Conflicts Regarding Decisions to Limit Treatment: Title and subTitle BreakA Differential Diagnosis

One of the most difficult situations physicians face is how to handle conflicts with families over forgoing life-sustaining treatment. Physicians may feel their competence or judgment is not trusted and turn to legal or ethics consultants for help with what they feel is a wrong decision. Families may feel isolated, misunderstood, or abandoned and begin to doubt the health care team's commitment to the patient's well-being. What should be a cooperative effort to achieve goals can turn into an exercise in frustration and distress.

Herein we present a "differential diagnosis" of the "pathophysiology" of family-physician conflicts regarding life-sustaining treatment. By applying this familiar type of analysis, we hope to ensure that physicians carefully consider all of the possible explanations for the disagreement rather than dealing emotionally with the manifest problems. This scientific model, applied to a psychosocial aspect of medical care, may help promote an understanding of the causes of conflicts and stimulate a creative problem-solving approach. Identifying the underlying basis for disagreements may encourage efforts to defuse or prevent them, similar to the way understanding coronary blood flow and atherosclerosis have led to effective treatment and prevention of myocardial infarction.

We focus on family-physician conflicts and not conflicts between physicians and patients for 3 reasons. First, while there is literature on negotiating conflicts with patients,¹ little has been written on family-physician conflicts.² Second, in the intensive care unit, where many conflicts occur, the patient is often incapacitated because of illness and thus decision making ethically and legally rests with his/her surrogates. Finally, a competent patient's preferences are given much more respect than the judgments of surrogates, opening the door for physicians to question whether a surrogate's decisions reflect the patient's views or best interests.

A man in his 80s with severe Alzheimer dementia and severe congestive heart failure is admitted to the general medicine service for evaluation and treatment of recurrent aspiration pneumonia. A swallowing evaluation reveals aspiration of both liquids and solids. The speech therapist recommends that the patient not be fed. A percutaneous endoscopic gastrostomy (PEG) feeding tube is considered and the gastroenterology service is consulted.

The gastroenterologist feels that a PEG tube is not indicated because of the patient's severely impaired mental status, poor prognosis, and the lack of data showing that a PEG tube decreases aspiration.^{3 - 4} The patient scores poorly on the Mini-Mental State Examination (10/30), is disoriented to person, place, and time, and does not recognize his family.

The patient, a widower, lives with his son and daughter-in-law. The son requests that "everything be done," including placement of a PEG tube. The son also explicitly requests that the patient be resuscitated in the event of cardiopulmonary arrest. The family wants to take the patient home and care for him until he dies. The internal medicine resident who cared for this patient and family during several bouts of pneumonia does not agree with the request for a PEG tube (or resuscitation) and tells the family that this is a case of "medical futility." Negotiations seem to have reached an impasse.

In the differential diagnosis of family-physician conflicts, we distinguish 3 general categories of contributing factors: family features, physician features, and organizational and social features. These factors are often present in combination and may interact.

Families may disagree with clinicians because they do not understand the medical situation. Health care providers should remember that the current episode is typically a small part of a much longer story and that they are merely one source of information among many. For example, if the patient's family in the case report had been told years earlier that he might recover, this might explain their continued insistence that everything be done. Asking families to put in their own words the patient's history, what other health care providers have told them, and their understanding of the clinical situation (eg, "Can you tell me what others have told you is going on with your father?") can help identify gaps in or incorrect information (Table 1). Asking them to explain their choice(s) (eg, "Can you tell me why you have chosen/are leaning toward having the PEG tube placed?") allows the physician to assess reasoning skills and competence. These and similar questions also elucidate the family's beliefs about the causes, consequences, and prognosis of their loved one's illness.⁵ Eliciting the family's beliefs is vital to "diagnose" misunderstandings or discrepancies between the clinical team and the family.

There are a number of reasons why a family member may not understand the medical facts. The family may be psychologically unprepared to hear the patient's diagnosis or prognosis. Denial is the inability to explicitly recognize facts because of unacceptable psychological consequences, such as overwhelming grief or guilt.⁶ Denial may produce displacement—a focus of concern on trivial, but controllable, matters.^{7 - 8} Physicians who misdiagnose denial as inadequate understanding may respond by attempting to state the medical facts more explicitly. Frustration follows when these cognitive interventions do not result in increased understanding or acceptance. Better techniques for managing denial include open-ended listening; nondefensive, neutral responses; silence; and frequent, regular opportunities for the patient or family member to communicate with a consistent health care provider. Especially valuable is reflecting and validating family members' emotions. Physicians' opening statements such as, "It must be very hard for you to see your husband so ill," and "You've been a wonderful caregiver for him for many years" can initiate a frank exchange between the family member and the physician that begins to deal with the grief, guilt, or anger causing denial. These approaches can help physicians identify and explore the cause of the denial rather than respond to its manifestations.

