Disease Management, Advance Directives, and End-of-Life Care in Heart Failure Education and Counseling

Advance Directives and End-of-Life Care

HF has a worse prognosis than many common cancers,155 and premature death from progressive decompensated HF or sudden cardiac death (SCD) is frequent. Recent advances in HF treatment have resulted in substantial reductions in annual mortality from these modes of death. Nevertheless, the mortality rate in HF remains high, making advance directives and end-of-life care important issues for patients with this condition.

Hospice services or other end-of-life care should only be implemented after full and appropriate application of evidence-based pharmacologic and cardiac device therapies (ie, cardiac resynchronization therapy [CRT]), unless documentation of intolerance or contraindication to such treatments is present. For critically ill patients, clinicians should acknowledge to the patient and their family the potentially life-threatening nature of their condition, and supportive care should be implemented as indicated. In most cases, adequate time (weeks to months) must be given to allow medical therapies to exert a beneficial therapeutic effect. In addition, issues such as access to care, adherence to medications and other self care behaviors, and knowledge about HF must be addressed. End-of-life care most often includes continuing HF therapies, which may effectively ease symptoms and stabilize or improve quality of life. Failure to implement evidence-based therapies or to comply with quality measures for HF is associated with higher patient mortality.156 In one hospital system, HF patients with do-not-resuscitate (DNR) orders were less likely to receive quality measures including ACE-inhibitor/angiotensin receptor blocker (ARB) use, non-pharmacologic counseling, or assessment of left ventricular (LV) function as compared to patients without DNR orders, after adjustment for other factors.157 Discontinuation of medications at the end of life may be considered when taking them becomes burdensome (eg, the patient has difficulty swallowing) or if they do not impact symptoms (eg, statins). Drugs should be discontinued one at a time so that worsening symptoms can be correctly attributed to discontinuation of a specific drug.

A discussion about HF course and prognosis should be conducted with all patients to the extent that they are willing to participate in such a conversation. Several tools, including the Enhanced Feedback for Effective Cardiac Treatment (EFFECT) score158 and the Seattle HF Score,159 may help clinicians identify the patient's general prognosis. Secondary analyses of registries and trials have identified several common predictors of death, including low sodium, elevated BUN or serum creatinine, advanced age, and low hemoglobin.160-164 Data generated from these scores should be presented to the patient and family as an estimated range of times, with the caveat that patients may live longer or shorter than expected.

A discussion of prognosis should acknowledge the fact that death in HF may occur suddenly and unexpectedly in patients who are otherwise well compensated, so patients should be educated on the available strategies to reduce the risk of SCD. In the Candesartan in Heart Failure: Assessment of Reduction in Mortality and Morbidity (CHARM) Preserved study that enrolled patients with NYHA Class II-II HF and preserved left ventricular ejection fraction (LVEF), 29% of subjects died of SCD while 20% died of progressive HF. Non-cardiovascular causes accounted for 30% of deaths, and the remaining 21% died of other cardiovascular causes.165 In CHARM-Added, 40% of patients with NYHA Class II-III HF and reduced LVEF died of SCD, 28% died of progressive HF, and 20% died of non-cardiovascular causes. However, in subjects with a reduced LVEF, mortality was double that observed in those with a preserved LVEF so the absolute mortality from SCD was much higher in patients with reduced LVEF compared to those with preserved LVEF.166 In the Metoprolol CR/XL Randomised Intervention Trial in-Congestive Heart Failure (MERIT-HF), the mortality in subjects randomized to the metoprolol CR/XL group was 5.3%, 8.1%, and 16.7% per patient-year of follow-up for NYHA class II, III, and IV, respectively. The deaths due to progressive HF increased from 12% to 26% to 56% in NYHA Class II, III, and IV, respectively, while SCDs declined from 64% of deaths in NYHA Class II subjects to 59% in class III and to 33% in Class IV subjects.167 The absolute number of SCDs was 6, 11, and 19 per 100 subjects in NYHA classes II, III, and IV. In the Comparison of Medical Therapy, Pacing, and Defibrillation in Chronic Heart Failure (COMPANION) Trial, NYHA Class IV CRT subjects experienced about 15 SCDs per hundred.168 Thus, while the percentage of deaths due to SCD is lower in patients with NYHA Class IV HF, the absolute number of SCDs is quite large. This general concept of a decline in the absolute percentage of SCD but a high absolute number has been confirmed recently.169

Patients and families may want more specific information about their likely course. A greater proportion of patients with less severe symptoms tend to die from SCD, whereas death is attributed to progressive HF more often in patients with more symptomatic disease.167 In young patients, progressive HF death typically is heralded by a period of severe symptoms, frequent hospitalizations, and obvious, unremitting clinical deterioration. Some individuals, especially, older, frail individuals, may have severe fatigue as a sign of progressive HF. Shortness of breath can be well managed for most patients and should not be presented as inevitable.

