Do-not-resuscitate (DNR) orders rule out cardiopulmonary resuscitation in patients who lack a perceptible pulse and become clinically unresponsive; nonetheless, patients with DNR orders continue to receive standard medical treatment. An estimated 15 percent undergo some form of surgery, either to treat a condition unrelated to their illness (for example, a fracture), to reduce pain, or simply to attempt to gain some extra time. Up to now, however, studies of the outcomes of various surgical procedures have not focused on this particular group of patients.

A trio of researchers led by Julie Ann Sosa, M.D., associate professor of surgery and medicine, decided to address this gap in knowledge. “We undertook this study because of our own experience as surgeons being summoned in the middle of the night to the bedside of a patient with a terminal illness facing a surgical crisis,” Sosa said. “Trying to have a life-and-death discussion without being able to provide real data is daunting.”

In a close analysis of more than 8,000 case records from hospitals throughout the United States, Sosa and her colleagues found that, in the first month after surgery, the mortality rates of DNR patients are nearly three times those of their non-DNR counterparts. Moreover, DNR patients who survived suffered a higher rate of complications, both major (such as pneumonia or severe sepsis) and minor (urinary tract infection or peripheral nerve injury), than similar patients without DNRs.

Why would patients with a DNR order fare so much worse after surgery than their non-DNR counterparts? In their study, published online on April 18 in the Archives of Surgery, Sosa and her colleagues discovered a partial answer: The DNR patients were more likely than non-DNR patients to have come from an acute or chronic care facility, to be suffering co-morbid conditions, and to show some degree of functional impairment.

Yet even when the researchers adjusted their statistical model for all these characteristics, the DNR order still stood out as an independent risk factor. “This finding suggests that besides the DNR patients being sicker to begin with, there’s something else going on that affects the outcome—something we haven’t been able to pinpoint yet,” said co-author Sanziana Roman, M.D., associate professor of surgery.

Are DNR patients somehow treated differently, or is this population different in its willingness to undergo surgery despite being in declining health? Perhaps these are individuals who wish to make one more major effort at life (the surgery) but who have also begun to face their own mortality (hence the DNR order). Then again, some may be agreeing to surgery that would normally be beneficial, but whose benefit they will probably not live long enough to see.

While the exact nature of the DNR order as a risk factor may elude researchers for some time, Sosa, Roman, and medical student Hadiza Kazaure, make one recommendation that can be put into practice right away. They urge patients and their physicians to begin a conversation about overall health goals much earlier in the course of treatment, before a medical crisis rushes the patient into surgery that might not be in his or her best interest. Said Roman, “We hope our paper could be helpful in such a conversation. We’d like surgeons, physicians, and patients to view this study as a tool that helps them to weigh any surgical options in the context of each particular patient’s overall treatment plan.”