Single arm, open label, high dropout. That pretty much tells us how much trust we can put in this study. Lancet apparently continues its tradition of publishing great sounding studies of low scientific quality.