The Electrocardiogram in Epidemiological Studies: Reproducibility, Validity, and International Comparison

Abstract

The means available for identifying coronary disease in the community are regrettably limited.A history of chest pain may be elicited and judicious questions may enable one to decide if this was cardiac or not.If cardiac pain has been experienced, clinical examination may reveal conditions other than coronary disease (congenital, valvular, or hypertensive disease, for example) which could be responsible.Finally an electrocardiogram (ECG) may be taken.If the resting ECG is normal, evidence suggestive of coronary disease may be obtained in a proportion of cases by repeating it under some form of stress, usually exercise.Unfortu- nately abnormal electrical patterns shown by the ECG, even when they are highly suggestive, are not specific for coronary disease.Furthermore, there may be considerable difficulty in differentiating normal from abnormal tracings (Simonson, 1958).It is not surprising, therefore, that a number of studies have revealed variation between different observers in their classification ofECGs as normal or abnormal and especially of whether the abnormality is due to coronary disease or not (Davies, 1958; Thomas,