INTRODUCTION — The concept of medical futility is an ancient one. Hippocrates counseled clinicians not to treat patients who were “overmastered by their disease.” The idea that doctors should not recommend or provide treatments that could not help a patient has been a part of most codes of medical ethics for centuries.

Modern controversies about medical futility arise when doctors and patients (or family members) disagree about whether a particular treatment is futile or inappropriate/inadvisable/nonbeneficial in a particular circumstance. Uncertainty about the appropriateness or benefit of interventions is a product of different perspectives on what constitutes benefit to patients and how individuals (patients/families as well as clinicians) can vary dramatically in how they define value as related to life. Medical interventions can sustain organ function, thereby prolonging life. However, at times, the quality of that life (eg, lack of awareness of surroundings and/or inability to interact with surroundings without anticipated recovery) would lead many to determine that such a life lacks the quality or value warranting continued interventions to sustain it. In short, medical technological capacities collide with deeply ingrained moral and ethical beliefs and frameworks.

This topic review will cover the concept of medical futility as it applies to palliative care patients. Ethical issues related to futile and potentially inappropriate treatments that are specific to issues that arise in the intensive care unit (ICU) from a critical care perspective are presented separately, as is a more extensive discussion of ethical issues that arise in palliative care. (See "Ethics in the intensive care unit: Responding to requests for potentially inappropriate therapies in adults" and "Ethical issues in palliative care".)

AUTONOMY AND MEDICAL FUTILITY — The concept of patient autonomy emerged in the last half of the 20^th century. Prior to that, doctors would routinely make unilateral decisions about what treatments to offer and what to withhold [1,2]. There was no legal or (perceived) ethical need to discuss such decisions with patients or families. That began to change when the concept of “informed consent” became embedded in the law governing doctor-patient communication. To ensure informed consent, the doctor was required to explain the risks and benefits of treatment to patients and family members. (See "Ethical issues in palliative care", section on 'Framework for ethical reasoning' and "Informed procedural consent".)

The late 20^th century also saw the invention and promulgation of life-sustaining medical technologies that enabled the prolongation of life in patients who were far too sick to survive without such technologies [3]. These technological developments were synergistic with the legal and ethical changes in medical practice described above in changing the landscape of medical ethics. There were more choices to be made, and there was more of a need to include patients and families in the decision-making process.

However, this had advantages and disadvantages. While patients and families were now allowed to discontinue treatments they deemed to be at odds with their goals, they were also allowed to request treatments that clinicians believed to be inappropriate or futile. In the 1980s and 1990s, reports began to be published of patients and families insisting on medical interventions that would not be able to achieve the ends of reversing death and sustaining meaningful life. Many began to argue that clinicians should be allowed to unilaterally withhold treatments that they considered futile or potentially inappropriate [4]. Decisions in many of the early cases were made by courts of law.

Representative cases — The following cases highlight the ways in which controversies about medical futility arose and were played out in the courts:

●Helga Wanglie – Helga Wanglie was 86 years old when, in December 1989, she fell at home and broke her hip. After receiving care at the county hospital, she was discharged to a nursing home for rehabilitative care. There, she developed respiratory distress and was transferred back to the hospital where she was subsequently intubated and ventilated. She was unable to breathe on her own and remained on the ventilator for many months. She was transferred to another long-term care facility for chronic respiratory failure. During this time, she was conscious and aware of her environment. Ultimately, during attempts at weaning and separating from ventilation, she suffered a cardiopulmonary arrest with severe anoxic brain injury.

She was diagnosed as being in a persistent vegetative state (PVS). She underwent months of ongoing support with mechanical ventilation, gastrostomy feeds, and antibiotics for multiple recurrent infections. Her husband and children acknowledged the severity of her brain injury and recognized the medical view that this was irreversible.

The medical staff believed further care to not be in the best interest of Mrs. Wanglie. Her family believed that the medical staff should not “play God” and that stopping medical care would be against their religious views. Multiple family conferences were convened, and an ethics consult was obtained. The family eventually agreed to placement of a do-not-resuscitate (DNR) order but did not want her removed from the ventilator or other treatments withheld. Medical staff and administration agreed with the ethics committee recommendation to stop interventions. The hospital and family sought another hospital to take over Mrs. Wanglie’s care, which could not be achieved [5,6].

The hospital requested that the family obtain a court order to continue care. When the family did not, the hospital filed a petition with the court. The court appointed Mr. Wanglie as his wife’s guardian and supported his decision to continue care. The decision came three days before Mrs. Wanglie's death on July 4, 1991 at the age of 88 from sepsis [7].

●Baby K – Stephanie Keene, referred to as Baby K during the ethical debates that surrounded her life, was an infant with the prenatal diagnosis of anencephaly, with a brainstem to maintain heart rate, blood pressure, and some respiratory effort but no higher cognitive abilities. Her mother refused the recommendation to terminate the pregnancy on religious grounds, and Baby K was born on October 13, 1992.

Consistent with her underlying pathology, Baby K had significant respiratory insufficiency at birth and throughout her life. Her mother refused to place a DNR order as advised by medical staff. Baby K was on mechanical ventilator support for the first six weeks of life, during which time attempts were made to transfer her to another hospital, but no other facility would accept her. She was eventually weaned from the ventilator and transferred to a long-term care facility for other medical needs. However, she had continual episodes of respiratory distress for which she was frequently transferred back to the hospital for ventilator support. Baby K’s mother insisted upon continued resuscitation and intubation. Baby K’s father, who was not married to the mother, supported the hospital in its objections to ongoing resuscitation.

The hospital filed a petition with the courts to appoint a guardian and to allow the hospital to provide only palliative care. The American Academy of Pediatrics wrote a brief in support of the hospital. The court gave a sweeping decision that in providing only palliative care against the wishes of the mother, the hospital would be in violation of multiple federal laws, including the Emergency Medical Treatment and Labor Act (EMTALA), the Americans with Disabilities Act, and the Rehabilitation Act, as well as the mother’s constitutional rights as protected by the 14^th Amendment [8]. The decision was affirmed by the federal court of appeals on only the basis of EMTALA [9].

There was clear disagreement within the medical ethics community, with Dr. Michael Grodin, a pediatrician and head of the Program on Medical Ethics at the Boston University School of Medicine and Public Health, stating, “For a baby that has no brain and is born dying, life for just another minute or another day is not an appropriate goal. It’s saying that it’s the technology that has life, not the patient.” On the other hand, Dr. Robert M. Veatch, director of the Kennedy Institute of Ethics at Georgetown University, stated that the mother’s and court’s decision “is well within the range of parental discretion.” The discrepant opinions that surrounded this case reveal the heated debate over the concept of medical futility and how it is defined and applied [10].

Notably, in both cases, the courts ruled in favor of family members advocating for continued medical intervention and against the medical providers who were arguing that such intervention was “futile” or not warranted due to the compromised state of the lives of the individual patients in question. On the other hand, the procedural posture in these cases was unique. Since almost all these treatment conflicts concern inpatients, EMTALA is almost never as relevant as it was for Baby K. In Wanglie, the hospital asked the court to rule on the propriety of the surrogate instead of the propriety of the treatment itself.

●Charlie Gard – Charlie Gard was born on August 6, 2016. At two months of age, he became lethargic and developed respiratory failure. He was admitted to the intensive care unit (ICU) at Great Ormond Street Hospital (GOSH) in London and started on mechanical ventilation. He was eventually diagnosed with a mitochondrial DNA depletion syndrome (MDDS), a rare clinically and genetically heterogeneous condition [11].

