Chapter 9: Futile Interventions


Patients or surrogates sometimes request medical interventions that physicians consider irrational or pointless. The concept of futility seems an appealing way to resolve such disagreements. The term futility comes from a Latin word meaning “leaky” [1]. In classical mythology, the gods condemned the daughters of Danaus to carry water in leaky buckets. No matter how hard they tried, they could never achieve their goal of transporting water. By analogy, futile medical interventions would serve no meaningful purpose, no matter how often they are repeated.

Physicians might claim that judgment of futility is a matter of professional expertise and that they may decide unilaterally to forego futile interventions rather than share decision making with patients or surrogates. Because the term futility gives decision-making power to physicians, however, it must be used with caution. The term is fraught with confusion, inconsistency, and controversy.

Strict Definitions of Futility

Physicians use the term in different ways [2] – [4]. In three strictly defined senses, medical futility justifies unilateral decisions by physicians to withhold or withdraw interventions (Table 9-1).

Table 9-1.When Is an Intervention Futile in a Strict Sense?

  • Intervention has no pathophysiologic rationale.
  • Cardiac arrest occurs after refractory progressive hypotension or hypoxemia.
  • The intervention has already failed in the patient.

Intervention Has No Pathophysiologic Rationale

Case 9.1. Antibiotics not active against organism

A 74-year-old woman has progressive septic shock with methicillin-resistant Staphylococcus aureus (MRSA) infection despite treatment with appropriate antibiotics, fluids, and vasopressors. The patient’s family requests an antibiotic that they learned about on the Internet. The patient’s organism is resistant to this antibiotic.

In this case, there is no pathophysiologic rationale for the antibiotic because the organisms causing this patient’s illness are resistant. No clinical or physiologic benefit can be expected. Even if the family insists on the drug, there is no medical reason to administer it.

Cardiac Arrest Occurs After Progressive Refractory Hypotension or Hypoxemia

Case 9.2. Progressive multiorgan failure

The patient in Case 9.1 is now comatose, on renal dialysis, and on a ventilator. Despite fluid replacement and increasing doses of vasopressors, her mean arterial pressure is below 60 mm Hg. Her physicians want to write an order not to resuscitate her in case of a cardiopulmonary arrest.

In Case 9.2, cardiopulmonary arrest occurs because of progressive hypotension despite maximal support of the patient’s circulation and oxygenation. Effective circulation cannot be sustained in this patient despite all appropriate therapy. If her refractory hypotension progresses to cardiopulmonary arrest, cardiopulmonary resuscitation (CPR) could not restore effective circulation. Similarly, CPR would be ineffective if cardiac arrest occurs after progressive hypotexia that is refractory to treatment.

Intervention Has Already Failed in the Patient

Case 9.3. No response to CPR

A 54-year-old man suffers a cardiac arrest in the emergency room. CPR and advanced cardiac support are initiated promptly. The initial rhythm is asystole. After 30 minutes, there has not been any return of spontaneous cardiac rhythm or circulation. All measures recommended in the American Heart Association guidelines have been attempted. His family insists that resuscitation be continued.

An adequate clinical attempt of CPR has failed to achieve the fundamental goal of restoring effective circulation and breathing. It is pointless to continue or repeat interventions that have already failed.

These three strict senses of “futility” are as plain as the root metaphor of carrying water in leaky buckets. Such interventions will not achieve the goals set by the patient, and all physicians would agree that the interventions are useless. The determination that an intervention is futile in these strict senses is based on objective data or judgments within the expertise of physicians. Physicians have no ethical duty to provide interventions that are futile in these strict senses; indeed, they generally have an ethical obligation not to provide them.

Loose Definitions of Futility

The term futility is also used in several looser senses that are confusing, involve contested value judgments, and do not justify unilateral decisions by physicians to withhold interventions [3], [5], [6]. The phrase “not medically indicated” commonly is used in similar ways.

