The present technology is an automated method for determining whether a patient-specific electrocardiogram (ECG) is either (a) Normal and can be excluded from manual review or (b) Abnormal and included for manual review. In one embodiment, the method comprises comparing a plurality of characteristics of the ECG with predetermined subthreshold levels that are set less than clinically significant levels of abnormality of the characteristics, wherein the characteristics of the ECG are selected from the group including T-Wave inversion, ST-Depression, QT segment duration, delta wave character, anterior S-wave character and ectopic or pre-mature beats. The method continues by selecting the ECG for manual review if the plurality of selected characteristics exceed the predetermined subthreshold levels yet are below the corresponding clinically significant threshold levels of abnormality of the characteristics.