Assessing the validity and responsiveness of a generic preference quality of life measure in the context of posttraumatic stress disorder