Case 9.4. Recurrent aspiration pneumonia and severe dementia

A 74-year-old man with severe dementia is hospitalized for the third time in 6 months for aspiration pneumonia. At baseline, he sometimes recognizes his daughter and smiles when watching television or listening to music. The daughter, his only surviving relative, insists that he be treated with antibiotics. The resident exclaims, “Treating him is futile! His dementia is not going to improve, and it’s inhumane to keep alive someone with such a poor quality of life.” The resident also argues that a Do Not Attempt Resuscitation (DNAR) order should be written on the basis of futility because CPR is so unlikely to succeed.

The Likelihood of Success is Very Small

Some physicians contend that an intervention should be considered futile if the likelihood of success in a given situation is extremely small—for example, no success in the last 100 attempts or less than a 1% chance of success [7]. There are problems, however, in setting a quantitative, probabilistic concept of futility. Why set the threshold at 1%? Some patients or families might consider a likelihood of success of 1% worth pursuing in some circumstances. However, some physicians might desire to make unilateral decisions to forego interventions whose likelihood of success is 2% or even 5%. Indeed, physicians commonly describe interventions as futile when the likelihood of success is far greater than 1% [8], [9].

No Worthwhile Goals of Care Can Be Achieved

Futility can be defined only in terms of the goals of care [4]. Some ethicists contend that the proper goal of medicine is not simply to correct physiologic derangements. For these writers, it is inappropriate to prolong life if the patient will not regain consciousness or leave the intensive care unit (ICU) alive [1], [7].

Individual patients or the public, however, may have sharply different views. Some people regard life as precious even if the patient will not regain consciousness. Indeed, some states have public policies that favor prolonging life in patients who will not regain consciousness [10]. Thus, physicians cannot define goals of care unilaterally, but should be guided by the patient’s values.

The Patient’s Quality of Life is Unacceptable

In some situations, some physicians might declare an intervention futile because they consider the patient’s quality of life unacceptable, for example, if he is in a persistent vegetative state (PVS) [1]. These physicians contend that sustaining biologic life is not an appropriate goal when the patient has no likelihood of regaining consciousness or interacting with other people. However, patients generally view their quality of life more favorably than family members or physicians. Quality of life needs to be assessed according to the goals and values of the patient and cannot be determined unilaterally by physicians.

Prospective Benefit is Not Worth the Resources Required

An intervention might be termed futile because the expected outcomes are not considered worth the effort and resources required. Allocation of resources, however, should be decided by society as a whole, not an individual physician acting unilaterally at the bedside (see Chapter 30). Asserting that such interventions are futile closes off this difficult but essential debate [11].

Practical Problems with the Concept of Futility

Several problems occur in practice when physicians make unilateral decisions to withhold “futile” interventions.

Judgments of Futility Are Often Mistaken or Problematic

Physicians often err when they claim that an intervention has a very low probability of success. One study analyzed cases in which residents had written DNAR orders on the basis of a probabilistic definition of futility [9]. In 32% of such cases, residents estimated the probability of survival after CPR to be 5% or higher. In 20% of cases, the estimated probability of survival after CPR was 10% or greater. Thus, the term futility was applied inappropriately when the probability of success was considered much greater than the 1% threshold for futility proposed in the literature.

Problems also occur when determinations of futility are based on quality of life. When residents judged that CPR would be futile because of unacceptable quality of life, they discussed quality of life with only 65% of competent patients [9]. It is ethically problematic for physicians to judge a competent patient’s quality of life without talking to the patient because doctors underestimate the extent to which patients believe their lives are worth living [12].

Unilateral Physicians Decisions Polarize Disagreements

Attempts by physicians to resolve disputes by claiming the power to act unilaterally commonly antagonize patients and surrogates. Many surrogates do not agree with physicians’ judgments of prognosis. In one small study, almost two thirds of surrogates of patients in critical care units doubted the accuracy of physician’s predictions of futility, and almost a third would continue life support with less than a 1% estimate of survival [13].

A larger study of relatives of patients who required mechanical ventilation and had a high probability of dying illuminated why relatives might reject physician’s estimates of prognosis. Less than 2% of surrogates based that their beliefs about the patients’ prognoses exclusively on prognostic information from physicians. Other additional factors included the patient’s character and will to live, the patient’s history of illness and survival, the surrogate’s observations of the patient’s appearance, and the surrogate’s optimism, intuition, and faith [14].

