Justification for Withholding and Withdrawing Life-Sustaining Therapy

The withholding and withdrawal of life support is justified by four ethical principles ( Table 104-1 ). The first principle is beneficence : the physicians’ obligation to do good for patients. Relieving pain and suffering, rather than sustaining life at all costs, may be beneficent in certain situations. The second principle is nonmaleficence : the physicians’ obligation to avoid harm. Life-sustaining interventions may be painful and unlikely to sustain life in a way judged meaningful by the patient, and foregoing such interventions may reduce harm. The third principle is autonomy : respect for the patients’ right of self-determination. Patient autonomy is reinforced when patients are allowed to refuse unwanted life-sustaining therapies. The fourth principle is justice : the fair allocation of medical resources. Justice may be served if life support is withdrawn from one patient with a poor prognosis to help another patient with a better prognosis, such as during a mass casualty event or in an influenza pandemic.

In the United States, the withholding and withdrawal of life-sustaining therapies is justified legally by the principles of informed consent and refusal, which have strong roots in common law. The right of adults who are capable of making medical decisions either to consent to or to refuse treatment was first established in Schloendorff v. Society of New York Hospitals in 1914. In this case, the New York Court of Appeals declared: “Every being of adult years and sound mind has the right to determine what should be done with his own body; and a surgeon who performs an operation without his patient’s consent commits an assault, for which he is liable in damages, except in cases of emergency when the patient is unconscious or when it is necessary to operate before consent can be obtained.”

The right of adults with decision-making capacity to refuse treatment was advanced in cases such as Bartling v. Superior Court and Bouvia v. Superior Court in California. In the first of these cases, the Court of Appeals allowed a man with advanced COPD to have a mechanical ventilator removed against the wishes of his physicians and of the hospital. In the second, the appeals court ordered a hospital to stop force-feeding a quadriplegic woman against her will even though she might die in the process. Most states have dealt with similar cases, and the principle that adults with capacity can refuse unwanted therapies is now widely accepted.

The U.S. Congress supported the importance of respect for patient autonomy by passing the Patient Self-Determination Act. This statute mandates that patients admitted to medical facilities be asked whether they have advance directives and, if they do not already have them, that they be assisted in drawing up directives. Advance directives are of two types: instructional directives that state what patients want done in a given situation (e.g., living wills) and proxy directives that appoint surrogates to make decisions for them in these situations (e.g., the durable power of attorney for health care). Although there are well-documented shortcomings, advance directives have the potential for extending patient autonomy beyond the point of incapacity, even though this potential has not been fully realized.

Because of their illness and sedation, many critically ill patients cannot partake in the medical decision-making process. In this circumstance, family members and other surrogates, if available, may consent to or refuse treatment for them. Parents have a long-standing right, indeed an obligation, to speak for their dependent children. The legal right of surrogates to act for incapacitated adult patients was established in In re Quinlan in which the Supreme Court of New Jersey allowed the parents of a vegetative patient to refuse mechanical ventilation for their daughter through the mechanism of surrogate decision making. Through this mechanism, family members also were allowed to make life-determining decisions for their adult relative in Barber v. Superior Court in California. In this case, the court determined that not only mechanical ventilation but also nutrition and hydration—indeed, any therapy that was not clearly benefiting the patient—could be forgone.

The U.S. Supreme Court dealt with the issue of surrogate decision making in the case of Cruzan v. Director, Missouri Department of Health . In this case, a family argued that it was unconstitutional for a chronic care facility in Missouri to deny their request to remove a feeding tube from their vegetative adult daughter. The facility argued that the daughter had not specified her wishes not to be fed artificially in advance of incapacity. In its Cruzan decision, the Supreme Court allowed Missouri and other states to require “clear and convincing evidence” of patients’ prior wishes before care was forgone for them. Nevertheless, it did not demand this requirement of other states, and it accepted the argument that the right of adult patients with decision-making capacity to refuse any and all therapies was protected under the Constitution.

Substituted judgment, through which surrogates make inferences about what treatment decisions patients would make if they were able to make decisions for themselves, is the highest standard under which surrogates may make medical decisions for incapacitated patients. Nevertheless, surrogates also may consider the patients’ best interests. One delineation of a best-interests standard is that of the New Jersey Supreme Court relative to In re Conroy . In this case, a nephew asked that a feeding tube be withdrawn from his elderly aunt, who had not declared her wishes before becoming demented, on the grounds that forgoing nutrition and hydration was in her best interest. The court would allow the best-interests standard in this case only because the burdens of continuing the patient’s life outweighed the benefits and because the recurring, unavoidable pain of life with treatment was such that administering therapies such as nutrition and hydration was inhumane. These requirements were imposed because the court considered the best-interest standard less compelling than the substituted-judgment standard.

