<p>Background: The efficacy of antifungals for onychomycosis has been determined in randomized controlled trials (RCTs); interestingly their control arms have demonstrated some therapeutic effects. These controls constitute either placebos (inert pills) or vehicles (all but the antifungal component of the creams). The objective of this research was to determine (i) whether RCT controls exhibited statistically-relevant efficacy rates (i.e. beyond the "placebo effect"), (ii) whether oral and topical controls differed in their efficacies, and (iii) if the efficacy rates of the controls correlated with those of the active comparator associated with that control.</p> <p>Methods: RCTs of oral and topical monotherapies for dermatophyte toenail onychomycosis were identified through a systematic literature search. For our meta-analyses of cure rates the double arcsine transformation was used. The N-1 chi squared test was used to determine whether the cure rates significantly differed between topical and oral controls. Correlation was investigated using Kendall rank correlation tests.</p> <p>Results: The pooled mycological, complete, and clinical cure rates of all control interventions (n = 19 trials) were 9%, 1%, and 6%, respectively. The pooled efficacy rates for oral and topical controls were: mycological cure rate, 7% and 12% (p=0.0016); complete cure rate, 1% for both; and clinical cure rate, 4% and 8%, respectively (p=0.0033). For oral RCTs, the respective cure rates of the active therapies were not correlated with controls. However, for topical RCTs, as the mycological and clinical cure rates of the active therapy increased, so did those of the topical vehicle associated with the active therapy in question, and vice versa.</p> <p>Conclusions: The topical vehicle cure rates were often higher than the oral placebo cure rates, likely due to the presence of non-antifungal chemicals (e.g. moisturizers, urea) with antifungal and debriding properties, which are not present in oral controls.</p>

This content is only available as a PDF.