While one may be able to quibble with their overall methods and statistics, my primary goal will be to further the cause for more studies dealing with end-of-life care. As the authors mentioned, the small numbers in their study make extrapolation of these data somewhat tenuous. However, while it is not a definable problem with this specific article, I believe that we need a major shift of emphasis in articles such as this. This study was conducted at a university hospital that had attending physicians, fellows, other house staff, medical students, and an established palliative care team. Since most patients in this country who are critically ill are actually cared for in a different environment, that is, in the ICU of a community hospital, there should be a means of pairing studies such as this with a community hospital. This would have a tremendous potential benefit. It could serve to break down the town-and-gown mentality, it could raise the level of care and interest in community hospitals, and any meaningful data or conclusions from such studies would be more generalizable to the medical community and therefore to more patients. In this way, we would be able to more immediately and effectively use these data rather than waiting for the study that applies to one’s particular environment. Combining academic resources with an increasing number of excellent, resourceful, and progressive community hospitals could be a practical approach. It has been thought that a teaching center presence in a community may raise the standard of medical care for that community.