Through their own research, Charlie’s parents discovered an experimental therapy called “nucleoside therapy.” They asked the doctors at GOSH to consider a trial of nucleoside therapy, and the physicians agreed, but then Charlie developed status epilepticus, suffering further brain damage; the doctors then concluded that nucleoside therapy would thus be futile, and they recommended withdrawal of life support. The parents requested instead that Charlie be transferred to the United States and to the care of a doctor who was willing to try nucleoside therapy. The doctors at GOSH thought that the treatment would be futile and thus transfer would not be in Charlie’s best interest, and GOSH petitioned the United Kingdom High Court of Justice to allow Charlie’s mechanical ventilation to be discontinued over the parents’ objections.

The High Court concluded that Charlie was suffering and therefore it was in his best interests to have life support withdrawn. The High Court judge also stressed that the decision was not based in any way on the cost of treatment [12].

The parents appealed. Charlie’s lawyers argued that even a small chance of benefit from a trial of therapy outweighed the potential harms. More importantly, they argued that the well-informed and caring parents, in collaboration with a doctor of their choice, should make the judgment about the balance of harms and benefits. The Court of Appeals upheld the High Court’s opinion [13]. The United Kingdom Supreme Court decided not to hear an appeal [14]. As a last resort, the parents appealed to the European Union Court of Human Rights. That court concluded that the decision to withdraw life support did not violate Charlie’s human rights [15]. Eventually, the outside consultant came to agree that Charlie’s disease had progressed to the point where therapy would no longer be beneficial. The parents gave up their fight, and Charlie was transferred to hospice, where he died.

This case has implications for the process of shared decision-making in such circumstances. Generally, when physicians disagree about the care of a child, we defer to parents and allow them to choose which physician’s recommendations to follow [16]. In the United States, this is true even in states such as Texas that permit unilateral physician decisions to withdraw life support [17]. Under Texas law, when doctors determine that further life support is inappropriate, parents are given 10 days to find another physician. If they do, their child will be transferred to the care of that physician [18]. In California, this was even true for a child who was declared brain dead [19]. The precedent in the Gard case suggests that, instead, courts may make an independent evaluation of which medically endorsed treatment is best.

●Tafida Raqeeb – In dozens of cases, many with extremely public profiles (like Alfie Evans), British courts sided with clinicians in medical futility disputes as they did in the Charlie Gard case [20]. But in 2019, the court did not agree with clinicians at Barts Health NHS Trust that stopping life-support treatment was in the best interest of five-year-old Tafida Raqeeb [21]. Tafida had suffered from a brain arteriovenous malformation, which resulted in a burst blood vessel in her brain. Key to the court’s decision was the family’s Muslim beliefs and their plan to transfer Tafida to a hospital in Italy. The court found Tafida had a minimal awareness of her surroundings, was possibly able to interact with them, and was not in any pain. The court further found that while Tafida was unlikely to recover significantly, she might be able to breathe without ventilation. Finally, the court noted that children in the United Kingdom in similar conditions to Tafida are already receiving treatment of a similar nature. The United Kingdom cases demonstrate that neither physician nor parental authority is absolute, and that patient nuances must be considered in respect to other cultural values.

●Tinslee Lewis – Tinslee Lewis was born prematurely on Feb 1^st, 2019, in Texas. She was diagnosed with Ebstein’s anomaly with pulmonary atresia. She underwent a number of cardiac surgeries to palliate her heart disease but developed worsening chronic lung failure with severe pulmonary hypertension. She remained in the ICU, ventilator-dependent and heavily sedated. When she reached six months of age, her doctors’ opinions were that further interventions could not save her. They believed even routine care caused pain and suffering. They recommended the withdrawal of life support, but the mother disagreed. The doctors followed the processes outlined by Texas law for such cases (see 'Texas Advance Directives Act' below). They sought and received the concurrence of the hospital ethics committee. They notified the mother of this. They reportedly contacted 20 other hospitals to see if they would accept Tinslee in transfer. None would [22].

A lower court ruled in December 2019 that it would be legally permissible to remove Tinslee from the ventilator. But in July 2020, the Texas Court of Appeals reversed, sending the case back for trial in the Tarrant County District Court [23]. The appellate court held that since the physicians and hospitals would be “state actors” when withdrawing treatment over parental objections, they must comply with requirements of constitutional due process. The hospital had failed to do that because they did not provide the mother with sufficient notice or opportunity to participate in the ethics committee deliberations. Moreover, the committee lacked sufficiently objective standards by which to reach its decision. The mother appealed.

DEFINITIONS

Futility — Medical futility has proven difficult to define. In the Oxford English Dictionary, the word futility is defined as “leaky, vain, falling out of the desired end through intrinsic defect.” The word is derived from the Latin “futtilis,” meaning a vessel wide at the top and narrow at the bottom that tipped over and poured out its contents whenever it was filled. These definitions and word origins invoke the images of the Danaïdes from Greek mythology, eternally filling an unfillable vessel to atone for their sins [24]. Doctors and nurses caring for extremely ill patients could resonate with the feeling of carrying water in a leaky bucket, continually working towards a goal that could not be achieved.

Attempts to provide clarity to the definition of futility have led to multiple qualifiers of the term, including physiological futility, quantitative futility, qualitative futility, and potentially inappropriate or inadvisable therapy.

Physiological futility — Physiological futility describes medical interventions that could not possibly result in a physiological goal [25,26]. Using a specific antimicrobial, for instance, for a bacterial infection that is resistant to that antimicrobial is physiologically futile. Another example is performing cardiopulmonary resuscitation (CPR) on a patient with a ruptured myocardium.

This definition carries the assumption that each medical intervention has a specific physiological goal and that the clinician is able to determine if achieving that goal is at all possible. An action or treatment cannot be considered futile unless its goal is clearly stated. Critics of the use of the term “physiological futility” argue two points:

●Goals may exist other than the perceived physiological goal. For example, in performing CPR on a terminally ill patient, perhaps the true goal is not to restore a spontaneous heartbeat but rather to give the family a measure of certainty that everything has been attempted. In this instance, CPR could be considered “physiologically futile” but not necessarily futile if we were to apply other definitions.

●The term “physiological futility” implies that a clinician knows with certainty the likelihood that an intervention cannot reach a physiological goal. However, the clinician’s predictive abilities may not be that absolute.

Quantitative futility — The term quantitative futility focuses on the numeric probability of achieving the intended goal of therapy, which is generally presumed to be cure and survival [27]. It is used interchangeably with physiological futility by some, although some argue that quantitative futility, the likelihood or chance that a treatment will benefit a patient, differs from physiologic futility, which measures the outcome achieved, whether likely or not, by the effect on the patient’s physiology [28]. Properly understood, physiological futility is quantitative futility with a probability of zero.

Those who advocate the use of this term recognize that a possibility of death or survival is not absolute and that some probability threshold should be established. This approach utilizes published data on medical outcomes incorporated with personal clinician experience to determine a probability of survival following a given treatment.

While this mechanism of applying data attempts to make decisions more straightforward, it requires a judgment call on what level of likelihood should be the threshold for making a decision. Some recommend that clinicians use empiric evidence, including personal experience or published data, to assess the likelihood of a treatment being successful. One proposal is that if the last 100 cases experienced or reported indicate that the treatment did not achieve the desired end, then it could be regarded as futile [27]. However, is ≤1 percent possibility of survival an appropriate threshold [29]? What if another clinician or family member believes that a 0.5 percent chance of survival warrants aggressive therapeutic interventions? Whose threshold should be used for making the decision? Investigators have shown wide variations in the threshold of judging quantitative futility when physicians and laypeople are compared [30].