Discussion of end-of-life care can occur when the patient has progressed to a state of severe, refractory HF. This discussion is easier if the patient and family are aware early in the course of HF care that HF leads to death, often over a period of many years. Early in the course of care, clinicians should discuss dying from HF with patients. This conversation should include a discussion about the effectiveness of medication management, the use of CRT (if indicated) to modify the course of illness, and the potential risks and benefits of implantable defibrillators to reduce the chance of a SCD. Some data suggest that patients prefer to be informed about issues related to their disease and its prognosis when they are relatively well.170 In addition, patients want to be aware of the prognosis of their condition, but they desire that this information be balanced with hope that they have the potential to respond to available therapeutic measures.170 To optimize interventions and approaches to care, it is important to understand whether a patient would want an attempt at resuscitation or natural death.171

In considering these issues, it is important to understand the distinction between advance directives and end-of-life care. Advance directives are decisions or legal documents made or created by individuals and shared with loved ones and health care providers that identify desired or undesired treatments if an individual becomes incapacitated and incapable of making decisions about care. Examples of legal advance directives are shown in Table 8.4. All patients with HF should be encouraged to have advance directives in place before the end-of-life is imminent and should designate proxy decision makers in the event they are not able to speak for themselves. The use of advance directives has not been well-studied in patients with HF. End-of-life care refers to care designed to provide symptom relief, comfort, and support for patients and their families when optimal treatments have failed to halt progression of the illness or relieve symptoms and the likelihood is high that death is imminent within the coming weeks to months.

Recommendations

8.11

It is recommended that patient and family or caregiver discussions about quality of life and prognosis be included in the disease management of HF. (Strength of Evidence = C)

8.12

It is recommended that:

  • Seriously ill patients with HF and their families be educated to understand that patients with HF are at high risk of death, even while aggressive efforts are made to prolong life.
  • Patients with HF be made aware that HF is potentially life-limiting, but that pharmacologic and device therapies and self-management can prolong life. In most cases, chronic HF pharmacologic and device therapies should be optimized as indicated before identifying that patients are near end-of-life.
  • Identification of end-of-life in a patient should be made in collaboration with clinicians experienced in the care of patients with HF when possible.
  • End-of-life management should be coordinated with the patient's primary care physician.
  • As often as possible, discussions regarding end-of-life care should be initiated while the patient is still capable of participating in decision-making. (Strength of Evidence = C)
8.13

End-of-life care should be considered in patients who have advanced, persistent HF with symptoms at rest despite repeated attempts to optimize pharmacologic, cardiac device, and other therapies, as evidenced by 1 or more of the following:

  • HF hospitalization172,173 (Strength of Evidence = B)
  • Chronic poor quality of life with minimal or no ability to accomplish activities of daily living (Strength of Evidence = C)
  • Need for continuous intravenous inotropic therapy support174,175 (Strength of Evidence = B)

Background

Identification of Patients Who Are Near the End of Life. Some patients with HF exhibit episodes of frequent decompensation requiring hospitalization. Although a roller-coaster pattern of decompensation may occur in advanced HF despite aggressive therapy, in some patients, events will be related to reversible causes, such as dietary sodium or fluid indiscretion, medication nonadherence, contraindicated medications, inadequate medical therapy, new onset atrial fibrillation, or acceleration of ventricular rate in patients in chronic atrial fibrillation.

After identifiable causes of decompensation are eliminated and proven therapies have been aggressively applied, end-of-life care should be considered if patients still experience a marked decline in functional ability and quality of life. Typically, these patients have severe LV systolic dysfunction or severe restrictive diastolic dysfunction and evidence of marked cardiac decompensation. They often have significant renal insufficiency and hypotension that may limit the application of effective therapy. This clinical picture persists despite intensive attempts at pharmacologic management both in inpatient and outpatient settings. Elderly patients with HF may also approach the end of life in the context of progressive frailty or with other significant medical problems. HF in this population is often accompanied by cognitive problems and increasing need for assistance with care.

