Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

© 2015 Elsevier Ireland Ltd. All rights reserved. Background: Screening for atrial fibrillation (AF) using 12-lead-electrocardiograms (ECGs) has been recommended; however, the best method for interpreting ECGs to diagnose AF is not known. We compared accuracy of methods for diagnosing AF from ECGs. Methods: We searched MEDLINE, EMBASE, CINAHL and LILACS until March 24, 2014. Two reviewers identified eligible studies, extracted data and appraised quality using the QUADAS-2 instrument. Meta-analysis, using the bivariate hierarchical random effects method, determined average operating points for sensitivities, specificities, positive and negative likelihood ratios (PLR, NLR) and enabled construction of Summary Receiver Operating Characteristic (SROC) plots. Results: 10 studies investigated 16 methods for interpreting ECGs (n = 55,376 participant ECGs). The sensitivity and specificity of automated software (8 studies; 9 methods) were 0.89 (95% C.I. 0.82-0.93) and 0.99 (95% C.I. 0.99-0.99), respectively; PLR 96.6 (95% C.I. 64.2-145.6); NLR 0.11 (95% C.I. 0.07-0.18). Indirect comparisons with software found healthcare professionals (5 studies; 7 methods) had similar sensitivity for diagnosing AF but lower specificity [sensitivity 0.92 (95% C.I. 0.81-0.97), specificity 0.93 (95% C.I. 0.76-0.98), PLR 13.9 (95% C.I. 3.5-55.3), NLR 0.09 (95% C.I. 0.03-0.22)]. Sub-group analyses of primary care professionals found greater specificity for GPs than nurses [GPs: sensitivity 0.91 (95% C.I. 0.68-1.00); specificity 0.96 (95% C.I. 0.89-1.00). Nurses: sensitivity 0.88 (95% C.I. 0.63-1.00); specificity 0.85 (95% C.I. 0.83-0.87)]. Conclusions: Automated ECG-interpreting software most accurately excluded AF, although its ability to diagnose this was similar to all healthcare professionals. Within primary care, the specificity of AF diagnosis from ECG was greater for GPs than nurses.

Original publication

DOI

10.1016/j.ijcard.2015.02.014

Type

Journal article

Journal

International Journal of Cardiology

Publication Date

01/01/2015

Volume

184

Pages

175 - 183