Will the electrocardiograph replace the electrocardiographer?

Abstract

Professional groups, endorsed by individual cardiologists involved in the practice of electrocardiology, have recently issued official policy statements declaring that there is no computer program for electrocardiographic interpretation that can replace interpretation by a skilled physician, and consequently, all computer electrocardiographic interpretations require careful overreading by a physician qualified to interpret electrocardiograms (ECGs). A somewhat different point of view is presented in this study, which asserts that the arguments brought forward against automated electrocardiographic interpretation may have been biased to some degree. The key issue in these deliberations is whether or not electrocardiographic analysis can be considered a laboratory test. This review elucidates factors that have hindered more widespread acceptance of automated electrocardiographic interpretation and presents possible solutions to some of the logistic problems observed. It is concluded that with relatively minor design improvements in electrocardiographic acquisition and interpretation software, a considerable fraction of ECGs can be identified that are amenable for electrocardiographic processing as an automated laboratory test. In turn, the remaining ECGs will require more advanced electrocardiographic software for differential diagnostic classification in combination with professional review using more comprehensive supplementary information than is presently provided in most routine electrocardiographic processing applications.