Recognition of End-stage HF. Patients with HF and their caregivers often do not appreciate the life-limiting nature of their illness.176 HF is a chronic disorder and often progresses to death.177 Despite the concern that a discussion of prognosis might be discouraging and have a negative impact on psychological and physical morbidity,178 discussions about dying should occur in the course of care for patients with HF. These conversations should be coupled with discussions on ways patients can manage HF (i.e., through self-care maintenance and management behaviors (see sections 8.1-8.6).179 Early in the course of illness and in the context of a discussion of the importance of self management, the clinician should acknowledge that HF is rarely curable and will ultimately lead to death. This information should be partnered with encouragement that good quality of life can often be achieved with evidence-based pharmacologic and device therapies. Patients should be educated on the possibility of SCD and available strategies to reduce the risk of this event, and clinicians should assess the patients' wishes regarding the implementation of these strategies (i.e. placement of implantable cardioverter defibrillator (ICD), cardiopulmonary resuscitation [CPR]). When patients develop refractory HF despite aggressive medical therapy, clinicians should discuss their worsened prognosis and options for care. It is reasonable to have discussions about the possibility of death with the patient and their family during any period of severe instability (i.e. during hospitalizations for HF, and/or in the setting of hemodynamic compromise or hypoxemia). Recent evidence has shown that the majority of family members or surrogate decision makers of critically ill patients wanted physicians to accurately inform them of the patient's prognosis.180

Decision-Making at End of Life. Experience from HF centers caring for patients dying from progressive HF suggests that decisions about termination of life-prolonging therapy are usually made by the patient and family after discussions with their health care provider about prognosis and goals, although such open discussions can be challenging for patients and clinicians.171,181,182 Decisions related to end-of-life care may be made during periods of relative compensation; however, clinicians should be prepared to guide patients and families in decision making in situations of decompensation as well. Because patients with HF approaching end of life may have periods of confusion, delirium, somnolence, or inattention and need someone else to make decisions, a designated surrogate decision maker or durable power of attorney for health care is especially important at this time.

Recommendations

8.14

It is recommended that end-of-life care strategies be individualized and include core HF pharmacologic therapies, effective symptom management and comfort measures, while avoiding unnecessary testing. New life-prolonging interventions should be discussed with patients and care-givers with careful discussion of whether they are likely to improve symptoms. (Strength of Evidence = C)

8.15

It is recommended that a specific discussion about resuscitation be held in the context of planning for overall care and for emergencies with all patients with HF. The possibility of SCD for patients with HF should be acknowledged. Specific plans to reduce SCD (for example with an ICD) or to allow natural death should be based on the individual patient's risks and preferences for an attempt at resuscitation with specific discussion of risks and benefits of inactivating the ICD. Preferences for attempts at resuscitation and plans for approach to care should be readdressed at turning points in the patient's course or if potentially life-prolonging interventions are considered. (Strength of Evidence = C)

8.16

It is recommended that, as part of end-of-life care, patients and their families/caregivers have a plan to manage a sudden decompensation, death, or progressive decline. Inactivation of an implantable defibrillation device should be discussed in the context of allowing natural death at end of life. A process for deactivating defibrillators should be clarified in all settings in which patients with HF receive care. (Strength of Evidence = C)

8.17

Patients with HF receiving end-of-life care should be considered for enrollment in hospice that can be delivered in the home, a nursing home, or a special hospice unit. (Strength of Evidence = C)

Background

Reassessment of Decision-Making. The Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment (SUPPORT) evaluated 936 patients with severe HF and showed a 19% change in resuscitation decisions over a 2-month period among patients who survived their enrollment hospitalization.183 In 50% of the cases, the physician's perception of the patient's preference was inaccurate. An analysis of the SUPPORT and Hospitalized Elderly Longitudinal Project (HELP) showed that a HF diagnosis was an independent predictor of attempted resuscitation, whereas other severe diseases (chronic obstructive pulmonary disease, cirrhosis, coma, colon or lung cancer, or multi-system organ failure) were independent predictors of not receiving resuscitation.184

End-of-Life Care. The goals of end-of-life care are to meet patients' and their families' goals for length and quality of life to the extent possible, manage debilitating symptoms, and provide support for emotional, social, and spiritual distress. Bereavement support should be provided during the patient's illness and for the family after the patient's death. In most cases, evidence-based HF care or even aggressive treatment should be continued to meet these goals. In some cases, time-limited trials of aggressive treatment can be used to help providers and patients understand whether or not a patient may be responsive to such treatment. Hospitalization for management of congestion or a trial of intravenous treatment at home or under hospice care to reduce symptoms are examples of appropriate end-of-life medical care for HF.

