Purpose Recommendations for interpretation of the athlete’s electrocardiogram (ECG) has evolved in recent years, namely with the European Society of Cardiology (ESC, 2010), Seattle (2013) and more recently, the Refined (2014) criteria. A principle aim of these developments has been to improve the sensitivity and specificity for identification of young athletes at risk of sudden cardiac death and reduce false-positive rates. Whether these criteria are effective in adolescents is yet to be validated. Our study aimed to compare these criteria in a large cohort of adolescent (aged 14–17 years) athletes.

Methods Between July 1998–August 2012, 7993 adolescent, professional male footballers aged between 14 and 17, were assessed through pre-participation screening. This included a history, examination, 12-lead ECG and echocardiogram. A retrospective analysis of the ECG data was undertaken with the application of the ESC, Seattle and Refined criteria.

Results Athletes were aged 16.1 ± 1.5 years and were all male. 87.8% were Caucasian, 10.3% were Afro-Caribbean and the remainder being Oriental and South-East Asians. 1017 (12.7%) subjects were deemed to have an abnormal ECG when the ESC criteria were applied. This compared with 312 (3.9%) athletes against the Seattle criteria (p < 0.0001). The Refined criteria reduced this further to 155 (1.9%, p < 0.0001). By comparison to the ESC criteria, the Refined criteria significantly reduced the rate of ECGs deemed abnormal by 85% (p = 0.0001). By comparison to the Seattle criteria, the Refined criteria significantly reduced the rate of ECGs deemed abnormal by 50% (p = 0.0001). Rates of abnormal ECGs differed by criteria predominantly due to variations in T wave inversion criteria, and cut offs for QT intervals and right bundle branch block.

Conclusions/implications Recent developments of recommendations for interpretation of the athlete’s ECG, most significantly through the Refined criteria, shows a reduction in the number of adolescent athletes deemed to have cardiac abnormalities. Our results suggest that current screening criteria are applicable in adolescent athletes as well as older counterparts. The Refined criteria will further improve the specificity of screening and therefore reduce the burden posed by high false-positive rates.