LAUSR

BACKGROUND Accurate electrocardiogram (ECG) interpretation in competitive athletes requires the distinction of physiological adaptations from findings suggestive of a pathological condition. The purpose of this study was to compare the performance of the Seattle Criteria to the International Criteria in a large dataset of NCAA athletes screened with an ECG. METHODS ECGs from 5258 NCAA athletes who underwent prior ECG screening were re-examined by two experts in the field of sports cardiology and by ECG interpretation software by Cardea (© 2018 Cardiac Insight Inc.) using the Seattle and International Criteria. Each ECG was classified as normal or abnormal and the specific ECG abnormalities noted. Chi-squared analysis was used for statistical comparisons. RESULTS The total number of ECGs flagged as abnormal by expert over-read decreased from 158 (3.0%) using the Seattle Criteria to 83 (1.6%) using the International Criteria (p<0.0001). Likewise, the total number of abnormal ECGs using ECG interpretation software by Cardea decreased from 278 (5.3%) using Seattle Criteria programming compared to 134 (2.5%) using International Criteria programming (p<0.0001). The most common ECG abnormality by expert over-read using the International Criteria was T wave inversion 40 (48%). The newer definition of pathological Q waves reduced the number of ECGs flagged as abnormal from pathologic Q waves from 69 (Seattle) to 11 (International) (84% reduction; p<0.0001). Expert over-read using both criteria and both Cardea interpretation programs identified all 13 athletes with cardiac pathology associated with sudden cardiac death. Cardea software using the International Criteria had a higher false-positive rate (2.3%) than expert over-read (1.3%) (p=0.0001). CONCLUSIONS Use of the International Criteria for ECG interpretation significantly reduces the total abnormal and false-positive ECG rates compared to the Seattle Criteria without compromising sensitivity. Cardea interpretation software performs well and may be a useful tool to assist clinicians.