Second, even if frank denial is absent, "bad news" is often poorly processed and imperfectly remembered.⁹ Individuals only remember a portion of information given to them under relatively controlled classroom situations; in stressful environments in which they are given bad news, the amount of information retained may be much less. Repeating key concepts, giving written as well as verbal information, encouraging questions ("I expect you will have questions about what we discussed today. Write them down so you can remember them when we talk tomorrow"), and periodically assessing the family's understanding of the situation ("To make sure we are all on the same page, can you tell me your understanding of the situation?") may improve information transfer and retention.

Third, sometimes a physician's communication style increases misunderstanding. The use of jargon (eg, aspiration and PEG) may confuse families. Terms such as "usually," "most of the time," or "we cannot rule out" are often used by physicians in an attempt to convey the inherently probabilistic and uncertain nature of medicine. Families, hearing these hedges, may attribute a higher or lower rate of success than the physician intends, and thus opt for different care approaches. Differences in semantics, such as "death with dignity," "everything done," "living like a vegetable," or, in the case above, "futile," also can lead to misunderstandings.¹⁰ Communication can be enhanced by using language appropriate to the family's educational level and by frequently assessing their understanding.

Fourth, families obtain information from multiple sources, including television, the Internet, and friends and relatives. For example, an optimistic belief that cardiopulmonary resuscitation will be successful may be derived from the 80% success rate of this procedure on television.¹¹ The family above may have been told by a neighbor about a friend who recovered completely from Alzheimer disease. Furthermore, when numerous clinicians are involved in a patient's care, as in this case, the chances for misunderstanding can increase. The nurse's aide may say that the patient seems much more alert today, leading the family to think that he is recovering; consultants or nurses may disagree with the primary care physicians about the prognosis or the proper course of action. If this multiplicity of voices is contributing to family confusion, it may be helpful to choose one health care professional to serve as the primary communicator or have regularly scheduled family meetings.¹²

Even if they understand and acknowledge the situation, family members may make decisions with which the health care team disagrees. A number of factors may be operative in such a situation. In many cases, the family is grieving, which may make any decision about forgoing life-sustaining treatment difficult. Family members who say "I will not be able to live without him" are commenting about their inability to deal with the relative's death. A family member's grief may make it almost prohibitive to participate in making any decisions that affect the dying process. Decisions that may result in the patient's death and thus increase the grieving are particularly difficult. Supportive, open-ended dialogue is critical to allowing the family to recognize, express, and begin to work through their grief. Some family members benefit when physicians help them identify and validate anticipatory grief, or grieve over a loved one they "lost" to dementia or debility. Other hospital staff, particularly social workers or chaplains, are often skilled in grief counseling and can prove invaluable.¹³

Family members' guilt, often manifest when a relative dies, may contribute to an unwillingness to make particular kinds of decisions. Avoidance of future guilt for decisions currently being faced is an understandable, even honorable, influence. Guilt is a recognizable motivator when family members say, "I cannot do this," or "I will not be able to live with myself." Physicians may unwittingly increase feelings of guilt by asking family members to take responsibility for medical decisions (eg, "Do you want us to put in a PEG tube?"). Philosophically and emotionally, this is the wrong question. As the patient's surrogate, the family member's major role is to represent their loved one's wishes (substituted judgment).^{14 - 15} The appropriate question is, "What do you think your father would want us to do?" One can also more explicitly attend to the family's potential guilt by taking responsibility for medical decisions (eg, "I am not asking you to make these decisions. Your job is tell me about what was important to your dad, then I will recommend what we should do for him.").

Family members' guilt may be tied to concerns regarding abandoning their loved one, especially if this was a fear held by the dying patient.¹⁶ Again the choice of language—"withdrawing care," "comfort measures only," "stopping treatment"—may exacerbate these feelings. A physician can attend to these concerns by asking the family to set positive goals and objectives (eg, maximizing comfort) and then recommending actions to achieve those goals. When appropriate, physicians can (and should) also acknowledge family members' care in respecting their loved one's values and wishes.

