Examining choice to cease mechanical ventilation

Perception of patient’s wishes forms decision

Rather than age or severity of illness, the strongest determinants of the withdrawal of ventilation in critically ill patients are often the physician’s perception that the patient preferred not to use life support or had a low chance of survival in the intensive care unit, a recent study1 by the Canadian Critical Care Trials Group and the Level of Care Study Investigators has found.

Decisions about withdrawal of life support are among the most complex and difficult choices faced by critical care physicians and families of patients, says Maurene A. Harvey, RN, MPH, FCCM, a critical care nurse educator and past president of the Des Plaines, IL-based Society of Critical Care Medicine.

In the study, researchers evaluated adults receiving mechanical ventilation in 15 intensive care units in the United States and Canada. They recorded baseline physiological characteristics, daily Multiple Organ Dysfunction (MOD) scores, the patient’s decision-making ability, the types of life support administered, any use of do-not-resuscitate orders, the physician’s prediction of the patient’s status, and the physician’s perceptions of the patient’s preferences about the use of life support.

They then examined the relationship between all of the separate factors and the withdrawal of mechanical ventilation. Of the 851 patients in the study, 539 (63.3%) were successfully weaned from the ventilator, 146 (17.2%) died while receiving mechanical ventilation, and 166 (19.5%) had mechanical ventilation withdrawn.

The need for isotropes or vasopressors was associated with the withdrawal of the ventilator, as were the physician’s prediction that the patient’s likelihood of survival in the intensive care unit was less than 10%, the physician’s prediction that future cognitive function would be severely impaired and the physician’s perception that the patient did not want life support used.

"I think the researchers found it surprising that the severity of the patient’s illness and the MOD scores didn’t affect the decision," Harvey says. "Not that they felt it wasn’t good that the physicians did take the patient’s and family’s wishes into account, but that it indicates an evolution of our understanding of the issues."

Research into end-of-life issues in intensive care is a young science, notes Mitchell Levy, MD, FCCM, director of the medical intensive care unit (ICU) at Brown University and Rhode Island Hospital in Providence.

"We don’t know a lot about how people die in critical care — that is, what that experience is like," Levy explains. "As we begin to look at things like whether these patients have a good death,’ we are starting to look at the factors influencing death in the ICU."

Recent studies indicate that approximately 90% of deaths in the ICU are preceded by withdrawal of life support of some kind, almost double the percentage reported in studies as recent as five years ago, he adds.

Therefore, the researchers felt it important to look at factors influencing the withdrawal of life support — particularly mechanical ventilation.

Discerning appropriate level of care

Historically, in critical care, clinicians have used all available interventions at their disposal to treat patients, and, conventional wisdom would hold that decisions would largely be based on clinical criteria, such as age or severity of illness, the study’s authors note.

But in recent years, clinicians, ethicists, and patient advocates have questioned the use of advanced interventions in so many cases, urging more judgment about which patients would be likely to benefit, Harvey adds.

Making these determinations is very difficult with critically ill patients. The illness and injuries they have, and the treatments available to use, are very complex and difficult to explain to family members. And not all patients may react in the same manner to available treatments. Patient and family perceptions about quality of life, and their subsequent wishes regarding treatment, may differ significantly from patient to patient.

Clinicians must evaluate complicated sets of data about each patient’s condition, convey this complicated information to the patient and/or family members, and then help them come to decisions about how to proceed.

Ideally, the physician will delineate the available choices and options for treatment, provide information about the possible outcomes of each course, and then help the patient or family member make a decision that is consistent with their values and wishes.

In the old days, it was "cure at all costs," Harvey says. With move toward more patient autonomy and decision making, the pendulum sometimes swings too far, with family members of critically ill patients faced with decisions that they find overwhelming and unprepared for.

The challenge for critical care practitioners now is to determine how to provide patients and families the appropriate context for making such decisions.

"More than any other area, in critical care, you really have to blend the science and the art of medicine," Harvey explains. "In the past, we tended to do research on critical care medicine and focus on the science of different interventions. It thrills me to see researchers finally focusing on the art. We keep pushing the envelope in science, about how we can help people survive things they never would have survived 10 years ago. We keep pushing that envelope; but at the same time, we need to keep pushing the envelope of our ethical discourse as well — to make sure we keep developing our experience and judgments."

Need for better info on patient wishes

The study indicates that physicians are making clinical determinations about the patient’s condition and prognosis and then using that information in conjunction with what information they have about the patient’s wishes, Levy says.

The bad news is that previous studies have indicated that, when physicians try to guess about a patient’s wishes at the end of life, they frequently are wrong. Studies of family members indicate they are often no better at guessing correctly about what measures a relative would or would not want, he adds.

"If we had a lot of confidence that physicians could reliably predict a patient’s wishes, then this would be very good news," Levy says. "But that is not the case."

Most people still do not enter an intensive care unit with documented advance care plans in place, and family and caregivers often have no information about what the person would have wanted.

In general, people do not discuss their wishes with family members because they either feel that their relatives’ opinions mirror their own or because they are uncomfortable talking about death, Levy says.

The public and clinicians need to become more comfortable discussing death — even when death is not imminent. In particular, health care providers should initiate conversations about advance planning with patients when they are diagnosed with chronic conditions, he adds.

ICU stays

Harvey also highlighted the study authors’ comparison of the outcomes of their patients with published outcomes of critical care patients treated in intensive care units a decade ago. The patients in the current study who remained on the ventilator were the ones most likely to survive in the intensive care unit. Ten years ago, patients were more likely to remain on a ventilator until dying in the ICU.

"I think this indicates our decision making is getting better," she adds. "We are not just pushing all the time. We are more likely to bring up the discussion [about terminating ventilation] early rather than late, instead of just keeping on until everything fails."

As further evidence of this, she points out that patients in the study who received vasopressors or isotropes were more likely to have ventilator support withdrawn.

Patients with lung disease, who are on ventilators, can do well if they have strong hearts that can compensate and help increase blood oxygenation. However, patients with lung disease, or lung injury, who also have cardiac problems, frequently are too ill to improve and derive much benefit from mechanical ventilation.

"One of the worst combinations you can have in critical care is a patient with lung disease on a ventilator who also has a bad heart," she notes. "The [use of medications to support blood pressure and cardiac function] related to withdrawal of ventilation indicates that they were making decisions appropriately, step by step, deciding each day — is this level of care appropriate for this patient at this time?"

Although patients or their surrogate decision makers may want to try intensive interventions at first, hoping for the best, once the patient’s condition worsens past a certain point, they may decide such interventions are not appropriate — and that is reflected in the current study’s information, Harvey notes.

It’s important for more research to be done into the "art" of critical care, so that best practices and ethical decision-making skills can be developed, Harvey says.

"We all have our clinical experience to draw from, and you need experience," she says. "But you also need solid research like this to back it up."


1. Cook D, Rocker G, Marshall J, et al. Withdrawal of mechanical ventilation in anticipation of death in the intensive care unit. N Engl J Med 2003; 349:1,123-1,132.

For more information

The Society of Critical Care Medicine (SCCM) has published guidelines for withdrawal of ventilator support. They are available on the society’s web site at www.sccm.org. Click on the heading "Professional Resources" and the subheading "Guidelines" to find them.

SCCM also has several patient information pamphlets available for download. These pamphlets are designed to explain different aspects of critical care and intensive care units to patients and families. They also are available on the SCCM web site under the heading "Patient and Family Resources."


Maurene A. Harvey, RN, MPH, FCCM, Society of Critical Care Medicine 701 Lee St., Suite 200, Des Plaines, Illinois 60016.