Timing and Meaning of Do-Not-Resuscitate Orders in the Palliative Care Setting

A Conversation With Mary F. Mulcahy, MD

Get Permission

Mary F. Mulcahy, MD

Guest Editor: Jamie H. Von Roenn, MD

Having ongoing conversations with patients over the course of their cancer progression will help ensure that there is no misunderstanding about patients’ expectations for their end-of-life care.

—Mary F. Mulcahy, MD

Although a do-not-resuscitate (DNR) order does not mean “do not treat,” that is how it is often interpreted, according to a study examining the level of care oncology inpatients at a tertiary care hospital received.1 The study found that the interpretation of DNR orders among oncology nurses and especially among physicians is variable and results in less aggressive care in general, and that DNR-status patients are more likely to die compared to full code–status patients, regardless of their disease severity. Other studies also show DNR orders influence care more broadly than perhaps patients intended.2,3

Having ongoing conversations with patients from the time of their initial diagnosis and throughout their illness about their likely cancer trajectory, treatment options including benefits and alternatives, and their goals of care are crucial to ensuring that patients’ treatment wishes are followed during a medical emergency.

“If patients have a very clear feeling about what they would or would not want at the end of their life and understand they have a terminal illness and they do not wish to be resuscitated at the time of their death, then a DNR order is appropriate,” said Mary F. Mulcahy, MD, Associate Professor in Medicine, Hematology/Oncology at Northwestern University Feinberg School of Medicine in Chicago, and cofounder of Life Matters Media. “But a lot of patients are afraid of signing a DNR order because they think their treatment is going to be compromised or withheld.”

A more comprehensive directive than a DNR order is the Physician Orders for Life-Sustaining Treatment (POLST) paradigm (, which augments traditional methods for advance care planning and treatment preferences, including those for cardiopulmonary resuscitation (CPR), artificial nutrition by feeding tube, and antibiotic use, according to Dr. Mulcahy.

The ASCO Post talked with Dr. Mulcahy about the timing of DNR and POLST orders, the difficult end-of-life medical decisions for physicians even when a DNR order is in place, and how devices like pacemakers and implantable cardioverter-defibrillators are impacting end-of-life care.

Clarifying DNR

At what point in a patient’s care should DNR orders be applied?

Once a patient is diagnosed with an incurable cancer, regardless of how long that person may live, having an advance directive stating the patient’s wishes about the extent of medical care at the end of life is critical. The most important thing is that the directive be a living document and that there is ongoing conversation between the patient and the oncologist as the status of the cancer and the patient’s health changes.

Having a do-not-resuscitate order only applies at the time of the patient’s death. It is a directive for when the patient has stopped breathing and the patient’s heart has stopped beating. If a patient has lung cancer and develops pneumonia, what should the oncologist do if the patient has a DNR order in place—not treat the pneumonia? Well, maybe, or maybe not. A study investigating the preferences for other treatments among patients with DNR orders and those with attempt CPR orders found that half of the patients with DNR orders wanted comfort measures only and half wanted a higher level of treatment at a hospital.4

The POLST paradigm tries to address this middle ground, which is the hardest part to interpret. Just because patients have a DNR order in place, physicians cannot assume that they do not want medical intervention. We are asking patients if they want to be resuscitated at the point of death, but we are not asking terminally ill patients with no further treatment options, “What would you like us to do in the event you develop pneumonia; would you want to go to the hospital for treatment?” This is the gap in conversation that POLST addresses.

I prefer the POLST paradigm approach to end-of-life medical planning because it includes two possible orders for resuscitation: do not attempt CPR (DNR) or attempt CPR, so it gives the patient more options. I had a patient—a physician—with metastatic colon cancer, and when she handed me her advance care directive, she said to do everything regardless of cost and chances of recovery, because she had seen firsthand that medical staff back off if a patient has a DNR order, and she was afraid that she wouldn’t be treated for a problem that was potentially treatable. That experience really hit home with me and made me appreciate what patients feel when they sign DNR orders.

Some patients are very clear about the care they want at the end of life, but they are in the minority. Most patients do not like to commit to a DNR order in the event a medical issue arises that is potentially fixable.

Ongoing Conversations

How can oncologists help patients in their end-of-life care decision-making?

More important than just checking off a DNR or resuscitate box on a form is understanding what someone’s goals are for care. That comes from having conversations with patients that outline those goals and may include possible medical scenarios as their cancer progresses and their health declines.

