They did a study of intensive care patients. They found that patients characterized as "difficult" by hospital staff actually had a higher success rate (# of days in ICU and lower death rate) than non-difficult patients. They didn't assign a cause but postulated better nursing care due to demanding it or more "will to live."