Critical Care Plus: Withdrawing Life-Sustaining Support Calls for Clear Protocols 

Family education, consensus essential to successful process 

By Julie Crawshaw, Editor, CRC Plus

Because withdrawing or withholding life-sustaining support in the icu is the leading cause of ICU deaths, both practices call for well thought out protocols, says Gordon Rubenfeld, MD, MSc. Rubenfeld, who is assistant professor of medicine in the Division of Pulmonary and Critical Care at the University of Washington in Seattle, notes that though the legal and procedural issues surrounding the decision to withdraw support are well-documented, scant attention has been paid to the actual procedures this requires.

One reason may be that the very clear process and outcomes measures available for other types of in-hospital care don’t exist for end-of-life ICU care. "We don’t have validated tools for measuring end-of-life care or the quality of death and dying," Rubenfeld says. "Obviously, you can’t use "live or die" as a measurement for end-of-life care."

Stepped withdrawal from full life-sustaining support to no support is commonplace, Rubenfeld notes, but the practice needs to be dispassionately examined because it’s more often done for the clinicians benefit than for the welfare of the patient and family members.

"I don’t think there’s anything wrong with stepped withdrawal per se, but the data suggest that it happens to make the doctors and nurses feel better," Rubenfeld says. "I think we clinicians are actually negotiating with ourselves during that step-down process. We don’t want to feel like we’re giving up, and when we remove support gradually we lose that feeling that the decision we make leads to the patient’s death."

Many critical care clinicians feel more comfortable withholding life-sustaining measures than withdrawing them. Rubenfeld says he frequently hears physicians argue that they use stepped withdrawal to make family members feel better, yet some of the orders physicians issue are very strange, such as orders not to add a second blood pressure medication, Rubenfeld observes.

"It’s difficult to imagine a lay family member saying it’s okay for the patient to be on one blood pressure medication but not two. However, it’s a very common thing for doctors to do because we don’t want to say "no blood pressure-raising medicine" but we also don’t want to go so far as to add a second one."

To counter this, Rubenfeld advocates talking with the patient’s family about the goals of care overall but not discussing specific treatments. "It’s our job to operationalize those goals and we need to look at our own practices and biases very, very carefully and not allow those to drive our decision-making," he says.

Joanne Lynn, MD, president of Americans for Better Care of the Dying and director of the RAND Center to Improve Care of the Dying, agrees that it’s difficult to know whether reducing and withdrawing life-sustaining measures in steps will help the patient’s family.

"It may or may not be more merciful to give the family a little time to pull themselves together, to hold Mom in their arms once more or they may find it heartless and cruel to extend the time," she says. "Maybe that’s where the stepwise process makes a little bit of sense because people need to think about it a little bit."

Lynn adds that in addition to not succumbing to their own biases, ICU clinicians must be aware of the many ways in which family members communicate their feelings. "It’s communication on a very deep level, really understanding what people are feeling, all their ambivalences," Lynn notes.

She adds that sometimes communication is much more in the body language and whether the family still finds it possible to visit. Awaiting clear communication may appear to be a stepwise reduction in support when in fact physicians don’t have the permission they need to proceed, as when family members spend a few hours with the patient in order to gain clarity.

Communication Problems Can be Preempted

The transition from cure to comfort has a different trajectory for each patient, Rubenfeld observes, but many communication problems can be pre-empted by educating family about what death looks like. Most modern Westerners are almost completely insulated from the process and—thanks to television—most are also conditioned to expect that dying is a quick and quiet process in which it appears the patient is merely going to sleep.

"My experience is that we need to explain to family members that they are likely to see signs they would normally interpret as manifestations of discomfort but that those are actually part of the normal dying process," Rubenfeld says.

Families also need to understand that the purpose of doctors and nurses in end-of-life care is to make patients as comfortable as possible, but not to speed up the natural course of events. He outlines the principles of withdrawing life support as:

1. The goal of withdrawing life-sustaining treatments is to remove treatments that are no longer desired or do not provide comfort to the patient.

2. The withholding of life-sustaining treatments is morally and legally equivalent to their withdrawal.

3. Actions with the sole goal of hastening death are morally and legally problematic.

4. Any treatment can be withheld or withdrawn.

5. Withdrawal of life-sustaining treatment is a medical procedure.

6. Corollary to 1 and 2: when circumstances justify withholding an indicated life-sustaining treatment, strong consideration should be given to withdrawing current life-sustaining treatments.

Rubenfeld emphasizes the importance of distinguishing palliative care for patients who are dying in general from palliative care for patients who are dying in the ICU. He points out that there’s a very robust body of data indicating pain treatment is inadequate in acute care hospitals but far less evidence that such problems exist in the ICU.

There is, however, evidence that physicians don’t listen carefully enough to requests by patients and surrogates and that many patients receive overly aggressive end-of-life care. The best evidence came from a study done more than five years ago, which Rubenfeld says needs to be updated.

Family members and surrogates who want the ICU team to do more than is medically warranted pose another problem and they often need reassurance that everything that could have been done was in fact done.