Another legal perspective on the standards of substituted judgment and best interests came in the case of Wendland v. Wendland . This case involved a middle-aged man who was conscious but hemiplegic, uncommunicative, and unable to feed himself after a motor vehicle accident. After his feeding tube repeatedly became dislodged, his wife, who was his conservator, refused to authorize its reinsertion. In support of her decision, she cited her husband’s statements before his accident that he would not want to live in a severely debilitated condition. At this point, his mother went to court to block the plan not to reinsert the feeding tube, and it was reinserted. Mr. Wendland subsequently died, but the California Supreme Court deliberated on the issue of whether a conservator could withhold or withdraw life support from a patient who was conscious but incapable of articulating his current wishes. The Court ultimately ruled that feedings could be discontinued only if there was clear and convincing evidence that “the patient wished to refuse life-sustaining treatment or that to withhold such treatment would have been in his best interest,” which was not thought to be true of Mr. Wendland. The decision in his case suggests that the courts are unlikely to allow surrogates to limit treatment in patients who are neither terminally ill nor permanently unconscious unless the patients have specified what they would want done in such a situation.

Legal guidelines are less clear when incapacitated patients lack family members or other surrogates. Some states (e.g., Hawaii, Connecticut) allow physicians to make decisions for such patients on the basis of wishes expressed to the physicians when the patients had decision-making capacity. However, no state explicitly allows physicians to make decisions on the basis of a best-interests standard. Be that as it may, as shown in a study from seven U.S. medical centers, some physicians do make decisions to withhold or withdraw life-sustaining therapy from incapacitated patients on the basis of best interests, usually after consultation with their colleagues or with a hospital ethics committee. Other physicians ask the probate court to appoint conservators or other advocates for the patients, presumably to ensure a fair, transparent, and deliberate process of decision making for these vulnerable patients.

Before leaving the subject of justification for withholding and withdrawing life support, it is important to mention that decisions regarding this process are different for brain-dead patients than they are in patients who have lesser degrees of neurologic impairment. In the United States, death is defined as the total and irreversible loss of either cardiopulmonary function or the function of the entire brain. The determination of death by whole-brain criteria requires the demonstration of coma, indicating the loss of cerebral hemispheric function, and the documentation of absent corneal, oculovestibular, and ventilatory reflexes, indicating loss of function of the brain stem. Absence of ventilatory reflexes is demonstrated by an abnormal apnea test in which intubated, nonparalyzed patients removed from mechanical ventilation and given oxygen fail to initiate respiratory muscle efforts despite an increase of their arterial P co ₂ to at least 60 mm Hg.

Life-sustaining therapy usually can be forgone without a formal declaration of brain death. Conversely, some surrogates may insist that therapy be continued until patients die, in which case determining death by whole-brain criteria may be desirable. That said, the only patients in whom brain death must be determined are those who seem to be dead and whose organs will be transplanted after life support is withheld or withdrawn. Regarding the latter patients, one might argue that withdrawing life support from someone who is dead is an oxymoron. Because of this argument, and because brain-dead patients usually receive mechanical ventilation and other interventions only until their organs can be harvested, “life support” for them can be likened to “organ support.”

The ethical and legal justification for removing therapies from brain-dead patients is not just that the therapies generally are unwanted by patients and surrogates in this situation but also that the patients are dead and cannot benefit from them. Inasmuch as the patients are dead, physicians have no obligation to treat them, and they do not need to obtain consent from surrogates before withholding or withdrawing treatment. Nevertheless, consent is required for organ retrieval and transplantation. Furthermore, the families of brain-dead patients frequently do not understand the concept of brain death and consider the patients to be alive because their chests rise and fall as the ventilator cycles and a tracing of their cardiac rhythms is displayed on the bedside monitor. Families also may reject the concept of brain death for religious reasons.

Motivated by consideration for such families and by a desire to gain their consent for organ donation, many physicians are appropriately nonconfrontational in addressing the issue of brain death. They take time to explain what death means in a biologic and legal sense, how brain death is determined, and how valuable transplantation can be for organ recipients, as well as for the families of patients who serve as organ donors and even for the patients themselves. This approach is generally helpful to families, regardless of whether they ultimately approve transplantation. It is also useful in maintaining emotional equilibrium during a difficult time for all parties in the ICU.

