901 DIAGNOSTIC STRATEGIES FOR THE ASSESSMENT OF SUSPECTED STABLE CORONARY ARTERY DISEASE: A SYSTEMATIC REVIEW AND NETWORK META-ANALYSIS

Abstract

Abstract Background It is unclear which diagnostic strategy for detecting coronary artery disease (CAD) yields better outcomes. We aimed to compare the clinical accuracy and efficacy of non-invasive and invasive diagnostic strategies for the initial assessment of patients with suspected stable CAD. Methods On March 16, 2022, we searched PubMed, Embase, and CENTRAL databases for randomised controlled trials comparing diagnostic strategies for CAD detection among patients with symptoms suggestive of stable CAD. Diagnostic modalities included coronary computed tomographic angiography (CCTA), cardiovascular magnetic resonance (CMR), exercise electrocardiography, invasive coronary angiography (ICA), single photon emission computed tomography-myocardial perfusion imaging (SPECT-MPI), and stress echocardiography. Functional tests were grouped into a single node for the primary analyses, while being analysed as separate nodes for the secondary analyses. Frequentist random-effect network meta-analyses were conducted to summarise the evidence. GRADE frameworks were applied to rate the certainty of findings. The primary efficacy outcome was trial-defined major adverse cardiovascular events (MACE) and the primary accuracy outcome was the rate of unnecessary angiography. The study was registered with PROSPERO (CRD42022329635). Results Twenty trials (n=27753 participants) were included. Compared with direct-ICA referral, CCTA and functional testing provided no significant difference in the risk of MACE (incidence rate ratios [IRRs] 0·87 [95% CI 0·65-1·17] and 1·19 [95% CI 0·87-1·62], respectively; moderate certainty) and a large reduction in the rate of unnecessary angiography (odds ratios 0·04 [95% CI 0·02-0·09] and 0·08 [95% CI 0·03-0·18], respectively; high certainty). Among non-invasive diagnostic strategies, CCTA significantly reduced the risk of MACE compared with functional testing (IRR 0·73, 95% CI 0·57-0·95; high certainty). The worse prognostic performance of functional testing was mainly driven by the poor efficacy of exercise electrocardiography (high certainty) and stress echocardiography (moderate to low certainty), whose risk of MACE was significantly reduced by CCTA (IRRs 0·56 [95% CI 0·42-0·75] and 0·59 [95% CI 0·42-0·81], respectively), SPECT-MPI (IRRs 0·56 [95% CI 0·41-0·77] and 0·59 [95% CI 0·43-0·81], respectively), and direct-ICA referral (IRRs 0·64 [95% CI 0·45-0·91] and 0·67 [95% CI 0·48-0·92], respectively), while non-significantly reduced by CMR (IRRs 0·71 [95% CI 0·47-1·06] and 0·74 [95% CI 0·53-1·04], respectively). Conclusion In patients with suspected stable CAD, an initial assessment with non-invasive diagnostic strategies is safe and addresses the low diagnostic yield of direct-ICA referral. CCTA is among the most effective strategies in terms of both clinical accuracy and efficacy. Among functional non-invasive tests, exercise electrocardiography and stress echocardiography may not be deemed effective diagnostic alternatives, while SPECT-MPI and CMR are valuable and feasible options, respectively.