Best supportive care in clinical trials: review of the inconsistency in control arm design