Dr. Franklin said that inclusion of DNR status or withdrawal of care in a data model could be problematic unless the terms are clearly defined beforehand. He noted, for example, that a DNR order may not be written in the chart for some patients, but that it is clear physicians are writing for comfort care measures when they extubate a patient, increase their pain medications, and move them out of a monitored setting. Although the potential exists to game the system and data entry needs to be reinforced, the data system itself is okay, according to Dr. Franklin.
Finally, it was noted by members of the audience that the medical community has not done a good job of educating the public on such issues as withdrawal of care, DNR orders, and palliative care, and that these issues will become more pressing as the population ages. When asked whether the researchers looked at race, Dr. Franklin said they did not, but that anecdotally, minorities were less likely to withdraw care.
Dr. Franklin and his coauthors disclosed no relevant conflicts of interest.