A bias for action in cancer screening?