Another problem is that most published studies that purport to guide determinations of futility are based on insufficient data to provide statistical confidence for clinical decision-making [31]. In addition, available published data on survivability may already be skewed by health care perceptions on futility. Published mortality rates for a specific patient demographic may be 100 percent or close to it because attempts at improving this outcome are not made because of the poor prognosis, thus creating a self-fulfilling prophecy.

Qualitative futility — Viewing futility from a qualitative viewpoint shifts the focus from achieving a level of certainty to a focus on the quality of the potential benefits [32]. In other words, there is a distinction made between an effect of treatment that is limited to some part of the patient’s body and a benefit that improves the patient as a whole that can be perceived and appreciated by that patient [28].

If a therapeutic intervention might prolong survival even a small amount, with minimal suffering and an improvement in quality of life, one could refute its futility. On the other hand, if an intervention would cause significant suffering while offering only a brief (or possibility of brief) extension to the length of life, the opposite could be argued. This means of determining futility, however, is subjective and requires individuals to make judgment calls on what constitutes quality and a decision on what is “good enough.” This may lead to disagreements between clinicians, patients, and family members, who may view the balance of benefit and harm from certain therapeutic options quite differently.

Imminent demise futility — Other ways to define futility include “lethal condition futility” and “imminent demise futility” [25]. In both of these situations, the underlying condition is terminal in the short term (days, weeks, or months) and cannot be reversed or impacted by treatment. While an intervention may have a physiological impact upon the patient, such as CPR restoring a spontaneous heartbeat, they argue that the intervention will not alter the trajectory of the underlying illness and is therefore futile.

Two states have implemented this concept of futility in their Physician Orders for Life-sustaining Treatment (POLST) laws. Vermont permits a physician to write a do-not-resuscitate (DNR) order without consent when CPR would not prevent “imminent death” [33]. Similarly, Maryland permits a physician to write a DNR order without consent when CPR would be “medically ineffective,” which means that it would not prevent “impending death” [34].

Because of the multiple difficulties in creating a clear definition and criteria for futility, some have argued that the term is so value-laden that it is impossible to apply in an objective manner. These critics suggest that the term be abandoned altogether [35]. Some suggest that, instead of the term “futility,” we should talk about “potentially inappropriate treatment” [36].

Futility versus cost-effectiveness — Some believe that the issue of futility will inevitably be submerged in the debate about health care reform and utilization of limited resources, and the need for society to determine where resources are appropriately spent [37]. Societies with nationalized health care have certainly made decisions about how money and resources should be allocated based upon cost-effectiveness.

However, there is a clear distinction between cost-effectiveness and futility, and between futile and rationed treatments [38]. Clinicians should not label interventions as futile simply because they believe the treatment is not cost-effective. Care that may be expensive or resource intensive may not necessarily be futile, as it may provide benefit to the patient and achieve specific goals. Focusing on the impact to society minimizes the values of specific benefits or harms to the patient, which should be the primary factors considered.

Futile versus potentially inappropriate treatment — In 2015, five critical care societies, the American Thoracic Society (ATS), the American Association of Critical-Care Nurses (AACN), the American College of Chest Physicians (ACCP), the European Society of Intensive Care Medicine (ESICM), and the Society of Critical Care Medicine (SCCM), officially endorsed new and specific terminology that should be used when discussing and categorizing treatments of questionable benefit [36] (see 'ATS/AACN/ACCP/ESICM/SCCM (2015)' below):

●The five-society statement maintains that the term “futile” should be used only in the rare circumstance that an intervention definitely cannot accomplish the intended physiological goal. The statement provides the example of performing CPR on a patient with a ruptured myocardium. (See 'Physiological futility' above.)

●The term “potentially inappropriate” should be used, rather than “futile,” to describe the more common situation in which treatments have at least some chance of accomplishing the effect sought by the patient, but clinicians believe that competing ethical considerations justify not providing them. In contrast to “futile” treatment, disagreements over a “potentially inappropriate” treatment are value based.

We agree with these definitions.

Other treatments for which benefit is considered unlikely, uncertain, or controversial should be discussed when applying the terminology of potentially inappropriate, rather than futile, treatments. These issues, as well as suggestions as to appropriate responses to requests for inappropriate/inadvisable treatments in the intensive care unit (ICU), are discussed in more detail separately. (See "Ethics in the intensive care unit: Responding to requests for potentially inappropriate therapies in adults", section on 'Definitions' and "Ethics in the intensive care unit: Responding to requests for potentially inappropriate therapies in adults", section on 'Responding to requests for potentially inappropriate interventions'.)

Proscribed and discretionary treatment — In addition to distinguishing futile and potentially inappropriate treatment, the five-society statement (ATS/AACN/ACCP/ESICM/SCCM) also carves out two other categories [36]:

●“Legally proscribed” treatments are those that are prohibited by applicable laws, judicial precedent, or widely accepted public policies. The statement offers the example of circumventing organ allocation policies.

●“Legally discretionary” treatments are those for which there are specific laws, judicial precedent, or policies that give physicians permission to refuse to administer them. The statement offers the example of POLST laws in Maryland and Vermont that permit clinicians to write DNR orders without patient or surrogate consent when CPR would not prevent the patient’s “imminent death.”

THE COMPLEXITY OF PROGNOSTICATION — One of the major barriers to the definition and acceptance of medical futility and potentially inappropriate treatments is the medical community’s struggle with prognostication. Some degree of uncertainty almost always exists in any prognostic estimate that is delivered to patients and their families. Clinicians often struggle with this uncertainty and how to honestly communicate this appropriately to families [39,40]. As an example, in Kalanithi’s book, When Breath Becomes Air, Dr. Kalanithi, a neurosurgeon, is diagnosed with metastatic cancer. He asks his oncologist how long he has to live. She refuses to give him a number and says, instead, that he needs to think about what he wants to do while he is alive, rather than how long he will live [41]. Giving a precise prognosis, while necessary to resolve intractable disagreements about potentially inappropriate treatments, is always an matter of probabilities rather than certainties. (See "Communication of prognosis in palliative care", section on 'The science of estimating prognosis' and "Survival estimates in advanced terminal cancer", section on 'Accuracy of survival estimates'.)

Attempts have been made to use individual patients’ data to predict the likelihood of survival. Many of these have been used to predict the likelihood of survival among critically ill patients in the intensive care unit (ICU). Predictive scoring systems such as these cannot be [42] used to predict outcomes in a specific individual patient, and they are not appropriately used for determining issues of futility. (See "Predictive scoring systems in the intensive care unit".)

MAJOR POLICY STATEMENTS — Due to the need for consensus on how to address concerns regarding futile or potentially inappropriate treatments, multiple groups, including professional societies, have established policies or guidelines.

American Medical Association (1997) — In 1997, based on the report from its Council on Ethical and Judicial Affairs, “Medical Futility in End-of-Life Care,” the American Medical Association (AMA) released its policy in Opinion 2.037 [43]. It states: “When further intervention to prolong the life of a patient becomes futile, physicians have an obligation to shift the intent of care toward comfort and closure.” In order to help with this process, it recommends that all hospitals develop medical futility policies, and it further includes a suggested algorithmic approach to ensure due process.

American Academy of Hospice and Palliative Medicine (2011) — The American Academy of Hospice and Palliative Medicine (AAHPM) has an official position on withholding and withdrawing nonbeneficial medical treatments, which are considered ethically indistinguishable and both ethically and legally permissible [44]. The policy, however, does not discuss utilizing the concept of futility, and the term is not addressed. It does state that an ethics consultation should be considered if disagreements exist between the health care team and patients or their surrogates.

