Inter-observer agreement in athletes ECG interpretation using the recent international recommendations for ECG interpretation in athletes among observers with different levels of expertise.

INTRODUCTION International criteria for the interpretation of the athlete's electrocardiogram (ECG) have been proposed. We aimed to evaluate the inter-observer agreement among observers with different levels of expertise. METHODS Consecutive ECGs of Swiss elite athletes (≥14 years), recorded during routine pre-participation screening between 2013 and 2016 at the Swiss Federal Institute of Sports were analysed. A medical student (A), a cardiology fellow (B) and an electrophysiologist (C) interpreted the ECG's independently according to the most recent criteria. The frequencies and percentages for each observer were calculated. An inter-observer reliability analysis using Cohen Kappa (κ) statistics was used to determine consistency among observers. RESULTS A total of 287 ECGs (64.1% males) were analysed. Mean age of the athletes was 20.4±4.9 years. The prevalence of abnormal ECG findings was 1.4%. Both, normal and borderline findings in athletes showed moderate to good agreement between all observers. κ scores for abnormal findings resulted in excellent agreement (κ 0.855 in observer A vs C and B vs C to κ 1.000 in observer A vs B). Overall agreement ranged from moderate (κ 0.539; 0.419-0.685 95% CI) between observer B vs C to good agreement (κ 0.720; 0.681-0.821 95% CI) between observer A vs B. CONCLUSIONS Our cohort of elite athletes had a low prevalence of abnormal ECGs. Agreement in abnormal ECG findings with the use of the recently published International recommendations for ECG interpretation in athletes among observers with different levels of expertise was excellent. ECG interpretation resulted in moderate to good overall agreement.