Detection performance of an automatic lead reversal detection module

Background: Lead reversal and lead mismatch can simulate ECG abnormalities such as ectopic beats, intraventricular conduction disturbances, chamber enlargement or ventricular pre-excitation. Further simulation or concealing of myocardial ischemia or infarction might be seen on the ECG trace as well which might lead to thewrong decision of therapy. Lead reversal such as right arm (RA) and left arm (LA) usually can be detected by an experienced ECG reader or by computer algorithms. In such cases, a second ECG can be taken. Lead reversal has been reported to occur in 0.4–4% of all cases in standard 12-lead ECGs. Methods: We performed a retrospective analysis of 532 16-lead ECGs including standard 12-lead plus right precordial leads V3r and V4r as well as the posterior leads V8 and V9. The ECGs were acquired from chest pain patients at an emergency department. The ECG device did not give any feedback to the user with respect to lead reversal or lead mismatch at all before, during or after acquisition. A lead reversal detection module based on correlation analysis of 3 subsets within the acquired 16 ECG leads was investigated. The numbers of possible lead reversal in each subset were computed. Results: We found in 24 cases possible lead reversal in the standard limb leads (I, II, III, aVR, aVL, aVF), 40 in the standard chest leads (V1–V6) subset of the standard 12-lead, 99 in the subset of V3r and V4r and 43 in the V8 and V9 subset. Therefore, the amount of possible lead reversal found was 5%, 8%, 19% and 8% in the corresponding lead subset with the highest number in the right precordial leads V3r and V4r. Conclusions: Lead reversal might be much higher than estimated. Clinically accepted lead systems that are known and used since some time such as the standard 12-lead system give compared to the combinations of possible mistakes the lowest relative lead reversal number.