Symptom Management. Patient-centered care dictates that symptoms be managed to the level desired by patients and families when possible. Inadequate symptom relief is distressing to patients and their families and negatively affects quality of life, as well as the ability of patients to complete life closure tasks.185

Since some therapies to manage HF symptoms may influence duration of survival, it is important for physicians to fully assess a patient's desires regarding the balance of symptom management and prolongation of life. In a time-trade-off study of 287 patients with advanced HF in the ESCAPE (Evaluation Study of Congestive Heart Failure and Pulmonary Artery Catheter Effectiveness) study, 49% indicated they would not be willing to trade survival time to feel better. In the subgroup of patients who survived for <105 days, 31% reported they would trade almost all of their remaining days to feel well for the time they had left. In contrast, only 6% of patients who survived for 180 days were willing to trade most of their survival days to feel well (P=0.0015).186

In an analysis of 91 HF patients (48 with NYHA class II symptoms and 43 with NYHA class IV symptoms), treatment preferences were assessed in relationship to the time trade off score. In this analysis, two specific patient groups were identified. Treatments that improved quality of life at the expense of survival were preferred by 55% of the patients, whereas 45% preferred medical management that prolonged survival time.187 The description of end-of-life symptoms that may occur with medical management (severe shortness of breath or gasping for air) may have influenced the outcome of this study. Interestingly, these preferences correlated poorly with quality of life, symptom, and overall health scores.187 This evidence suggests that for some patients, treatment preferences can be decided early in the course of the illness.

As compared to patients with other manifestations of coronary heart disease, patients with HF have a worse health status at the end of life and tend to have more hospitalizations, and more commonly rate their health as fair or poor, have activity restrictions, and report symptoms.188 An array of symptoms are seen among patients with end stage HF,189-191 including pain (78%), dyspnea (61%), low mood (59%), sleeplessness (45%), loss of appetite (43%), confusion (40%), constipation (37%), nausea and vomiting (32%), anxiety (30%), and urinary incontinence (29%).192 Families rated pain, dyspnea, low mood, anxiety, urinary incontinence, and confusion as being the most distressing to patients in the last year of their lives. In the SUPPORT study of patients admitted for acute HF who were considered to have end-stage HF, the three most common symptoms reported by family members in the last 6 months of life were dyspnea, pain, and confusion.189 The percentage of patients experiencing escalating rates of perceived severe dyspnea and pain increased significantly as death approached. In the last 3 days of life, 63% of all patients with HF experienced severe dyspnea. Current experience may differ as a result of advancements in medical therapies since the early 1990s when SUPPORT was conducted, but nonetheless, adequate symptom management is a high priority for patients and their families. In one study, during the last week of life, 70% of patients' families rated their quality of life as poor to fair. Increases in emotional symptoms, such as anxiety and depression, were reported by families during the 3 days before death. Other studies have confirmed these findings.193 Analysis of medical records of 80 patients diagnosed with HF revealed that the most common symptoms experienced in the last 6 months of life were breathlessness (88%), followed by pain (75%) and fatigue (69%). Investigators concluded that end-stage patients with HF experience similar symptoms to end-stage cancer patients. In two studies of HF programs, the course to death for patients with advanced HF was frequently progressive metabolic disarray and decreased consciousness.194,195

One of the most important components of end-of-life care is good listening and open communication, with particular attention to patients' concerns about management of symptoms, attitudes about dying,177 ease to access of services, and emotional and spiritual concerns. Symptoms should be treated to the level of comfort desired by the patient and family, recognizing that in some situations a compromise is required between alertness and decreased symptoms.

