Belief is only half the truth - or why screening for heterophilic antibody interference in certain assays makes double sense