Interpreting a survival curve II