Patients and families can be uncertain not just about what treatments are worth pursuing, but which goals or outcomes are desirable. A family may be unsure whether their previously healthy loved one who was just in a motor vehicle collision would find life "worth it" as a mentally impaired double amputee. Moreover, families come to these discussions with their own histories. They may not get along, or may be competing to show they are the "most caring" or "most responsible." Focusing the discussion on what the patient's goals would be (could he/she voice them) can divert attention from conflicts and unite the family in a common purpose. Obtaining information from several family members, a social worker, primary care physician, or member of the clergy familiar with the family dynamics also may help physicians negotiate these difficult situations.

The issue of secondary gain may lead the family to make a decision with which the health care team disagrees. It is important to understand that "conflict of interest," often suspected when conflict arises, is always present to some degree when intimates make decisions for and about each other. A wife's wish to postpone her husband's death because of grief, or a spouse's (possibly subconscious) fear of the responsibility of caring for a dependent partner, are examples of conflicts of interest operating in end-of-life decisions. The question is less whether conflicting interests exist and more what are the degree, influence, and legitimacy of those interests.¹⁷ Wanting to postpone the death of a spouse is very different, for instance, from keeping a patient alive to collect a pension. Physicians should be careful to have evidence to support any belief that family members are acting against the patient's interests to pursue unreasonable or illegitimate interests of their own.¹⁷ In these situations, rumor and innuendo are often rampant and are unfair to families honestly struggling with their loved one's dying.

Finally, there may be a gap between the physicians' values and those of patients or their families.^{2 - 18} A vitalist patient's family may have a very different idea of what makes a life worth living than a physician who prizes cognitive function. Individuals also may have vastly different perceptions of what treatments are reasonable to pursue. These differences in values can be "diagnosed" by attending to the reasoning of the family. For instance, if the family in the case report were asked to explain their choices of resuscitation and PEG tube placement, they might reveal beliefs that the patient held about the sacredness of life.

There are a number of physician characteristics that may contribute to conflicts regarding end-of-life decisions (Table 1). Physicians, like patients, also may be uncomfortable with prognostic uncertainty.¹⁹ This discomfort may lead them to approach limiting treatment decisions in overly hesitant or confident ways. For example, statements such as "He won't leave the hospital alive" or "He has less than 6 months to live" fail to take into account the near-universal uncertainty in prognosticating. Families who have experiences with mistaken prognostication will likely distrust these pronouncements. Conversely, other physicians may constantly hedge their information or recommendations, leading to confusion or false hope.²⁰

Like patients and families, physicians may be uncomfortable discussing death or troubled by the thought of a medical "failure." Spikes and Holland²¹ discuss the troublesome responses that can stem from a physician's need to preserve his/her image as a "powerful healer," able to master any situation. Feeling like he/she has failed the dying patient, a physician may respond by acting defensively, wishing the patient would die (to avoid having to deal with the situation), or overtreating (to ensure that "everything has been done"). Anxiety about one's own mortality may lead physicians to avoid frankly discussing end-of-life issues with families or falsely reassuring them that "everything is okay."²¹

Physicians' attitudes toward the chronically ill also can lead to conflicts. Physicians tend to underestimate chronically ill patients' quality of life, and are more likely than patients or families to think that such patients would choose to forgo life-sustaining treatment.^{22 - 23} In the case presented above, the resident physician's belief that the severely demented are "not really alive" may make it difficult for him to understand the family's desire to take the patient home. Other physician attitudes that can increase the risk of conflict include religious tenets about the sanctity of life, beliefs regarding the proper role of families in end-of-life decision making, difficulty with the radically different value systems of others, or insecurity about one's competence or skill. Insight into one's own limitations and beliefs regarding death and dying may help a physician understand his/her feelings of anger and frustration with certain families and more clearly discuss with the family the areas of disagreement.²⁴

Similarly, knowledge or skill deficits can catalyze family-physician conflicts. Physicians may be unaware of the prognosis or treatment options and misinform the family. They may not understand ethical, legal, or hospital policies surrounding end-of-life care. Misunderstanding regarding the legality of withdrawing ventilatory support or artificial nutrition, for example, may lead a physician to refuse to accede to the family's wishes.²⁵ A lack of training in symptom management²⁶ may lead to inappropriate care. For example, the family of a dyspneic man with end-stage chronic obstructive pulmonary disease may request intubation because they see him in distress and feel it is the only way to relieve his air hunger; a physician who knows alternate ways to relieve symptoms can offer these instead. Moreover, physicians may be ill-trained in interpersonal communication regarding end-of-life decisions, leading to misunderstandings, confusion, and frustrations.^{27 - 28} Recently, both the American Board of Internal Medicine and the American Medical Association designed educational programs to enhance physicians' knowledge and skills in end-of-life care.^{29 - 30}