Medical Futility

Although the ethical principle of autonomy and the legal principles of informed consent and refusal are the most compelling justifications for withholding and withdrawing life support, the concept of futility has also been used as a justification. This concept is invoked on the relatively rare occasions when patients or their surrogates request interventions (particularly those that are costly, scarce, or both) that physicians object to because they believe the patients cannot benefit from them. Some treatments, such as brain transplantation for brain-dead patients, are physiologically futile in that they cannot be accomplished. Other treatments, such as the use of mechanical ventilation in a permanently comatose or brain-dead patient, can succeed physiologically, but some physicians may consider them inappropriate because they cannot result in an outcome that the clinician judges to be valuable.

One group of investigators sought to define medical futility as an intervention that has been useless in the last 100 cases or that “merely preserves unconsciousness and cannot end dependence on intensive medical care.” In a similar vein, the American Thoracic Society argued that “a life-sustaining intervention may be withheld or withdrawn from a patient without the consent of the patient or surrogate if the intervention is judged to be futile. A life-sustaining intervention is futile if reasoning and experience indicate that the intervention would be highly unlikely to result in a meaningful survival for that patient. Here, meaningful survival specifically refers to a quality or duration of survival that would have value to that patient as an individual. Survival in a state with permanent lack of consciousness (i.e., completely lacking cognitive and sentient capacity) may be generally regarded as having no value for such a patient.”

Despite what has been called the “futility movement,” a broad concept of futility has never achieved consensus within the medical community in the United States or elsewhere, and the American Thoracic Society position has not gained broad support. A contrary position has been advanced by the Society of Critical Care Medicine, which holds that “treatments should be defined as futile only when they will not accomplish their intended goal. Treatments that are extremely unlikely to be beneficial, are extremely costly, or are of uncertain benefit may be considered inappropriate and hence inadvisable but should not be labeled futile. Futile treatments constitute a small fraction of medical care. Thus, employing the concept of futile care in medical decision making will not primarily contribute to a reduction in resource use.”

In some ways, the development of prognostic scoring systems can be seen as an attempt to predict which patients cannot benefit from therapies, especially those that are expensive, and thereby provide a rationale for denying such therapies to them. Yet most patients and families are quite willing to forgo such therapies even when physicians are not, as indicated by SUPPORT. Furthermore, a study of the theoretical implementation of a strict futility guideline for more than 4000 SUPPORT patients demonstrated that only minimal cost savings could be realized by not treating patients who were predicted to have a 1% or less 2-month survival. Nearly 75% of the savings in hospital days would have resulted from stopping treatment for 12 patients, one half of whom were younger than 51 years of age and 1 of whom lived 10 months when treatment was continued.

The emphasis on cost savings in the aforementioned study speaks to the fact that arguments about futility often have economic overtones. At one time, it appeared that American society could transition from a “rule of rescue,” in which large sums of money were spent on therapies providing only marginal benefit, to a “rule of reason,” in which this money could be used for primary and preventive care. Arguments based on futility seem to provide a rationale for forgoing treatment on the grounds that it is not worthwhile, apart from the matter of cost. Yet these arguments conceal value-laden assumptions, and they obscure the underlying issue of rationing.

Thus, the greatest problem with medical futility is not how it is defined but who defines it. In this regard, the debate over futility pits physicians, who believe that their training and experience enable them to know what therapies are truly useful and cost-effective, against patients and families, who may feel entitled to such therapies regardless of whether they pay for them directly. Physicians certainly are able to identify physiologic futility, and they are not legally or ethically required to perform procedures they consider nonbeneficial and below professional standards. Benefit, however, seems often to be in the eyes of the beholder: although physicians may regard mere prolongation of life as undesirable in certain circumstances, patients and families may consider it valuable. When these persons want treatments that physicians consider futile in more than a physiologic sense, which party should decide?

The American Medical Association Council on Ethical and Judicial Affairs has stated that, “Since definitions of futile care are value laden, universal consensus on futile care is unlikely to be achieved.” At the same time, the Council has recommended a process-based approach to futility determinations whereby patients or families who insisted on therapies that physicians considered futile would be required to undergo a dispute resolution process. Central to this process would be mediation by a hospital ethics committee. If resolution was not reached through mediation, the patients would be transferred to another institution or care would be terminated if transfer were impossible. So-called futility policies based on this model have been developed in Houston, Texas, and other cities.