American Thoracic Society (2015) — In addition to being part of the five-society statement on potentially inappropriate treatment, the American Thoracic Society (ATS) also has an updated policy statement on managing “conscientious objections” (ie, situations where the clinician has a moral objection to providing or disclosing information about a legal, professionally accepted, and otherwise available medical service) in intensive care medicine [45]. The policy recommendations are based on the dual goals of protecting patients’ access to medical services and protecting the moral integrity of clinicians. Their approach, which is outlined in an algorithm (algorithm 1), has many similarities to those of the AMA and the Houston Task Force. This statement emphasizes that an individual clinician should not use their conscientious objection to justify unilaterally deciding to forego what they perceive to be futile treatment.

ATS/AACN/ACCP/ESICM/SCCM (2015) — In addition to distinguishing futile, proscribed, discretionary, and potentially inappropriate treatment, as described above, the five-society statement (ATS, American Association of Critical-Care Nurses [AACN], American College of Chest Physicians [ACCP], European Society of Intensive Care Medicine [ESICM], Society of Critical Care Medicine [SCCM]), offers recommendations for how clinicians should respond to each type of request [36] (see 'Futile versus potentially inappropriate treatment' above):

●Clinicians should not provide futile or proscribed interventions. In responding to requests for futile, proscribed, or discretionary treatment, clinicians should carefully explain the rationale for treatment refusal.

●Responding to requests for potentially inappropriate treatment is more complicated. These value-based disagreements require a fair-process approach for conflict resolution. The recommended approach, as reflected in an algorithm to address disputed treatments in the intensive care unit (ICU) (algorithm 2), includes the following seven-step process:

•Enlist expert consultation to continue negotiation during the dispute-resolution process.

•Give notice of the process to surrogates.

•Obtain a second medical opinion.

•Obtain review by an interdisciplinary hospital committee.

•Offer surrogates the opportunity to transfer the patient to an alternative institution.

•Inform surrogates of the opportunity to pursue extramural appeal.

•Implement the decision of the resolution process.

The five-society statement recognizes that time pressures may sometimes make it infeasible to complete all steps of the process. In these cases, clinicians may refuse to provide the disputed treatment if they have a high degree of certainty that the requested treatment is outside accepted practice, and if they complete as much of the seven-step process as possible.

American College of Physicians (2019) — The 2019 statement from the American College of Physicians (ACP) states the following [46]:

“In the circumstance that a specific treatment desired by the patient is medically ineffective and potentially harmful, the physician is not ethically obliged to provide such treatment (although the physician should be aware of any relevant state law). The physician need not provide an effort at resuscitation that cannot conceivably restore circulation and breathing, but he or she should help the family to understand and accept this reality. The more common and much more difficult circumstance occurs when treatment offers some small prospect of benefit at a great burden of suffering… but the patient or family nevertheless desires it. If the physician and patient (or appropriate surrogate) cannot agree on how to proceed, there is no easy, automatic solution. Consultation with learned colleagues or an ethics consultation may be helpful in ascertaining what interventions have a reasonable balance of burden and benefit. Timely transfer of care to another clinician who is willing to pursue the patient’s preference may resolve the problem. Infrequently, resort to the courts may be necessary. Some jurisdictions have specific processes and standards for allowing these unilateral decisions… A physician who writes a unilateral DNR order must inform the patient or surrogate and give a detailed explanation when doing so.”

OTHER POLICY STATEMENTS — While the policy statements described above are the most recent and relevant, other groups and professional societies have established policies or guidelines.

President’s Commission on Biomedical Ethics (1983) — The Presidential Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research was convened in 1978 and continued its work through 1983. It released the report, “Deciding to Forego Life-Sustaining Treatment,” to address many of the evolving ethical concerns around advancing medicine technology [47]. The report stressed the importance of patient and surrogate autonomy in making decisions on life-sustaining treatment. It is the clinician’s responsibility to ensure that ethically appropriate decisions are made, and the clinician is not required to provide therapy that they find ethically objectionable; however, the clinician must not abandon their patient and instead should transfer care to another provider.

Houston Citywide Taskforce on Medical Futility (1996) — In response to the American Medical Association (AMA) report (see 'American Medical Association (1997)' above), the city of Houston, Texas developed a multi-institution collaborative policy on medical futility that was adopted by a number of large Houston hospitals [48]. They required steps nearly identical to the recommendations of the AMA to determine medically inappropriate interventions.

Society of Critical Care Medicine (1997) — The Ethics Committee of the Society of Critical Care Medicine released its statement on medical futility policies in 1997 [49]. It defined futility as providing treatments that provide no physiological benefit to the patient. In situations of true physiologic futility, they state the treatment should never be offered. However, as discussed above, they recognize that, in the majority of situations where the term “futility” is invoked, the correct terminology should be “potentially inappropriate” or “inadvisable” treatment. In these situations, the health care team is advised to explore and address the sources of conflict and work to define therapy goals. (See "Ethics in the intensive care unit: Responding to requests for potentially inappropriate therapies in adults".)

San Diego County Medical Society (2009) — The San Diego County Medical Society created its own policy addressing the use of nonbeneficial treatments [50]. This policy is unique in that it recognizes that not all disagreements about medical futility are between doctors and patients or family members. They categorize three types of disagreements: intraprofessional (between members of the treating team); intrafamilial; and between the treating team and the patient/surrogate. The policy outlines means of mediating the different types of conflicts.

Renal Physicians Association (2010) — The Renal Physicians Association recognizes that conflicts may occur between the patient/legal agent and the renal care team about whether dialysis will benefit the patient. In addition, conflicts also may occur within the renal care team or between the renal care team and other health care providers. A 2010 clinical practice guideline advises physicians to “try to understand [the patient/agent] views, provide data to support his/her recommendation, and correct misunderstandings. If dialysis is indicated emergently, it should be provided while pursuing conflict resolution, provided the patient or legal agent requests it.” [51]. (See "Kidney palliative care: Withdrawal of dialysis" and "Kidney palliative care: Principles, benefits, and core components".)

National Council on Disability (2019) — The National Council on Disability (NCD) recognizes that when a physician decides that providing or continuing health care treatment would be “medically futile,” there are a number of objective, evidence-based factors that can impact this decision. But the NCD is concerned that a physician’s subjective judgments about whether a patient’s life would be “worth living” may go unrecognized and can also play a role in decision-making [52]. Some health care providers, most unknowingly, harbor biases and misperceptions about the quality of life and capacities of people with disabilities. These assumptions can and have impacted physicians’ willingness to provide or continue life-sustaining care.

NOTABLE STATE LAWS

Texas Advance Directives Act — Recognizing the struggle to define futility, the state of Texas took a unique approach to this issue. The Texas Advance Directives Act (TADA) was passed in 1999 and amended in 2003 with the addition of pediatric patients. It provides definitions of “irreversible conditions,” “terminal conditions,” and “life-sustaining treatment.” It delineates steps for approaching disagreements, including first having the case reviewed by an ethics or medical committee [53,54]. The results from this meeting are then presented to the patient or patient surrogate, and they can be asked for in written form. If the clinician and the patient or surrogate continue to disagree, the clinician will attempt to transfer the patient to another clinician or facility. Life-sustaining support will be maintained during any pending transfer and up to 10 days following the written decision. If an accepting clinician or facility cannot be found within those 10 days, the clinician and facility are not required to continue support other than artificial nutrition and hydration [55].

TADA also provides legal support to the health care team and guarantees that a patient will be supported for a period of time while alternative treatment facilities are considered [56,57]. Arguments against TADA claim that the hospital ethics committee is given too much power and is representative of the health care team, not society as a whole. Opponents argue that these cases would best be handled by judicial courts [58].