Physicians, like patients and families, can be overworked, fatigued, frustrated, stressed, and otherwise beset by a number of competing concerns. Physicians in training often have heavy workloads and thus may be poorly motivated to spend additional time with patients or care for more of them, and may be especially intent on making decisions quickly.^{31 - 32} For example, an overextended resident physician may, consciously or subconsciously, feel that his/her scarcest resource, time, is "wasted" on a demented, terminally ill patient. The culture of the hospital, with its prioritization of emergent decisions, technology, and speedy discharges, as well as the lack of reimbursement for conversations with patients and families, may contribute to the emphasis on "high-tech" interventions and the avoidance of time-consuming family conferences.^{33 - 34} Increased awareness by the physician of his/her own emotional status may help, although larger cultural changes in medicine probably will be needed to deal with these issues.^{23 ,26 ,34 - 35}

We have alluded to some of the economic and peer pressures on physicians, the lack of reimbursement for time-consuming conversations with patients and families, and the culture of hospitals. More general and pervasive economic pressures at both the social and organizational levels also may exacerbate family-physician conflicts (Table 1). Hospitals are worried about going out of business, physicians are worried about being "deselected," and there is a pervasive social belief that too much money is being spent on medical care. Thus, when a patient's prognosis seems hopeless, physicians may believe that life-prolonging interventions should be stopped quickly to avoid wasting money.

Patients and families face economic pressures of their own. The economic impact of serious illness on families is large.^{36 - 37} As indicated above, financial concerns may push insured families and patients to prolong inpatient treatment because they do not bear most of the cost of the hospitalization, but would incur large personal and financial costs if the patient were discharged. In other cases, financial concerns about hospital bills may influence families to make decisions that bring about the patient's death quickly.

Hospital policies also may promote conflicts in end-of-life decisions. First, many intensive care and other units have restrictive visiting hours.³⁸ This minimizes contact between families and patients and may impair communication about end-of-life issues. It also prevents families from seeing and understanding what their loved one is going through.³⁸ Second, organizational policies often require that physicians write orders limiting life-support treatments. Nurses, social workers, or pastoral care personnel, however, may have greater opportunities to discuss illness and death with patients, either because of more time with the patient or a different sort of relationship with them.³⁹ Third, organizations and physicians may have legal fears regarding end-of-life decisions,⁴⁰ and these fears may influence decisions about life-sustaining treatment.

In the hospital, death is a routine occurrence to the caregivers, but not to patients and families. Given the emotional impact of decisions surrounding death and dying, conflicts are not surprising but are still disturbing to all parties involved. These factors can decrease trust between the physician and patient or proxy. This impaired trust profoundly influences the ability of families to believe or understand the prognosis and accept physicians' recommendations based on the patient's goals. Families may worry about subsequent abandonment or even mistreatment or experimentation on the part of the physician or the health care system. Physicians, on the other hand, may be angry and frustrated, distrust the family's motives, and worry about litigation.

We have tried to set out a "differential diagnosis" of causes of and influences on conflicts between families and health care workers regarding decisions at the end of life. By considering this list of potential sources of conflict, clinicians can identify more readily and accurately the causes of difficult interactions with families of terminally ill patients regarding decisions to limit treatment. The list is doubtless incomplete and would benefit from empirical studies of the epidemiology and etiology of such conflicts.

The next step is to develop and evaluate interventions designed to "treat" or "prevent" these conflicts. Advance directives, a potentially promising preventable measure, are used by only a minority of hospitalized patients, despite legislation encouraging their use.^{34 ,41} Based on our experience and empirical research, educational interventions are needed to foster higher-quality end-of-life care. Too often we have observed physicians, especially those in training, engaging in long, detailed, and contentious discussions with gravely ill patients and their families over minute technical aspects of care. Unwittingly, these physicians may exacerbate misunderstandings, grief, guilt, or cultural differences between patients and families and physicians. Education may help improve physicians' awareness of their own values and feelings when treating terminally ill patients; increase understanding of the psychology of grief, family dynamics, and cultural sensitivity; and improve physicians' ability to give bad news, deal with emotions, and negotiate end-of-life care goals. Hospitals and health care organizations also should experiment with structural changes, such as changing visiting times or increasing support for family meetings, to minimize conflicts and facilitate acceptable and relatively efficient solutions when they arise. It is hoped that by more accurately identifying the diagnosis, effective treatment (empathic end-of-life care), and prevention, early patient- or proxy-physician discussions and institutional change will follow.