Some patients absolutely know that they never want to be in a hospital again, never want to be transferred to an intensive care unit, and don’t want treatment for extenuating circumstances beyond their cancer. Other patients may want treatment for anything that is potentially treatable. Having ongoing conversations with patients over the course of their cancer progression will help ensure that there is no misunderstanding about patients’ expectations for their end-of-life care.

Complicating Factors

Please describe how DNR orders can inadvertently complicate medical care.

The POLST form clearly states the two possible orders for resuscitation, which we have discussed. There is also a section that addresses the aggressiveness of care at the end of life, which is the harder part in the decision-making process.

When a patient has stopped breathing and his heart has stopped beating, not resuscitating that patient is a clear-cut decision. What is more difficult is the case in which a patient has impending health decline and there is a question, for example, of whether to intubate because of a health complication like pneumonia. We tend to put all the attention on the DNR order, but we need more attention focused on impending health catastrophes.

One morning I was on service and had a conversation with a patient with advanced breast cancer that was progressing rapidly. She was not my patient, but she wanted to fill out a DNR order, and she did. That evening, the patient went to the radiology department for a brain scan and had a seizure. The rapid-response team called to ask me if she should be intubated.

I was really unsure about what to do. I was the one who had had the conversation with the patient about her DNR order. I had even helped her fill out the form and I signed the form, and I didn’t know what to do. I finally decided to intubate her because I wasn’t sure if she had had a chance to talk to her family about her decision. The patient never recovered, but at least her family had time to process what was happening to her and to learn about the conversation I had had with her that morning, and I think it helped them come to terms with her death.

Legal Issues

Were you concerned about potential legal jeopardy?

I wasn’t worried at the time, but when I think about this case and all the legal avenues it could have taken, I am troubled. I have questioned myself over and over again about whether I made the right decision, and I’ve concluded that I did. But the end result certainly could have gone another way.

The family could have faulted me for going against the patient’s wishes not to be resuscitated, and I could have gotten into trouble. But these types of cases are always going to present themselves, and it is always going to be a struggle knowing patients’ true values and goals, and what is important to them.

That is why advance care planning is an ongoing process and not simply a DNR box to be checked off on a form. It is a process in which the patient accepts his or her illness and whereby the patient, family members, and medical team work together to do the right thing for the patient.

Sometimes patients want us to tell them what to do, and doctors can’t be afraid to do that, because these are hard decisions to make. When I meet patients for the first time, I try to get a sense of their character and what they want because I want to guide them in the right direction.

Some patients are never going to accept that they are going to die from their cancer and will always hold onto hope that there is something more that can be done. As oncologists, we have to change that unrealistic hope [for futile care] to hope for comfort and peace at the end of life.

Implantable Devices

How are implantable devices like pacemakers and defibrillators impacting oncology care in the palliative setting?

They are definitely complicating care. When a patient has a terminal illness, he or she will eventually go into an arrhythmia, and that implantable device is going to fire and keep the heart beating. We have to tell these patients that there is going to come a time when it will be appropriate to shut off the pacemaker and let the heart beat naturally, but that is a hard discussion to have. When is that appropriate time to turn off these devices? Is it as the person is dying, because the heart is beating and maybe the patient is still reaping a benefit? I don’t always have the answer to that one, and this is where we need input and guidance from the patient’s cardiologist.

Part of what we are doing at Life Matters Media ( is providing a platform where people can share information and resources to help others participate in the decision-making for their advance care directive and presenting end-of-life issues in a narrative form to start this difficult dialogue. Our goal is to ensure that patients have the knowledge they need to make the end-of-life choices that are important to them, so they can die in dignity and peace. ■

Disclosure: Dr. Mulcahy reported no potential conflicts of interest.


1. Lund CR, Samant R: Interpretation of do-not-resuscitate (DNR) orders among oncology health care professionals and the potential for these orders to influence care for medical issues unrelated to resuscitation. J Clin Oncol 28(15 suppl):Abstract 9078, 2010.

2. President’s Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research: Deciding to forgo life-sustaining treatment, p 251. Washington, DC; U.S. Government Printing Office; 1983.

3. Beach MC, Morrison RS: The effect of do-not-resuscitate orders on physician decision-making. J Am Geriatr Soc 50:2057-2061, 2002.

4. Fromme EK, Zive D, Schidt TA, et al: POLST registry do-not-resuscitate orders and other patient treatment preferences. JAMA 307:34-35, 2012.


Addressing the evolving needs of cancer survivors at various stages of their illness and care, Palliative Care in Oncology is guest edited by Jamie H. Von Roenn, MD. Dr. Von Roenn is ASCO’s Senior Director of Education, Science and Professional Development Department.