-test probability of CAD for symptomatic patients, Framingham risk for asymptomatic patients, baseline ECG and ability to exercise. Based on the appropriateness criteria, tests were scored as having an appropriate (A), uncertain (U) or inappropriate (I) indications. Patients were followed up using review of electronic medical records for post test coronary angiography and percutaneous or surgical coronary revascularization.
Results: The majority of the tests were A (75.3%) vs. I (13.1%) and U (11.6%). African Americans and Women had more often A studies. While I studies were more often normal (91% vs. 84% for A), there was no difference in the rate of high risk scan between two groups (22.5% for I vs. 21% for A, p=NS). In patients with abnormal studies, the referral for cardiac catheterization (39% for A, 43% for U and 32% for I, p=NS) and early revascularization was not statistically different between the three groups (25% for A, 30% for U and 32% for I, p=NS). Using binary logistic regression, the independent predictors of referral to cardiac catheterization were abnormal SPECT-MPI, high risk scan findings, and low ejection fraction (p<.001 for all), while the category of appropriateness criteria was not predictive.
Conclusions: In a single-center university hospital setting, the yield of testing in I patients is low. However, the post-test resource utilization in the setting of an abnormal scan was the same in patients with I as compared to A test indications. The poor yield of inappropriate SPECT MPI scans will have cost effectiveness consequences.