Based on the Houston experience, the State of Texas adopted a law providing an extrajudicial due process mechanism for resolving medical futility disputes. A survey of Texas hospitals’ experience with this law indicates that a minority of the hospitals had used the due process mechanism and that life-sustaining treatment was discontinued against patients’ or surrogates’ wishes in only a small number of cases. Furthermore, the constitutionality of the Texas statute has not been determined at the appellate or U.S. Supreme Court level. As a result, the implications of the Texas statute for that state and for the rest of the United States are unclear.

To date, legal cases such as In re Helen Wanglie and In the Matter of Baby K suggest that judges are unwilling to allow physicians to withhold or withdraw support from patients when the physicians ask to do so and the patients or families object. In the first of these cases, a court in Minneapolis refused to replace a husband who was seeking continued life support for his wife, life support which her physicians considered nonbeneficial, with another conservator who might allow support to be forgone. In the second, a Virginia court required that physicians repeatedly resuscitate an anencephalic infant at the request of its mother but against their own wishes, reasoning that to do otherwise would violate the Emergency Medical Treatment and Active Labor Act.

In contrast, the case of Gilgunn v. Massachusetts General Hospital indicates that physicians are likely to obtain legal results more to their liking when they refuse to provide treatment they consider futile and defend their decisions in court as consistent with professional standards. In this case, a jury in Boston exonerated physicians at Massachusetts General Hospital for removing a patient from life support and writing a DNAR order for her over the objections of her daughter. The patient’s husband and other children did not protest the physicians’ actions, and the jury apparently believed that the physicians acted within the standard of care. This case does not set legal precedent because a written judgment was not rendered. Nevertheless, the jury decision in Gilgunn suggests some public support for the concept of futility.

How the debate over futility will be resolved is uncertain. Because patient autonomy is so widely accepted in the United States, because national health insurance does not exist there, and because patients and families are suspicious of managed care organizations, it seems unlikely that American physicians will soon be granted a mandate to restrict services on the basis of futility. This is not the case in other countries, however, where resources are limited, national health insurance exists, and physicians implicitly and explicitly are allowed to ration care. Whether physicians will continue to exercise this and other prerogatives if American-style patient autonomy becomes more prevalent in their countries remains to be seen.

Justification for Administering Palliative Care

The U.S. Supreme Court has provided ethical and legal justification for the administration of palliative care at the end of life, in the cases of Washington v. Glucksberg and Vacco v. Quill . These cases dealt with the constitutionality of laws prohibiting physician-assisted suicide in the states of Washington and New York. In Glucksberg , the Court decided that terminally ill patients do not have a liberty interest in committing suicide or in receiving a physician’s assistance in committing suicide because of the long tradition of prohibiting suicide in the United States and because of the states’ legitimate interest in continuing to make suicide illegal. In Vacco , it distinguished between assisted suicide and withholding and withdrawal of life support. “Everyone, regardless of physical condition, is entitled, if competent, to refuse lifesaving medical treatment; no one is permitted to assist a suicide,” the Court wrote. “When a patient refuses life-sustaining medical treatment, he dies from an underlying fatal disease or pathology; but if a patient ingests lethal medicine prescribed by a physician, he is killed by that medication.”

In Glucksberg and Vacco , five justices reasoned that Washington and New York could prohibit assisted suicide because these states had no barriers that prevented patients from receiving medications to relieve pain and suffering. However, as Justice Breyer wrote, “Were state laws to prevent the provision of palliative care, including the administration of drugs as needed to avoid pain at the end of life, an action against such law might be called for by the Supreme Court.” Through this and other statements, a majority of the justices suggested that being free of pain while dying was a liberty interest protected under the Constitution.

The Supreme Court distinguished assisted suicide from palliative care in Glucksberg and Vacco by accepting the ethical principle of double effect. Under this rule, acts such as giving sedatives and analgesics that lead to morally good effects, such as the relief of suffering, are permissible even if they produce morally bad effects, such as the hastening of death, provided that only the good effect is intended. The morally bad effect may be foreseen in that physicians are aware of its possibility and even its likelihood, but they may not wish it. The bad effect also may not be a means to the good effect, and the good effect must outweigh the bad one; that is, risking death is reasonable in palliating a terminally ill patient only if there are no less risky ways of relieving suffering.

