Is Surface ECG a Useful Surrogate for Subcutaneous ECG?

Background: Surface electrocardiograms (ECGs) have been used as surrogates for subcutaneous ECGs to optimize and evaluate subcutaneous devices, but differences between surface and subcutaneous ECGs remain poorly understood. This study evaluated the correspondence between surface and subcutaneous ECGs in Reveal (R) Plus (Medtronic Inc., Minneapolis, MN, USA) patients during various maneuvers. Methods: Surface electrodes were placed over the Reveal (R) electrodes of 48 subjects (23 men, age 60 +/- 14.3 years, body mass index 27 +/- 4.9 kg/m(2), implant time 45 +/- 29 weeks). Surface and subcutaneous ECGs were recorded simultaneously for 30 seconds during rest, isometric myopotential noise (pushing palms together), and artifact-inducing maneuvers (repetitive displacement of device, chest thumping on device, arm flaps, handshake, hallwalk). Results: During rest, subcutaneous and surface ECGs were highly correlated (R = 0.96), had similar R-wave amplitude (487 +/- 40 vs 507 +/- 49 mu V, NS), and signal-to-noise ratio (SNR) (13.4 +/- 0.8 vs 13.5 +/- 0.7, NS). During myopotential noise, subcutaneous and surface ECGs were highly correlated (R = 0.91) and had similar SNR (10.0 +/- 0.6 vs 9.7 +/- 0.6, NS). During artifact-inducing maneuvers, subcutaneous and surface ECGs were less correlated (R = 0.82 displacement, 0.84 chest thumping, 0.93 arm flaps, 0.90 handshake, 0.92 hallwalk) with subcutaneous SNR significantly higher than surface (11.4 +/- 0.7 vs 9.9 +/- 0.7 displacement, 11.1 +/- 0.6 vs 8.4 +/- 0.6 chest thumping, 11.5 +/- 0.4 vs 10.3 +/- 0.5 arm flaps, 9.5 +/- 0.4 vs 8.4 +/- 0.4 handshake, 12.0 +/- 0.4 vs 10.0 +/- 0.4 hallwalk, P < 0.05). Conclusion: Surface ECGs are adequate surrogates for subcutaneous ECGs in situations free from motion artifacts but not in situations involving movement of the device, surface electrodes, or recording equipment. During artifact-inducing maneuvers, subcutaneous ECGs are of higher quality and less susceptible to artifacts than surface ECGs. (PACE 2010; 33:135-145).