Sedation choices and mortality: a well-defined tandem?