Exercise electrocardiography for Pre-Test Assessment of the Likelihood of Coronary Artery Disease

Objectives. To develop a tool including exercise electrocardiography (ExECG) for patient-specific clinical likelihood estimation of patients with suspected obstructive coronary artery disease (CAD).
Methods. An ExECG-weighted clinical likelihood (ExECG-CL) model was developed in a Training cohort of patients with suspected obstructive CAD undergoing ExECG. Secondly, the ExECG-CL model was applied in a CAD Validation cohort undergoing ExECG and clinically driven invasive coronary angiography and a Prognosis Validation cohort and compared to the risk factor-weighted clinical likelihood (RF-CL) model for obstructive CAD discrimination and prognostication, respectively.
In the CAD Validation cohort, obstructive CAD was defined as >50% diameter stenosis on invasive coronary angiography. For prognosis, the endpoint was non-fatal myocardial infarction and death.
Results. The Training cohort consisted of 1,214 patients (mean age 57years, 57% males). In the CAD (N=408; mean age 55years, 53% males) and Prognosis Validation (N=3,283; mean age 57years, 57% males) cohorts, 11.8% patients had obstructive CAD and 4.4% met the endpoint. In the CAD Validation cohort, discrimination of obstructive CAD was similar between the ExECG-CL and RF-CL models: area under the receiver-operating characteristic curves 83.1% (95% confidence intervals (CI) 77.5-88.7) versus 80.7% (95%CI 74.6-86.8), p=0.14. By the ExECG-CL model, more patients had very-low (≤5%) clinical likelihood of obstructive CAD compared to the RF-CL (42.2% vs. 36.0%, p<0.01) where obstructive CAD prevalence and event risk remained low.
Conclusions. ExECG incorporated into a clinical likelihood model improves re-classification of patients to a very-low clinical likelihood group with very-low prevalence of obstructive CAD and favorable prognosis.