Aim: We evaluated the accuracy of multidetector computed tomography in detecting coronary artery disease and how it could change the indication to coronary angiography in patients with suspected cardiac chest pain.
Methods and results: We enrolled 142 consecutive patients who had already performed an exercise electrocardiogram test referred to our hospital and scheduled for coronary angiography for chest pain. According to the characteristics of chest pain and the results of exercise electrocardiogram, patients were divided into four groups: atypical chest pain and negative exercise electrocardiogram (group 1); typical chest pain and negative exercise electrocardiogram (group 2); atypical chest pain and positive exercise electrocardiogram (group 3); and typical chest pain with positive exercise electrocardiogram (group 4). We evaluated the accuracy of multidetector computed tomography and whether it could reduce the number of unnecessary coronary angiography in the study groups. Of 1801 segments larger than 1.5 mm, 1696 (94%) were assessable. In a segment based-model, sensitivity, specificity, negative predictive value, positive predictive value and accuracy were 81% (95% confidence interval 75-89%), 94% (95% confidence interval 90-98%), 96% (95% confidence interval 93-98%), 75% (95% confidence interval 69-82%) and 91% (95% confidence interval 89-93%), respectively. In a patient-based model, sensitivity, specificity, negative predictive value, positive predictive value and accuracy were 95% (95% confidence interval 91-99%), 78% (95% confidence interval 67-89%), 88% (95% confidence interval 79-97%), 89% (95% confidence interval 83-95%) and 89% (95% confidence interval 84-94%). Unnecessary coronary angiography may be avoided by multidetector computed tomography results particularly in group 2 (16%) and group 3 (24%), whereas in groups 1 and 4 the role of multidetector computed tomography in facilitating the correct indication to coronary angiography was less relevant.
Conclusions: Multidetector computed tomography is a particularly helpful technique in patients with discordance between the clinical features of chest pain and stress-test results. This technique may be introduced in the diagnostic work-up of patients with suspected coronary artery disease and may potentially reduce the number of unnecessary coronary angiography.