Limitations of exercise treadmill testing in evaluating patients for the presence of atherosclerosis

Background As subclinical atherosclerosis is increasingly recognized as a key predictor of cardiovascular events, there is growing interest in cost-effective and reliable screening tools for its detection. While exercise treadmill testing (ETT) is commonly used to assess ischemia, its sensitivity and specificity in detecting coronary artery disease (CAD) remain uncertain. Purpose This study compares ETT results with coronary CT angiography (CCTA) to evaluate its diagnostic accuracy in detecting both obstructive CAD and any presence of atherosclerosis, with implications for risk stratification. Methods A retrospective analysis was conducted on 1,326 patients without known CAD who underwent both ETT and CCTA. ETT results (normal, abnormal, or equivocal) were compared to CAD-RADS classifications. Two models were evaluated: the first assessed ETT’s ability to detect obstructive CAD (CADRADS ≥3), and the second assessed its ability to detect any atherosclerosis (CADRADS 1-5). Sensitivity, specificity, PPV, and NPV were calculated for both models. Results For detecting obstructive CAD, ETT had a sensitivity of 36%, specificity of 76%, PPV of 17%, and NPV of 89%. Disease (CADRADS ≥3) was correctly identified in 4% of patients, while 9% were missed. 67% of patients were appropriately categorized as not having obstructive disease, whereas 24% underwent unnecessary repeat testing despite having a CAD-RADS of 0-2. When evaluating ETT’s ability to detect any atherosclerosis, sensitivity was 27%, specificity 76%, PPV 61%, and NPV 43%. Atherosclerosis was correctly identified in 16% of patients, while 8% of patients with obstructive disease (CAD-RADS ≥3) and 35% of those with early atherosclerosis (CAD-RADS 1-2) were not detected. Appropriate reassurance was provided in 32% of cases, while 10% of patients underwent unnecessary repeat testing despite having no detectable disease (CAD-RADS 0). Conclusion ETT demonstrates limited accuracy in detecting both obstructive and non-obstructive CAD, with almost a third of patients misclassified in both models, leading to missed diagnoses and unnecessary testing. These findings highlight the limitations of ETT as a screening tool and suggest that coronary imaging may offer superior risk stratification by improving early atherosclerosis detection and reducing unnecessary testing.