Abstract

Since the introduction of the string galvanometer by Willem Einthoven more than a century ago, the ECG has become the most commonly performed cardiovascular test and an essential diagnostic tool in clinical cardiology.1–3 The transition from analogue to digital ECGs resulted in automated computer analysis of the ECG (ECG-C) assuming a larger role in the diagnostic interpretation.3 Yet, as the ECG-C has evolved into a necessary tool of modern medical practice, many physicians remain unaware of the hazards of relying on these preliminary diagnostic interpretations. The ECG-C clinical use should always be with the understanding that they are preliminary interpretations and require reading and confirmation by qualified electrocardiographer.3 With an increased reliance on these readings for point-of-care decision making, clinicians must remain mindful of many limitations of the ECG-C.2,3 Article see p 76 Multiple prior investigations on the ECG-C interpretation have demonstrated erroneous interpretation, resulting in unnecessary, potentially harmful medical treatment, and inappropriate use of medical resources.4–7 A large international study compared the performance of 9 ECG-C programs with that of cardiologists in interpreting ECGs in clinically validated cases of various cardiac disorders.4 The percentage of ECGs correctly classified by the ECG-C was lower than that for the cardiologists.4 The median sensitivity of the computer programs was also significantly lower than that of the cardiologists in diagnosing left ventricular hypertrophy, right ventricular hypertrophy, anterior myocardial infarction, and inferior myocardial infarction.4 The median total accuracy level (the percentage of correct classifications) was 6.6% lower for the computer programs (69.7%) than for the cardiologists (76.3%; P <0.001).4 More contemporary analysis of the accuracy of ECG-C similarly concluded that there are frequent errors in the interpretation of the cardiac rhythm. In evaluating the ECG-C interpretation of the cardiac …

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call