Virginia Health Care Decisions Act — In 2018, Virginia enacted a statute that authorizes a dispute resolution mechanism like TADA. First, all Virginia hospitals must develop a policy regarding the medical and ethical appropriateness of proposed medical care. The policy must provide for a second opinion, a review by an interdisciplinary review committee, and a written explanation of the committee’s decision. If a clinician determines that treatment requested by a patient or surrogate is medically or ethically inappropriate, the clinician must inform the patient or surrogate and provide a copy of the hospital’s policy.

If the conflict persists, then the hospital must give the patient or surrogate 14 days to attempt transfer to a facility that is willing to provide the disputed treatment. During this period, the current treating facility must both continue the disputed treatment and allow access to the patient’s medical records. If the patient is not transferred by the end of the 14 days, the hospital may stop treatment. Like TADA, artificial nutrition and hydration can be unilaterally withdrawn only when it is physiologically futile. But unlike TADA, Virginia allows court review of the hospital’s decision [59].

Treatment consent laws — While Texas and Virginia empower clinicians, other states have taken an opposite approach and have empowered patients and families. As an example, in 2012 Idaho enacted the Discrimination in Denial of Life-Preserving Treatment Act [60]. This statute mandates that health care “may not be withdrawn or denied if its provision is directed” by the patient’s surrogate. Oklahoma enacted similar legislation, the Nondiscrimination in Treatment Act, in 2013 [61].

DNR consent laws — More recently, states have enacted laws prohibiting clinicians from writing do-not-resuscitate (DNR) orders without consent. For example, between 2017 and 2020, several states enacted versions of a bill called “Simon’s Law.” The law is named after Simon Crosier, a baby with trisomy 18 for whom physicians instituted a DNR order without either his parents’ consent or knowledge. Kansas, Missouri, and Iowa enacted stronger versions of the law, directed at prohibiting unilateral DNR orders for minors [62]. Other states like Arizona, Idaho, and South Dakota enacted weaker versions directed at informing parents but not requiring their consent.

Simon’s Law typically imposes three mandates on providers. First, it requires hospitals and medical facilities with policies about withholding life-sustaining treatment to disclose those policies upon request. Second, the law prohibits physicians from instituting a DNR order on a minor patient unless at least one parent has first been informed both orally and in writing. Third, in its stronger version, Simon’s Law prohibits physicians from instituting a DNR order if either parent refuses consent.

Similarly, but going beyond the pediatric context, in a special 2017 legislative session, Texas enacted a statute that requires that clinicians must obtain patient or surrogate consent for DNR orders, except in one situation [63]. That exception is when in the reasonable medical judgment of the patient’s attending physician, both 1) the patient’s death is imminent, regardless of the provision of cardiopulmonary resuscitation (CPR); and 2) the DNR order is medically appropriate. Still, even in that situation, clinicians may not write a DNR order if that would be “contrary to the directions of a patient who was competent at the time the patient conveyed the directions.”

CLINICAL SCENARIOS IN PALLIATIVE CARE SETTINGS

Prematurity — The limit of viability is defined as the stage of fetal maturity that ensures a reasonable chance of survival. Determining the limit of viability is desirable so that futile, costly, and painful interventions can be avoided in the infant who does not have a chance of survival. However, deciding upon a threshold of viability is challenging because it remains uncertain which extremely preterm infant has a reasonable chance of survival. Factors that affect survival rates for extremely premature infants (gestational age below 26 weeks) include:

●Gestational age, which is the major factor in determining viability (table 1). While it is generally accepted that <22-week estimated gestational age infants are nonviable, there is a “grey zone” in regards to which neonates born at 22 to 23 weeks’ gestation should be aggressively resuscitated. Furthermore, there is a significant margin of error (eg, a few days to two weeks) in the estimation of gestational age using either prenatal or postnatal assessment tools. (See "Periviable birth (limit of viability)", section on 'Limitations of the data'.)

●Other factors that affect survival rate include birth weight at a given gestational age, gender, plurality, and the use of antenatal corticosteroid therapy. (See "Periviable birth (limit of viability)", section on 'Outcomes' and "Periviable birth (limit of viability)", section on 'Prediction tools'.)

Every facility that cares for high-risk pregnancies should have a consensus approach to the management of extremely premature infants that is consistently used by all providers. As of 2020, in the United States, many such facilities are revisiting their policies about the borderline of viability as evidence accumulates that babies born at 22 weeks can survive and thrive [64]. Futility issues surrounding critically ill newborns were specifically addressed by the Presidential Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research [47]. Several professional organizations have issued recommendations in this area, including the American Academy of Pediatrics. (See 'President’s Commission on Biomedical Ethics (1983)' above and "Periviable birth (limit of viability)".)

These issues are discussed in greater detail separately. (See "Periviable birth (limit of viability)" and "Preterm birth: Definitions of prematurity, epidemiology, and risk factors for infant mortality".)

Chromosomal abnormalities — Historically, it has been considered inappropriate to aggressively treat individuals with some specific genetic syndromes and chromosomal abnormalities. The treatment of babies with trisomy 13 and 18 is one example. The available data suggest that these patients have an extremely low likelihood of survival past one year of age, and those that survive beyond one year of age have severe intellectual disability, with a poor quality of life and significant utilization of health care resources [65]. The resuscitation of these infants has, in the past, been generally regarded as futile. Because of the lethality of the diagnosis, clinicians have routinely recommended pregnancy termination (if diagnosed prenatally) or limitation of aggressive interventions after birth, although this noninterventional paradigm is not universal. (See "Congenital cytogenetic abnormalities", section on 'Trisomy 18 syndrome' and "Congenital cytogenetic abnormalities", section on 'Trisomy 13 syndrome'.)

Published data on survival rates have been used to justify the lethality of the disease, and some have begun to argue that this creates a “self-fulfilling prophecy” on these children’s likelihood of survival. Survival beyond infancy is possible, especially in those who receive intensive treatment. When medical intervention is provided, survival rates improve, and families report that their infants have a better quality of life and length of survival than is reported in medical literature [66]. It is clear, today, that treatment, including intensive care and surgery, cannot be considered futile for babies with these trisomies. Instead, clinical decisions must be made on a case-by-case basis, considering the chromosomal anomalies along with other clinical information in order to decide whether treatment is appropriate [67]. This is a very controversial area that is discussed in detail separately. (See "Congenital cytogenetic abnormalities", section on 'Trisomy 18 syndrome' and "Congenital cytogenetic abnormalities", section on 'Trisomy 13 syndrome'.)

Persistent vegetative state — Resuscitation from cardiac arrest or other cardiopulmonary catastrophe may be complicated by hypoxic ischemic brain injury. Obtundation or coma is frequent early on, followed by recovery or evolution to brain death or persistent vegetative state (PVS). On the basis of available clinical data, PVS is judged to be permanent after three months if induced non-traumatically. (See "Hypoxic-ischemic brain injury in adults: Evaluation and prognosis", section on 'Persistent vegetative state'.)

PVS results in complete unawareness of oneself and one’s environment. These patients are incapable of interacting with others; they have no sustained, reproducible, purposeful, or voluntary behavioral responses to stimuli; and they have no evidence of language comprehension or expression. They have the presence of sleep-wake cycles with the preservation of hypothalamic and brainstem function. Because the patient is unaware of themself or their environment, they are not able to suffer or experience pain, despite what may appear as grimaces, crying, or other expressions of discomfort. The perception of pain and suffering is a conscious experience governed by the cerebral cortex, while the expression of pain may be elicited at any level of the nervous system, including the motor/behavioral, endocrinologic, and autonomic responses that may occur as reflexes in the absence of consciousness. (See "Hypoxic-ischemic brain injury in adults: Evaluation and prognosis", section on 'Family counseling'.)