Consensus is Important Even When Agreement Isn’t Reached

Though there is substantial data showing that coordinated care delivers better outcomes in end-of-life care, a difference of opinion about exactly when life-sustaining support should be removed remains common, Rubenfeld notes, with nurses frequently reaching the conclusion that it’s time to withdraw life support well before physicians do.

Even when consensus is not reachable, Rubenfeld says the process of trying to reach it is important for all concerned because it provides a forum for releasing thoughts and feelings. However, he strongly cautions against talking with the patient’s family before strong disagreements are resolved.

"If nurses and doctors can’t work together to deliver good ICU care in general, they’ll have a much tougher time agreeing on end-of-life care," Rubenfeld warns. "Everything depends on the culture of the hospital."

If consensus on withdrawing life-sustaining support can’t be reached, Rubenfeld advocates finding a mediator. "Ethics committees do a lot of this," he says. "They less frequently tell people what’s ethical than they help negotiate agreement, either between members of the team or members of the team and the family."

This is especially true in hospitals with "fractured" ICU care in units that lack a lead intensivist. Such ICUs may have a cardiologist, pulmonologist, nephrologist, infectious disease physician, and surgeon who are caring for different pieces of the patient but haven’t talked with each other about the big picture.

"The surgeon may think things are okay because his operation was a success while the nephrologist thinks they are terrible because the kidneys are doing poorly," Rubenfeld says. "You need to have someone quarterbacking the team."

David H. Beyda, MD, Chief and medical director of the Pediatric Critical Care department at Phoenix (AZ) Children’s Hospital, says that once the family and medical team have agreed that any further medical/therapeutic interventions are futile, they proceed rapidly to withdrawing all support except for pain and sedation medications.

As far as the family is concerned, Beyda says, there should never be an issue with them "coming to terms" with the process. "We spend a lot of time discussing the process, its ramifications, the changes they expect to see, their feelings, what support they have, and their involvement in it," Beyda says.

Withdrawing Support Calls for Specific Protocols

Rubenfeld and his colleagues have drawn up specific protocols for withdrawing life-sustaining support. He emphasizes the need to explain to family members what is being removed and why, and suggests that withdrawing life-sustaining support is a medical procedure that calls for the same levels of training and skills as in other components of critical care, including informed consent and documentation.

Here are the process steps he recommends for creating the appropriate setting and monitoring:

Separate the patient from the commotion of the ICU by moving the patient to a separate area or to an isolated room. In open units, curtains should be closed.

Turn off monitors, and, if possible, remove them from the room. Remove electrocardiographic leads, pulse oximeter and hemodynamic monitoring catheters. There is no point in monitoring physiologic parameters when the data generated will not alter care. Families attending the dying patient can become preoccupied with irrelevant numbers and waveforms instead of focusing their attention on the patient. Removing monitors also eliminates the alarms that would sound as patients die. Intensive nursing care supplemented by physical examination of the patient for blood pressure, pulse and respiratory rate is sufficient to identify manifestations of suffering and to determine when death occurs. We feel that removing patients from electronic monitoring is an essential step in the transition from curative to comfort care. Unfortunately, it is extremely difficult for clinicians to give up this technologic tether precisely because this step symbolizes the break from the physiologic monitoring that identifies the ICU.

Remove all tubes, lines and drains if this can be done without significant discomfort. Catheters may be left in place if removal would lead to painful obstruction, for example, of Foley catheters or biliary drains. Intravenous access should be maintained to administer analgesic medication. Remove unused intravenous pumps, resuscitation carts and other mobile technology from the room.

Liberalize visitation to the extent that it does not interfere with the delivery of care to other patients. Children should be allowed to visit if their parents approve.

Do not obtain further laboratory or imaging studies.

As feedback has come in about the procedures and processes he advocates, it has become discouragingly evident to Rubenfeld that he’s preaching to the choir. Most of those who attend the talks he and other members of his department give, or read the book they’ve written, are clearly only a small percentage of critical care clinicians.

"The challenge is how to get people who don’t attend or read to engage in this process," Rubenfeld says. "Frankly, I don’t know how to do that."

Rubenfeld says that if he could give only one piece of advice about end-of-life decision-making, it would be this: Unless clinicians sometimes withdraw or continue life support when they wouldn’t want it done that way for themselves, they are probably not fully engaging in a discussion with the family members and surrogates.

"It’s hard to imagine that families would always want what an ICU physician would want," he says. "If you only find yourself withdrawing or continuing life support when you fully agree with doing it, there’s some problem with the surrogate decision-making that’s happening. I’m not saying this should happen frequently, or that it should never be the case that you withdraw life support when you strongly feel that it shouldn’t be withdrawn. But from time to time there should be a case where you think This is what I would want, but I can understand how someone else would want it.’"


1. Curtis JR, Rubenfeld GD, ed. 2000. Managing Death in the ICU: The Transition From Cure to Comfort. New York, NY: Oxford University Press; 2000.

2. Rubenfeld GD. Withdrawing life-sustaining treatment in the intensive care unit. Respir Care. 2000;45(11):1399-1407, discussion 1408-1410.

3. For more information, contact Gordon Rubenfeld (206) 731-3356; Joanne Lynn (703) 413-1100 ext 5457; or David H. Beyda (602) 546-1784.