Discordant Results of Exercise Echo and ECG

Abstract & Commentary

By Michael H. Crawford, MD

Source: Al-Mallah M, et al. Long term favorable prognostic value of negative treadmill echocardiogram in the setting of abnormal treadmill electrocardiogram: a 95 month median duration follow-up study. J Am Soc Echocardiogr. 2008;21: 1018-1022.

Since both treadmill exercise ECG and echo have independent prognostic value when studied alone, what does the ECG response add to a negative exercise echo? To answer this question Al-Mallah et al from Henry Ford Hospital in Detroit evaluated consecutive Bruce protocol exercise echo studies and separated out 677 patients who had a negative echo study. They excluded those with uninterruptable ECGs. A positive or negative ECG response indicative of ischemia further separated the patients (598 negative, 79 positive). Patients were followed for a median of 95 months to identify major adverse cardiac events (MACE): death, myocardial infarction, or revascularization.

Results: MACE occurred in 8.6% of the patients with negative exercise echoes, with an annual event rate of 1%. Although the raw data exhibited more events in the ischemic ECG group (15 vs 8%, p = 0.025), when the data were adjusted for clinical and stress test variables, an ischemic ECG response was not independently predictive of MACE (p = 0.2). Also, for the first five years of follow-up, there was no difference between the ECG groups in the raw MACE. Independent predictors of MACE were female sex, prior coronary heart disease, exercise duration, and chest pain at peak exercise. Al Mallah et al concluded that patients with negative exercise echo studies have an excellent long-term prognosis that is not influenced by the exercise ECG results.


Not infrequently I will be called to assess an exercise echo on a patient of mine who had a positive ECG response. Often the echo is negative, and I reassure the patient, but there is always a lingering doubt concerning whether the ECG response is correct. This study relieves some of that anxiety, since patients with a negative exercise echo had an excellent prognosis, especially for the first five years. This suggests that the warranty on a negative exercise echo is about five years. This study is consistent with the results of previous studies with shorter follow-up periods. However, the study does show that you cannot look at the echo results in isolation. The raw data showed that a positive ECG response was predictive of MACE, but that it was explained by other clinical features such as female sex, low exercise duration, angina at peak exercise, and patient factors such as known coronary heart disease. Thus, the whole patient and all the features of the exercise response need to be factored into your clinical assessment.

There are several limitations to this study which may have biased the results. They included patients with known disease which probably increased the event rate. It was a retrospective single-center study, and charts were used for follow-up. The exercise studies were done before the use of contrast and harmonic imaging at their center; today's echo sensitivity for wall motion may be better. They do not discriminate between ECG leads. It is know that the specificity of the inferior leads is poor. About 80% of the patients achieved 85% maximal predicted heart rate, so 20% had inadequate stress tests. However, this is a consecutive series of real-world patients referred for stress testing and as such represents a valuable observation.