The Supreme Court’s approval of palliative care included sanctioning the practice of terminal sedation , in which patients are rendered unconscious while life-sustaining therapies, including nutrition and hydration, are withdrawn. Under Vacco , a state may allow terminal sedation if it is “based on informed consent and the double effect. Just as a state may prohibit assisted suicide while permitting patients to refuse unwanted lifesaving treatment, it may permit palliative care related to that refusal, which may have the foreseen but unintended ‘double effect’ of hastening the patient’s death.”

Some have argued that the rule of double effect has many shortcomings as an ethical guideline, in particular because it overlooks the complexity of human intention. Such complexity was demonstrated in a study of the administration of sedatives and analgesics during the withholding and withdrawal of life support in two ICUs. In this study, physicians indicated that they ordered these agents primarily to decrease pain, anxiety, and dyspnea—but also to hasten death—in 39% of critically ill patients. In another investigation, 16% of a sample of ICU nurses reported that they had engaged in assisted suicide or euthanasia while trying to relieve patient suffering, often without physicians’ knowledge.

Just as some physicians and nurses have mixed motives in caring for dying patients, so do some family members want to ease suffering and hasten death simultaneously in their relatives. That such motivation is widespread presumably accounts for the fact that few physicians who are suspected of participating in assisted suicide or euthanasia have been punished through the criminal justice system in the United States. In general, physicians and other caregivers are unlikely to be prosecuted or even criticized if they act compassionately in administering sedatives and analgesics to treat distressing symptoms in dying patients and do so with informed consent.

In Glucksberg and Vacco , the Supreme Court did not judge the laws prohibiting physician-assisted suicide in Washington and New York to be unconstitutional. Nevertheless, it also did not prevent other states from permitting physician-assisted suicide if they chose to do so. Physician-assisted suicide was legalized in Oregon in 1997 under that state’s Death with Dignity Act. Experience over the subsequent 2 years indicated that few patients requested lethal medications, that physicians granted few requests, and that palliative interventions led some, but not all, patients to change their minds about assisted suicide. By and large, the decision to request and use a prescription for lethal medication stemmed from patients’ concern about loss of autonomy or control of bodily functions, not from fear of intractable pain or financial loss.

Physician-assisted suicide (in which physicians prescribe potentially lethal medications that patients themselves can take) and euthanasia (in which physicians actually administer the medications) are practiced in several European countries. Although none of these countries has legalized the practices, physicians in the Netherlands have not been punished for performing them since 1991, when a national study revealed that assisted suicide and euthanasia were being performed. Instead, Dutch physicians are required to report all cases in which they administered or supplied drugs with the explicit intent of hastening death. Patient requests for assisted suicide or euthanasia rose from 8900 in 1990 to 9700 in 1995 and remained stable at 9700 in 2003 according to one Dutch study. Assisted suicide was listed as the cause of death on only 0.2% of all death certificates in the Netherlands during the same years. In 1990, 64% of Dutch physicians thought that patients have a right to decide about their own life or death; the percentage of physicians was the same in 1995 but fell to 56% in 2001. These data suggest that, in the Netherlands as in Oregon, the demand for assisted suicide has not increased; in fact, over time, Dutch physicians appear to have grown more reluctant in their attitude toward this practice. How common assisted suicide will become in other European countries and how their laws will deal with these practices are unclear.

Two Models of the Physician-Patient Relationship

Medical decisions may be made by physicians, patients, their families, or other surrogates alone or in combination. How decisions are actually made depends in large part on the model of the physician-patient relationship that is used.

Perhaps the oldest model is paternalistic ; this model has also been called parental or priestly. According to it, physicians act as guardians in defining what their patients’ interests are and then serving these interests as they see fit, with little or no input from the patients themselves. With pediatric patients, the families’ input similarly might not be sought or might be disregarded. The paternalistic model overlooks the strong social norm of respect for patient autonomy and the reality that in pluralistic societies it is unlikely that physicians can reliably discern what treatments most advance the interests of their diverse patients. The paternalistic model once dominated medical decision making at the end of life in the United States. In fact, the early cases such as Bartling, Bouvia, and Quinlan were brought because physicians and hospitals were unwilling to let patients or surrogates refuse therapy. It is ironic—and perhaps a cautionary tale about the limits of medical paternalism—that some physicians today demand that patients and families forgo treatment that other physicians once insisted they accept.

