Background: There is uncertainty about which diagnostic strategy for detecting coronary artery disease (CAD) provides better outcomes.
Purpose: To compare the effect on clinical management and subsequent health effects of alternative diagnostic strategies for the initial assessment of suspected stable CAD.
Data sources: PubMed, Embase, and Cochrane Central Register of Controlled Trials.
Study selection: Randomized clinical trials comparing diagnostic strategies for CAD detection among patients with symptoms suggestive of stable CAD.
Data extraction: Three investigators independently extracted study data.
Data synthesis: The strongest available evidence was for 3 of the 6 comparisons: coronary computed tomography angiography (CCTA) versus invasive coronary angiography (ICA) (4 trials), CCTA versus exercise electrocardiography (ECG) (2 trials), and CCTA versus stress single-photon emission computed tomography myocardial perfusion imaging (SPECT-MPI) (5 trials). Compared with direct ICA referral, CCTA was associated with no difference in cardiovascular death and myocardial infarction (relative risk [RR], 0.84 [95% CI, 0.52 to 1.35]; low certainty) but less index ICA (RR, 0.23 [CI, 0.22 to 0.25]; high certainty) and index revascularization (RR, 0.71 [CI, 0.63 to 0.80]; moderate certainty). Moreover, CCTA was associated with a reduction in cardiovascular death and myocardial infarction compared with exercise ECG (RR, 0.66 [CI, 0.44 to 0.99]; moderate certainty) and SPECT-MPI (RR, 0.64 [CI, 0.45 to 0.90]; high certainty). However, CCTA was associated with more index revascularization (RR, 1.78 [CI, 1.33 to 2.38]; moderate certainty) but less downstream testing (RR, 0.56 [CI, 0.45 to 0.71]; very low certainty) than exercise ECG. Low-certainty evidence compared SPECT-MPI versus exercise ECG (2 trials), SPECT-MPI versus stress cardiovascular magnetic resonance imaging (1 trial), and stress echocardiography versus exercise ECG (1 trial).
Limitation: Most comparisons primarily rely on a single study, many studies were underpowered to detect potential differences in direct health outcomes, and individual patient data were lacking.
Conclusion: For the initial assessment of patients with suspected stable CAD, CCTA was associated with similar health effects to direct ICA referral, and with a health benefit compared with exercise ECG and SPECT-MPI. Further research is needed to better assess the relative performance of each diagnostic strategy.
Primary funding source: None. (PROSPERO: CRD42022329635).