Furthermore, declaring one intervention futile might not settle other important issues in a case. For instance, a unilateral decision by physicians to withhold CPR in Case 9.2 might worsen disagreements about other interventions, such as mechanical ventilation, vasopressor support, and antibiotics for infection.

Physicians Confuse Futility and Best Interests

Physicians commonly confuse futility and best interests as a basis for their decisions [4]. Even if an intervention cannot be termed futile in a strict sense, physicians may recommend against it because the burdens outweigh the benefits, according to the patient’s values and goals, and try to persuade the patient or surrogate that the intervention is not in the patient’s best interests. Chapter 4 discusses in detail the concept of best interests.

Safeguards If Interventions Are Deemed Futile

Procedural safeguards ensure that physicians’ unilateral decisions to withhold “futile” interventions are appropriate. Open discussions help guard against errors and abuses. In the original meaning of “futile,” there is no controversy that a leaky bucket will not hold water. Similarly, it should not be difficult for a physician to persuade colleagues, the patient or surrogate that an intervention is futile in a particular case.

Obtain a Second Opinion

The physician who is considering a unilateral decision to forego a “futile” treatment should obtain a second opinion from a colleague or from the institutional ethics committee, to guard against inappropriate interpretations of futility.

Discuss the Intervention with the Patient or Surrogate

Some physicians believe that they need not discuss futile interventions with the patient or surrogate. To be sure, in Case 9.3 a vast array of interventions would be futile in a strict sense, such as cancer chemotherapy. It would be pointless to tell patients or surrogates of interventions that are irrelevant to the illness at hand. In some cases, however, physicians might not discuss pertinent interventions because they fear that the patient or surrogate will not agree they are futile. Physicians might use the idea of unilateral decisions about futility to avoid unpleasant discussions [15]. The best approach, however, is more discussion, not less.

Discussing “futile” treatments with patients or surrogates shows respect for patients and surrogates and clarifies their expectations, goals, values, concerns, and needs. It also helps physicians understand the patient-specific considerations that surrogates consider when they assess prognosis [14]. Sometimes after physicians better understand the values and goals of the patient, they are more willing to continue interventions.

Alternatively, after such discussions patients and families usually agree with the physicians that the interventions are highly unlikely to achieve the patient’s goals or that the burdens and risks are disproportionately heavy [16]. Chapter 14 gives specific suggestions for such discussions.

Establish Guidelines and Procedures on Futility

Hospitals should develop written guidelines about futile interventions [15], [17]. Written institutional guidelines demonstrate that unilateral decisions to forego futile interventions are based on carefully considered standards, not on ad hoc reasoning. Several states and cities have developed futility policies and procedures [15], [17].

Texas established a nonjudicial procedure for resolving disputes when a physician regards an intervention as “not medically indicated” and the patient or family disagrees [18], [19]. This is a broader category than futile interventions and similarly ambiguous. A hospital medical or ethics committee must be convened, and the patient or family must be notified and invited to the meeting.

If the committee agrees that the intervention is not medically indicated but the family disagrees, the hospital must try to work with the family to find another physician or institution willing to provide the intervention. After 10 days, if a transfer of the patient cannot be arranged, the physician and hospital is legally permitted to withhold or withdraw the intervention. The patient or family, however, may ask the courts to order that the intervention be continued beyond the 10-day period if it is likely that they will find a provider willing to accept the patient.

A study of the Texas law found that in 30% of cases, the committee determined that the intervention was medically appropriate and in another 3% of cases, the patient improved. Thus in one third of cases, the physician’s initial judgment of medically inappropriate was not confirmed [20]. In an additional 18% of cases, the intervention continued after the 10-day period. During the 10-day waiting period, 42% of patients died, and the family agreed to discontinue the intervention in 38% of cases [20].

The Texas experience also raises concerns about unfairness because a disproportionately high percentage of cases involved ethnic minorities [21], [22]. The Texas experience illustrates the need for policies and procedures to review cases when physicians seek to forego life-sustaining interventions over the objections of the patient or surrogate.


  1. The concepts of futility and “not medically indicated” are intuitively appealing but need to be used extremely carefully.
  2. When futility is strictly defined, physicians may—and indeed should—make unilateral decisions to withhold interventions.