Randomised controlled trials in severe asthma: selection by phenotype or stereotype