While aggressive treatments for individuals with PVS would not qualify as physiologically futile, as treatment might be able to maintain cardiopulmonary function, they may be considered qualitatively futile because of the lack of benefit in terms of quality of life. Common ethical challenges in these patients revolve around the pursuit of surgically placed feeding tubes and tracheotomies with chronic ventilator support. While the patient may not be considered to either suffer or benefit from any of these aggressive treatments because they are in a PVS, there is clearly an impact on the family, health care team, and society. (See "Ethics in the intensive care unit: Responding to requests for potentially inappropriate therapies in adults" and "Withholding and withdrawing ventilatory support in adults in the intensive care unit" and "Gastrostomy tubes: Uses, patient selection, and efficacy in adults".)

Decision-making regarding withdrawal of various levels of treatment is dependent on accurate prognostication. (See "Hypoxic-ischemic brain injury in adults: Evaluation and prognosis", section on 'Prognosis based on clinical findings'.)

Once this is established, discussions with surrogate decision-makers should focus on the patient’s known, stated preferences or advance directives, ie, whether ongoing life-supporting treatments are consistent with their prior stated wishes. In the absence of known wishes or directives, decision-makers should focus on whether ongoing treatments are consistent with the patient’s known values and goals. In the absence of either stated preferences or known wishes, families must make treatment decisions that are in the patient’s best interest. But these decisions can be very difficult. Some families, given time and information, can agree that pursuit of these treatments is not beneficial. However, some families and patients possess religious or cultural beliefs that support continuation of life even if the patient is unaware [32]. Disagreements between providers and families are common in these emotionally and spiritually charged circumstances. Most disagreements can be resolved, however, when clinicians and families meet and talk and consider the patient’s clinical situation and previously expressed wishes [68].

Brain death (death using neurologic criteria) — Brain death (death by neurologic criteria) is defined as the irreversible cessation of cerebral and brainstem function. There is no respiratory drive, and thus there are no spontaneous breaths regardless of hypercarbia or hypoxemia. There are no responses arising from the brain (including cranial nerve reflexes and motor responses) to stimuli, although spinal reflexes may persist. One is legally dead in the United States when neurologic criteria for death have been demonstrated. However, vital organ function may be maintained in these patients, sometimes for long periods, through artificial ventilatory support. Requirements for the declaration of brain death are discussed in detail elsewhere. (See "Diagnosis of brain death".)

The diagnosis of death using neurologic criteria is an area where, until recently, the meaning of futility was very clear. The concept first came into the literature in the 1950s, and in 1968 the Harvard criteria for brain death were published, putting forth an agreed-upon definition of the term and recommendations for its use [69]. The committee stated: “Death is to be declared and then the respirator turned off. The decision to do this and the responsibility for it are to be taken by the physician-in-charge, in consultation with one or more physicians who have been directly involved in the case. It is unsound and undesirable to force the family to make the decision.” Therefore, since the conception of this definition of death, it was deemed appropriate for clinicians to stop medical intervention following the declaration of death by neurologic criteria. (See "Diagnosis of brain death".)

Since then, declaring death by neurologic criteria has legally been accepted as a means of declaring death (the other being cardiopulmonary death), although New Jersey allows for religious exemptions (ie, nonacceptance of neurologic criteria as a final verdict for death). Three other states require brief reasonable accommodation [70]. While cases have been brought to court challenging a clinician’s ability to declare a patient dead by these criteria and to stop futile medical intervention, courts have generally honored the legal acceptance of brain death.

However, these issues once again came to the forefront with the case of Jahi McMath, a 13-year-old girl who suffered a cardiopulmonary arrest in the wake of surgery for sleep apnea, with anoxic brain injury progressing to death as defined by neurologic criteria. Following this declaration of death, the family requested and was granted a restraining order preventing the hospital from disconnecting Jahi from the ventilator, on religious grounds. Although the diagnosis of death by neurologic criteria was confirmed by consulting clinicians, the court order was extended, providing the McMath family time to find an accepting health care facility in New Jersey. For five years, she subsequently received a surgically placed feeding tube and was maintained through a tracheostomy on a ventilator. Those who supported the family stated that only patients and their surrogates should have the right to decide when medical treatment stops, not medical personnel [71].

Since the Jahi McMath case, other cases (mostly pediatric) have also reached United States and Canadian courts. These cases raise three questions [70]. First, must clinicians obtain family consent before administering brain death testing? Second, must clinicians accommodate religious objections to determination of death on neurologic criteria? Third, do accepted medical criteria for determining brain death comply with the legal standards in the Uniform Determination of Death Act?

Slow or short codes — In the 1970s and 1980s, the practice of unilateral, undisclosed do-not-resuscitate (DNR) orders and “slow codes” was discovered. Clinicians were independently deciding either not to perform cardiopulmonary resuscitation (CPR) or to deliberately provide inadequate or half-hearted CPR (a “slow code”), knowing that it would not be effective [1,72]. This practice has been deemed unethical [73], and most bioethicists and professional medical societies condemn the practice. The American College of Physicians Ethics Manual states, “Because it is deceptive, physicians or nurses should not perform half-hearted resuscitation efforts” [46].

Some distinguish three related concepts: “slow codes,” which refer to how timely CPR is administered, “show codes” or “Hollywood codes,” which refer to inadequate or half-hearted CPR, and “short codes,” which refer to the duration of CPR. While slow codes and show codes are condemned, the potential role of a “short code” continues to be debated [74]. When a family member refuses to consent to a DNR order in the instance of a patient with certain imminent mortality, the clinician is faced with either placing a unilateral DNR order or performing futile CPR [74]. If it is known that the CPR will be futile, it may be argued that adjusting the length or extent of the CPR might meet the family’s need (seeing that everything has been done) while limiting the suffering of or battery to the patient [75]. However, this practice is deceptive, and we recommend against it.

Geriatrics — Patients cannot be discriminated against because of their age; therefore, issues surrounding futility determinations should be no different for older adult patients. However, one important issue that arises in older patients is the need to accurately assess decision-making capacity. (See "Assessment of decision-making capacity in adults".)

Patients, particularly older individuals, need to be encouraged to make known their long-term wishes and appoint a surrogate, should they become incapacitated. End-of-life planning should be pursued by clinicians, giving time for patients to consider their wishes and values and to be as informed as possible about different medical interventions that could be pursued or refused. The fact that survival following out-of-home cardiac arrest decreases significantly with age should be disclosed when discussing DNR orders [76,77]. (See "Geriatric health maintenance", section on 'Advance directives and health care proxy' and "Advance care planning and advance directives" and "Medical care in skilled nursing facilities (SNFs) in the United States", section on 'Advance care planning' and "Hospital management of older adults", section on 'Advance directives and goals of care'.)

ADVANCE DIRECTIVES — Deciding upon issues of medical futility can be challenging enough when a patient is able to communicate their own goals and values. (See "Discussing goals of care".)

It is even more challenging when there is speculation upon what the patient would wish. Advance directives (ADs) are the documents a person completes while still in possession of decisional capacity about how treatment decisions should be made on their behalf in the event they lose the capacity to make such decisions. They are legal tools meant to direct treatment decision-making and/or appoint surrogate decision-makers. (See "Advance care planning and advance directives", section on 'Advance directives'.)

