On a quick skim sounds interesting (passes standard first things to check: viral count was not the only endpoint that improved (but I don't know how many they measured), studies had small hundreds of people in them (I don't know anything about demographics though)).