Dr. Lantz's comments on the utility of Holter monitoring raise several issues that deserve comment. Compared with our study, his observations showed a higher rate of clinically significant arrhythmias, probably due to the different populations studied (for example, persons with documented arrhythmias) or perhaps to different definitions of significant arrhythmias (length of consecutive beats and type of arrhythmia).

Both studies found that Holter monitoring can lead to confusion in the diagnosis and treatment of patients with cardiac arrhythmias. In our study, many of the arrhythmias recorded by Holter monitoring during symptoms were nonspecific and would not require specific therapy. In Dr. Lantz's report, half of the arrhythmias were asymptomatic, and most did not lead to changes in therapy. In contrast, our study showed that for the diagnosis of palpitations, event recorders provided data of better quality than did Holter monitoring. Any documented arrhythmias were specifically related to symptoms, and many could be modified by specific therapy directed at improving symptoms.

Dr. Lantz also raises the question of whether any form of cardiac monitoring leads to clinical management that improves symptoms, quality of life, or survival. Few randomized, controlled trials have addressed this issue, but considering only changes in specific therapy as an end point may underestimate the value of providing a patient with a diagnosis. Two of six patients with supraventricular tachycardia enrolled in our study declined an electrophysiology test or specific therapy but were happy to know that a physical cause accounted for their symptoms and that they did not have a psychological illness. In these patients, the diagnosis was the therapy, and this benefit may not be easily obtained in a retrospective review.

Regarding the ordering of other diagnostic tests in the investigation of possible cardiac arrhythmias, we agree with the principle of directing tests to certain clinical features of the patient, but this is a difficult and poorly investigated area in which few properly conducted trials have evaluated which diagnostic tests provide the best diagnostic yield (let alone improvement) in other end points [1]. Although the role of Holter monitoring in other situations is being debated, Holter monitoring is inferior to event recorders in the diagnosis of palpitations. Dr. Lantz's data support the view that Holter monitoring is of questionable value, at least for this indication.