Abstract 12100: Marked Overestimation of Obstructive Coronary Artery Disease Probability by the Diamond-Forrester Classification in Sympomatic Patients Undergoing Coronary CT Angiography: Results from the Multinational CONFIRM Registry (Coroanry CT Angiograph Evaluation For Clinical Outcomes: An International Multicenter Registry)
Background: The Diamond-Forrester Classification (DFC) of age, sex, and angina typicality is widely used to establish pre-test probability of obstructive coronary artery disease (Ob-CAD, =50% diameter stenosis in a major coronary artery) in symptomatic patients. However, DFC probabilities were calculated using autopsy Results and data from a separately pooled group of studies with invasive angiography, rather than direct in vivo observation. We examined how rates of Ob-CAD in a contemporary population referred for coronary CT angiography (CCTA) compare to those reported by DFC.
Methods: From 5 centers in 4 countries, we studied 4775 consecutive patients (51% men, median age 59 years, 10% diabetics) without prior CAD who underwent CCTA. Symptoms were categorized into 4 groups: Non-anginal pain (NonAng), atypical angina (AtypAng), typical angina (TypAng), and dyspnea (Dysp). All CCTA were performed on 64-slice scanners and were evaluated for Ob-CAD using a modified 18-segment coronary artery tree model.
Results: Frequencies of NonAng, AtypAng, TypAng, and Dysp in this population were 18%, 45%, 20%, and 17%, respectively. Observed rates of Ob-CAD were much lower than the pooled data used by DFC in patients with AtypAng (20% in 2143 patients vs. 50% in 1931 patients, p<0.001) and TypAng (34% in 949 patients vs. 89% in 2108 patients, p<0.001). When patients were stratified by age, sex, and angina typicality as described in DFC, observed prevalence of Ob-CAD was markedly lower than that reported by DFC in every AtypAng and TypAng subgroup (Figure). The highest prevalence was found in 60–69 year-old men with TypAng (57.4%).
Conclusion: In this large multicenter patient cohort, rates of Ob-CAD in patients with atypical and typical angina were considerably lower than those calculated by DFC. These findings highlight a need to reassess the current symptom-based approach for estimating Ob-CAD probability in patients referred for non-invasive cardiac imaging.
- © 2010 by American Heart Association, Inc.