Blinding in pharmacological trials: the devil is in the details