A meta-regression analysis of treatment effect modifiers in trials with flexible-dose oral sildenafil for erectile dysfunction in broad-spectrum populations.

Several trials and reviews provide evidence for the efficacy of phosphodiesterase-5 inhibitors in the treatment of erectile dysfunction (ED). However, little is known about the impact of treatment effect modifiers other than concomitant diseases. Our objective was to identify patient and trial characteristics as well as methodological and publication-related issues that are associated with the treatment effect measured in flexible-dose randomized controlled trials of oral sildenafil for ED. The MEDLINE and the Cochrane Central databases were searched for efficacy trials of sildenafil. Thirteen trials fulfilled all inclusion criteria. A series of meta-regression and graphical analyses were performed to test the impact of possible effect modifiers. Treatment effect was influenced by mean baseline disease severity and mean duration of the disease. These associations were at least partly mediated by placebo response. Trial duration, age of patients and etiology of ED in patients did not have any significant influence on the treatment effect. The year of publication of primary trials was also related to trial findings. Our analysis adds important data to enable the control of confounding variables in future trials and meta-analyses. It might also help the individual to assess the unbiased efficacy and true innovative potential of available and forthcoming pharmacological agents.