Having an AD clearly impacts outcomes. In a systematic review, the completion of an AD was associated with [78]:

●Decreased rate of hospitalization and the chances of dying in the hospital in two out of five studies

●Decreased use of life-sustaining treatment in 10 out of 22 studies

●Increased use of hospice or palliative care in five out of seven studies

However, it is not clear how much impact ADs would have on futility issues should they be more broadly adopted. One aspect is that physicians may not be aware of the presence of an AD [79]. Another problem is that disease progression can be unanticipated and prognosis may change over time. The AD, however, may remain stagnant and not accurately reflect the wishes of the patient at the moment when it is needed [3]. Concerns from clinicians that an AD may not always reflect the patient’s wishes in the exact clinical scenario may limit their utility. (See "Legal aspects in palliative and end-of-life care in the United States" and "Advance care planning and advance directives".)

Physician Orders for Life-Sustaining Treatment (POLST) — POLST forms are completed by health care professionals after discussion with the patient or their surrogate and are immediately actionable and transportable. As medical orders, POLST forms are followed by emergency medical service personnel in times of crisis, and by treating health care professionals once the patient has been transported to a health care facility. In short, seriously ill patients can reduce the chance of unwanted treatments by completing a POLST with their health care professional [80]. (See "Advance care planning and advance directives", section on 'Physician Orders for Life-Sustaining Treatment'.)

Do-not-escalate-treatment orders — Some surrogate decision-makers remain uncomfortable with being asked to withdraw or withhold treatments but may understand that certain limits of therapeutic intervention have been reached. As an example, if a patient is already on high doses of continuous inotropic support, the care provider can inform the family that increasing the medications beyond the current doses is unlikely to provide physiologic benefit and should not be pursued. Reframing the conversation to identifying where limits should be set, rather than asking for agreement to stop treatments altogether, may help to resolve conflicts.

For families, complying with a physician’s recommendation on these limits may feel less burdensome than making the decision to withdraw life-sustaining support. Some families may verbalize that withdrawing support feels like they are actively killing their family member, but placing limits to escalation can diffuse some of that burden. Do-not-escalate-treatment (DNET) orders may prolong the dying process as compared with immediate withdrawal of support, providing the family more time with their loved one. However, it may also increase the possibility of pain or suffering in the dying patient, and therefore the order should also include plans on what will be offered, including what medications will be provided to alleviate any suffering [81].

COMMUNICATION AND RESOLVING CONFLICT — Issues that arise surrounding futility may be indicative of a breakdown in communication and trust. Maintaining excellent communication between health care providers and patients/family members seems paramount in resolving conflict and reaching joint decisions. (See "Communication in the ICU: Holding a meeting with families and caregivers", section on 'Importance of communication'.)

However, good communication alone is not sufficient to avoid conflicts that arise about medically futile or inappropriate/inadvisable therapies, as was shown in the SUPPORT study [82]. The study had a prospective observational (phase I) component that followed outcomes in 4301 adults hospitalized with one of nine life-threatening diagnoses, then a randomized controlled clinical trial (phase II) component in which another 4804 patients hospitalized with one of nine life-threatening illnesses were randomly assigned to either an intervention or a control group. The intervention included providing the patients’ clinicians with frequent prognostic data, obtaining and updating the chart with patient and family care preferences, and providing skilled nursing to organize and help conduct family meetings. The phase I observational study revealed significant shortcomings in communication, as well as knowledge of patients’ wishes regarding cardiopulmonary resuscitation (CPR), and an overall high frequency of aggressive treatments. The hope was that the improved availability of prognostic information and communication with patients and families would improve these shortcomings. However, the investigators found that the intervention failed to improve care or patient outcomes, and it did not reduce use of hospital resources.

Fins and Solomon, in their studies of communication to resolve futility disputes, found that, in addition to good communication, other needed elements were clinicians’ self-awareness, early recognition of impending disputes, psychological insight into the cause of the dispute, and an institutional culture that promotes good communication [83]. Some physicians may not possess these qualities and may benefit from referral to or assistance from colleagues who better exemplify these characteristics; others may benefit from assistance from trained mediators.

Strategies to avoid futility disputes — Prendergast outlined strategies to avoid futility disputes: keep patients and families informed; identify other staff members, such as other physicians or nurses with whom the patient or family has a better rapport who can assist in facilitating good patient relations; promote realistic expectations; strive for accuracy in prognosis; maintain continuity of care; be compassionate and flexible; show firmness about limits; and beware of making decisions based on economic market forces [84].

Goal formulation is also pivotal. As discussed in the definitions of futility, a treatment may be deemed futile if it does not achieve a specific goal. While we may assume that the patient’s or family’s goal is recovery or survival, sometimes it may be to have a few more days or weeks with their loved one, or to die knowing they “fought” until the end, or to ensure that there is minimal pain or suffering. We must inquire about what a patient would like to achieve, provide the realistic framework to shape their expectations, and hopefully arrive at a mutually shared goal of therapy [85]. If an agreed-upon goal can be achieved, futility issues can be minimized.

Unfortunately, there are times when escalation of ethical concerns is unavoidable. Pragmatic steps can be taken to prospectively be prepared for these instances. Most hospitals have policies that specifically deal with these situations.

Hospital futility policies — As listed above, many major professional associations have advised that all hospitals develop formal policies to address futile and potentially inappropriate/inadvisable therapies [43,49]. (See 'Major policy statements' above.)

The Ethics Committee of the Society of Critical Care Medicine (SCCM) listed five criteria for these polices [49]: “a) be disclosed in the public record; b) reflect moral values acceptable to the community; c) not be based exclusively on prognostic scoring systems; d) articulate appellate mechanisms; and e) be recognized by the courts.” It has also been suggested that hospitals, particularly those in close geographical proximity, work together to ensure uniformity and standardization, which will provide clarity for the general public [32]. Some communities, such as San Diego and Houston, have taken it upon themselves to adopt futility policies, as discussed above. These types of community-driven policies can ensure that hospitals and clinicians are acting in a manner that is consistent with the community’s values and that members of the community may be educated on medical futility issues. (See 'Houston Citywide Taskforce on Medical Futility (1996)' above and 'San Diego County Medical Society (2009)' above.)

The nudge/assent model — In some cases, it may also be appropriate for clinicians to make a recommendation about limiting treatment without specifically asking surrogates to explicitly verbalize agreement. This is referred to as the “nudge/assent model.” The patient may veto but is not asked for consent ex ante, only assent. The patient or surrogates’ “informed non-dissent” may be a way of accepting limits of therapeutic intervention without necessarily explicitly endorsing those limits [86,87]. This method may be particularly useful when the surrogate’s guilt keeps them from being able to make a decision but they are aware and able to acknowledge the patient is dying (as long as others make the decision).

Framework for clinicians to use in resolving conflict — Written policies should outline a clear, stepwise approach to negotiating disputes about futile and potentially inappropriate/inadvisable treatments. An algorithmic approach to managing disputed treatment requests in the intensive care unit (ICU), which is also applicable to other health care settings, is outlined in the algorithm (algorithm 2) [36]. Recommended steps for clinicians to use to resolve conflict when a requested treatment is deemed potentially inappropriate are outlined in the table (table 2) [36].

Developing a clear process for deliberation that is transparent to health care professionals, patients, and families may alleviate some of the distress that accompanies these difficult situations and help obtain a viable solution.

Approach to patients who “want everything done” — Communication is paramount in situations where, when asked about setting limits on medical treatment in the face of a severe, life-threatening illness, patients and/or their families respond that they “want everything.” Although physicians might reasonably conclude that the patient is asking for every imaginable treatment no matter how harsh, invasive, or unlikely to be of benefit, a more appropriate response, after acknowledging the request, is to discuss the patient’s underlying treatment values and nonmedical concerns (table 3). The phrase “doing everything” means different things to different people. The clinician must figure out what it means to the particular individuals involved in a particular case. Often, it can mean, “do everything that has a reasonable chance of helping my loved one.” Once the meaning is clarified (table 4), the clinician must provide accurate information about the patient’s illness and prognosis and the treatment options that seem appropriate or inappropriate [88]. With that information, the physician can make more informed recommendations, and patients can make better decisions about what should (and should not) be done to help them achieve their goals. An algorithmic approach to such patients is outlined in the algorithm (algorithm 3). A separate topic review on discussing goals of care is available. (See "Discussing goals of care".)