In another model, the deliberative or shared model, physicians and patients/surrogates collaborate to help patients define their health-related values, discuss treatment options, and together decide which alternative is best. This process is more than the mere passing of information: It is a moral deliberation, based on a mutual understanding of the medical facts and the patient’s values. The shared model of medical decision making is best suited for situations in which physicians, patients, and families have adequate time to deliberate, when reasonable treatment options exist, and when the best course of action may vary according to patients’ values and preferences. Evidence from audiorecording ICU family conferences suggests that the process of shared decision making about life support for incapacitated, critically ill patients is often suboptimal. For example, clinicians frequently failed to explain principles of surrogate decision making, elicit patients’ values and preferences, and discuss reasonable treatment options, including a purely palliative approach to care.

Medical decision making and the attitudes that underlie it vary from country to country. For example, although the shared model is gaining popularity in France, the paternalistic model is still sometimes employed at the end of life in ICUs there. Physicians are granted legal decision-making prerogatives in these ICUs that are not permitted for their American counterparts. Furthermore, French families have had no legal right to make decisions for patients who cannot make decisions for themselves. Variations in attitudes also exist within a given country depending on its ethnic and cultural composition. For example, in one study in the United States, Korean Americans and Mexican Americans were significantly less likely than European Americans and African Americans to believe that a patient should be told the diagnosis of metastatic cancer or that the patient should make the decision about the use of life-sustaining therapy. In another study, African Americans were more likely than European Americans to want to be kept alive on life support. Interviews of these African Americans subjects documented a deep distrust of the health care system and a fear that access to health care was based on one’s ability to pay.

Importance of Physician, Patient, and Family Communication

Of course, American families would not need to be so involved in medical decision making in the ICU setting if physicians facilitated such decision making when the patients’ diseases were less advanced and they could speak for themselves. Yet SUPPORT and other studies have shown that physicians and patients rarely discuss end-of-life issues in advance of the patients’ deterioration, even when patients reside in nursing homes. A study of patients with severe COPD enrolled in pulmonary rehabilitation revealed that almost all had health concerns, the most common of which was fear of increasing dyspnea. Although many of the patients had concerns about being intubated, only a minority had completed an advance directive describing their choices. Furthermore, although the patients generally wanted discussions with their physicians, only 19% had such discussions, only 15% had discussed life support, and only 14% thought that their physicians understood their end-of-life wishes.

Why patients with chronic and terminal diseases do not discuss end-of-life care with their physicians has been the subject of several investigations. In the study of patients with COPD enrolled in pulmonary rehabilitation, those patients who had not had discussions gave the most important reasons as their own procrastination and the fact that their physicians had not brought up the topic. When these patients participated in an end-of-life educational intervention as part of their rehabilitation, they were more likely to pursue such discussions with their physicians and to complete durable powers of attorney for health care.

In another study of patients with advanced AIDS, structured interviews of the patients and their physicians were used to identify barriers to communication. Patients who had not had discussions with their physicians about end-of-life care most frequently agreed that, “I don’t like talking about getting very sick” and “I would rather concentrate on staying alive than talking about death.” Physicians who did not initiate discussions most often agreed that, “There is too little time during our appointments to discuss everything we should” and “I worry that discussing end-of-life with [patient name] will take away his or her hope.” Many physicians in this study acknowledged that they felt uncomfortable discussing end-of-life issues, suggesting that their own discomfort was as important as the lack of time for discussions in inhibiting communication.

Limited communication between physicians and patients was also reported in SUPPORT. Furthermore, SUPPORT revealed that communication was frequently inadequate between physicians and the families of critically ill patients. This finding was corroborated by a study from an ICU in France in which half the families of critically ill patients reported the same inadequacy.

A follow-up investigation from that and other ICUs in France indicated that family satisfaction with their relatives’ care was due in part to their being of French descent and having a language and cultural values similar to those of the ICU caregivers. Family satisfaction also related to information provided by hospital physicians, a patient-to-nurse ratio of 3 : 1 or less, knowledge of the specific role of each caregiver, help from the family’s own physician, sufficient time spent giving information, and an absence of perceived contradictions in information provided by caregivers.

In a study from one U.S. hospital, families were interviewed about their experiences in the ICU and the decision-making process for withholding or withdrawing life support. They reported a high incidence of conflicts, the vast majority of which were between themselves and physicians. The conflicts most often involved problems in communication or perceived unprofessional behavior, such as disregarding the primary caregiver in treatment decisions. These families identified pastoral care and prior discussion of treatment preferences as sources of psychological support. They appreciated lenient visiting hours and the availability of family conference rooms. Most of the families singled out attending physicians as the preferred source of information and reassurance.