I n this issue of Circulation: Arrhythmia and Electrophysiology, Sanatani et al 1 report a randomized, double-blind study of a comparison of digoxin and propranolol for chronic therapy of supraventricular tachycardia (SVT) in infants <4 months of age. The key findings in 61 patients were that there were no differences in SVT recurrence for the 2 drug treatment groups and that there were no recurrences of SVT after 4 months of therapy with either antiarrhythmic agent. Another salient point to be taken includes an expectation that perhaps 10% of infants diagnosed with SVT can be expected to require subsequent hospitalization for management of SVT recurrences.
Article see p 984As is often the case with even thoughtful attempts to perform these types of clinical trials in pediatric electrophysiology, and pediatric cardiology generally, the study is underpowered to make feasible many of the desired statistical analyses between the 2 treatment groups. As a result of the relatively small numbers of patients in a given institution, multicenter studies in our field are frequently required to achieve the requisite N for these study designs. As the authors note, issues of patient enrollment, physician biases, parental consent, and inadequate follow-up plague this and other similar efforts. The history of this specific article is telling, in its attempts to clear the hurdle of peer review: it was submitted to 4 journals, critiqued by 12 reviewers, and underwent 8 revisions in response to seemingly unachievable demands for the statistical rigor of a large, randomized clinical trial. The study and its lead author became fixtures for a plea to the masses for patient enrollment at the Pediatric and Congenital Electrophysiology Society business meeting at Heart Rhythm Society for each of the past 6 years. With its key features intact, it has reached its well-deserved publication.From a practical standpoint, yes, the study is underpowered. It is still worthy of our attention and publication, even in its descriptive terms, for its effects on what we do, how we educate families, and how we lay the foundation of their expectations going forward. Personally, over the years, I have scaled back the duration of therapy for infants with SVT from 12 months to 8 months and then to 6 months. As a result of this current report, I have been stopping therapy for SVT in these non-Wolff-Parkinson-White, common SVT mechanism patients at 4 months. The data from Sanatani et al 1 have been available in review form for the better part of a year, and (anecdotally!) over that time period we have not seen an SVT recurrence. That aspect of the current article raises another value of this data: it provides us with a data set from which we can pose new questions, such as "do all infants with SVT warrant ongoing treatment, and if not, which patients?" As the authors query, do we need (or want) a placebo study for this population?This all brings us to an important consideration for the design of studies for rare (ie, low N) conditions. A valid critique of ra...