Time-limited trials — Given the difficulties with prognostication and a patient’s or family’s need for time and experience to feel truly informed and to come to terms with medical realities and grief, some have advocated for the use of time-limited therapeutic trials. This may be beneficial in scenarios where treatment may be considered futile or inappropriate/inadvisable therapies and disagreement exists between parties. The goal of the trial should be determined and agreed upon between the health care team and the patient or surrogate. A specific timeframe should be determined that reflects the clinical situation, the treatments being proposed, and the nature of the underlying illness [89]. Sometimes, a written document specifying the nature of the time-limited trial, and the endpoints to be used in determining whether the trial has or has not been successful, can be useful. During the trial, all parties will have the opportunity to get a better sense of prognosis and efficacy of treatment and to adjust to illness realities. If the goal of therapy is not achieved, all parties may be more confident in not pursuing further interventions [90,91].

Palliative care consultations — For non-palliative-care-trained clinicians struggling with communication and consensus with patients and families around end-of-life issues, obtaining a palliative care consult can provide benefit to communication. Individuals trained specifically in death and dying can provide a different perspective to families and help families and patients understand that transition from pursuing aggressive medical interventions to pursuing control of pain and suffering is not “giving up” on a patient but is, instead, providing the best treatment when cure is not possible [92]. (See "Benefits, services, and models of subspecialty palliative care" and "Palliative care: Issues in the intensive care unit in adults".)

Ethics consultations — Ethics committees and consultations have been discussed above, both as components of the Texas Advance Directives Act (TADA) and as part of the American Medical Association’s (AMA) recommendations for futility negotiations. Many articles support the involvement of ethics consultations prior to escalation of issues to that level. Asking ethics consultants to get involved earlier, particularly ones trained in conflict mediation, may help gain resolution earlier while maintaining positive relationships between all parties [93]. (See "Ethics in the intensive care unit: Responding to requests for potentially inappropriate therapies in adults", section on 'Six-step process to conflict resolution'.)

SUMMARY AND RECOMMENDATIONS

●Controversies about medical futility arise when doctors and patients (or family members) disagree about whether a particular treatment is futile or inappropriate/inadvisable in a particular circumstance. Uncertainty about the appropriateness or benefit of interventions is a product of different perspectives on what constitutes benefit to patients and how individuals (patients/families as well as clinicians) can vary dramatically in how they define value as related to life. (See 'Introduction' above.)

While the legal and moral acceptability of treatment limitation is well established, clarity in establishing goals of care, timing of the transition from cure to palliation, and communication of specific decisions to withhold further aggressive interventions remain problematic for both patients and clinicians. (See 'Autonomy and medical futility' above.)

●In 2015, five critical care societies, the American Thoracic Society (ATS), the Association of Critical-Care Nurses (AACN), the American College of Chest Physicians (ACCP), the European Society of Intensive Care Medicine (ESICM), and the Society of Critical Care Medicine (SCCM), officially endorsed new and specific terminology that should be used when discussing and categorizing treatments of questionable benefit [36] (see 'Definitions' above):

•The five-society statement maintains that the term “futile” should be used only in the rare circumstance that an intervention definitely cannot accomplish the intended physiological goal. The statement provides the example of performing cardiopulmonary resuscitation (CPR) on a patient with a ruptured myocardium. (See 'Physiological futility' above.)

•The term “inappropriate” should be used, rather than “futile,” to describe the more common situation in which treatments have at least some chance of accomplishing the effect sought by the patient, but clinicians believe that competing ethical considerations justify not providing them. In contrast to “futile” treatment, disagreements over a “potentially inappropriate” treatment are value based. (See 'Futile versus potentially inappropriate treatment' above.)

•“Legally proscribed” treatments are those that are prohibited by applicable laws, judicial precedent, or widely accepted public policies. The statement offers the example of circumventing organ allocation policies.

•“Legally discretionary” treatments are those for which there are specific laws, judicial precedent, or policies that give physicians permission to refuse to administer them.

We agree with these definitions.

●One of the major barriers to the definition and acceptance of the concept of medical futility and potentially inappropriate treatment is the uncertainty that often exists when prognosis is delivered to patients and their families. A high degree of certainty is required for the application of futility. (See 'The complexity of prognostication' above.)

●Having an advance directive (AD) clearly decreases the use of life-sustaining treatment and increases the use of hospice or palliative care. However, concerns from clinicians that an AD may not always reflect the patient’s wishes in the exact clinical scenario may limit their utility. (See 'Advance directives' above.)

Seriously ill patients can materially reduce the chance of unwanted treatments by completing a Physician Orders for Life-Sustaining treatments (POLST) form. POLST forms are immediately actionable and transportable. (See 'Physician Orders for Life-Sustaining Treatment (POLST)' above.)

●Most disagreements about futile or potentially inappropriate treatments can be resolved with time and continued efforts to come to consensus. Some surrogate decision-makers remain uncomfortable with being asked to withdraw or withhold treatments but may understand that certain limits of therapeutic intervention have been reached. Reframing the conversation to identifying where limits should be set (ie, do-not-escalate-treatment [DNET] orders), rather than asking for agreement to stop treatments altogether, may help to resolve conflicts. (See 'Do-not-escalate-treatment orders' above.)

●In some cases, it may also be appropriate for clinicians to make a recommendation about limiting treatment without specifically asking surrogates to explicitly verbalize agreement. The surrogates’ “non-dissent” may be a way of accepting limits of therapeutic intervention without necessarily explicitly endorsing those limits. (See 'The nudge/assent model' above.)

●We specifically recommend against unilateral, undisclosed do-not-resuscitate (DNR) orders and the use of “slow codes” (ie, deliberately inadequate or half-hearted CPR). (See 'Slow or short codes' above.)

●Policy statements on dealing with futile and potentially inappropriate therapies are available from several groups. (See 'Major policy statements' above.)

In addition, a number of states have passed laws mandating consent for withdrawal of treatment and DNR orders. (See 'Notable state laws' above.)

●An algorithmic approach to managing disputed treatment requests in the intensive care unit (ICU), as recommended by five joint critical care societies, is provided in the algorithm (algorithm 2); this approach may also apply to other health care settings. (See 'ATS/AACN/ACCP/ESICM/SCCM (2015)' above.)

A framework approach for clinicians to use in resolving conflict regarding potentially inappropriate treatments, also as suggested by the five joint critical care societies, is provided in the table (table 2). (See 'Framework for clinicians to use in resolving conflict' above.)

●A time-limited therapeutic trial may be beneficial in scenarios where treatment may be considered futile or inappropriate/inadvisable and disagreement exists between parties. The goal of the trial should be determined and agreed upon between the health care team and the patient or surrogate prior to the trial. (See 'Time-limited trials' above.)

●For non-palliative-care-trained clinicians struggling with communication and consensus with patients and families around end-of life-issues, obtaining a palliative care consult can provide benefit. (See 'Palliative care consultations' above.)

●If conflicts over the aggressiveness of care cannot be resolved, then the courts can be a last resort. Ethics committees and consultations should be sought prior to escalation of issues to that level. Asking ethics consultants to get involved earlier, particularly ones trained in conflict mediation, may help gain resolution earlier while maintaining positive relationships between all parties. (See 'Ethics consultations' above.)