As previously discussed, pain is present in two-thirds or more of patients with HF, and it is common for patients to have multiple sites of pain. Non-steroidal anti-inflammatory drugs should be avoided in patients with HF, so interventions for arthritis pain should include local steroid injections, low-dose opioids, and physical therapy. Pain related to ischemia is most effectively treated with nitrates and opioids. Dyspnea can be managed with diuretics and opioids. Morphine is inexpensive and effective for dyspnea, but active metabolites can accumulate in patients with end-stage disease because of poor renal function. This accumulation may lead to myoclonus, agitation and delirium. In addition, recent evidence from the Acute Decompensated Heart Failure National Registry (ADHERE) suggests that morphine use may be associated with higher risk of mortality, even after adjustment for other important risk factors.196 Clinicians should be vigilant for confusion or delirium and attempt to avoid medications or other insults that precipitate or worsen delirium. Antidepressants, sleep aids, sedatives, and complementary therapies can worsen confusion, particularly if pharmacokinetic or pharmacodynamic changes related to HF (i.e. poor hepatic or renal perfusion) or age are not considered. Dose adjustments or extended dosing intervals may be needed to optimize the benefits from these drugs while minimizing cognitive side effects. Gastrointestinal problems, such as loss of appetite, constipation, nausea/vomiting, and fecal incontinence can be managed with diet modifications, appetite enhancers, laxatives, or other medications. Urinary incontinence is often related to diuretic use and weakness of the urinary sphincter. It may be addressed with a change in the timing of diuretic doses, a urinary catheter, pads or incontinence underwear. Management of fatigue and activity intolerance may require lifestyle modifications including periods of rest with feet elevated or rescheduling of activities to take advantage of changing energy levels. Energy conservation techniques, such as breathing retraining, spacing activities, meditating, or using assistive devices, may be helpful. A low-level exercise program may have both physical and psychological benefits. Home health aides and homemakers can be very helpful in assisting patients to manage activities of daily living and thus conserve energy.

Use of Continuous Intravenous Inotropic or Vasoactive Support and End-of-Life Care. Patients undergoing end-of-life care may respond to continuous intravenous inotropic agents with temporary symptomatic improvement. Utilization of inotropic agents must be undertaken with the understanding that these drugs likely will reduce survival due to an increase in SCD. Health care providers skilled in HF management may use intravenous inotropic infusions for end-of-life care when oral HF pharmacologic therapies fail to stabilize symptoms. The use of inotropic therapy in this population is highly variable.197 Patients should be informed about the potential risks of inotropic therapy including proarrhythmia and other adverse clinical outcomes such as sepsis due to chronic indwelling venous catheters that might reduce life expectancy despite a possible period of symptomatic improvement.

Periodic reevaluation of continuous intravenous inotropic support is mandatory, because the patient's response to treatment may diminish over time, or the patient may decide that the quality of life gained is offset by the intensity of therapy required. Continuous intravenous inotropic therapies must not be considered as acceptable alternatives to core evidence-based HF pharmacologic and cardiac device based treatments. They should be applied only after careful attempts to manage patients with evidence-based drug and cardiac device therapies. Hospices vary in their provision of intravenous and other therapies, based on agency size and staff education.

Referral to Hospice. Data from 2000 indicate that 8% of all hospice patients have a diagnosis of HF.198 A survey of hospice centers published in 2005 reported similar findings, with an average of 9% of patients under hospice care having a primary diagnosis of HF.199 Patients with cancer are routinely referred for hospice care and comprise the majority of hospice patients nationally. Only 1.6% of the 182,898 hospitalization episodes from 2001-2005 in ADHERE resulted in hospice referral; however, ADHERE enrollment was not limited to an end-stage population.200 Hospices vary in their expertise and practice caring for patients with HF.201 Clinicians providing care to patients with HF should partner with local hospice agencies to create a plan of care to meet patients' needs. For select patients, referral for hospice services may be an appropriate method of providing palliation when symptoms are refractory, quality of life is poor, and there is functional decline.202 The Medicare hospice benefit was developed so that individuals could choose such supportive care and still receive Medicare funding.

To be eligible for the hospice benefit, the patient's physician and hospice medical director must certify that the patient has a likely life expectancy of 6 months or less, and the patient must consent to receive hospice in lieu of Medicare A-reimbursed care for his or her terminal illness. This agreement does not preclude other treatments for illnesses or injuries not related to HF, nor does it necessitate abandonment of appropriate HF medical therapy. Patients may withdraw from the hospice program and reenroll at a later date with no penalty. Hospice care is not limited to 6 months; however, the patient's prognosis must be identified as approximately 6 months at specified certification periods (the first two periods are 90 days, followed by an unlimited number of 60 day certifications).

The Medicare hospice benefit includes coverage for all medications and treatments associated with the hospice diagnosis, symptom management, homemaker and home health aide assistance, and chaplain and bereavement support for patients and families. Nursing care, medical supplies and appliances, therapy services and a wide variety of other professional support services necessary to improve quality of life are covered. Physician oversight of care may be provided by the hospice medical director or by a physician of the patient's choice.

There are four levels of hospice care. In the United States, 70% of hospice care is delivered in patients' homes or place of residence (including nursing homes).203 Respite care up to 5 days per certification period is generally provided in nursing homes under contract. "General inpatient care" is provided to manage symptoms or provide services that cannot be provided in other settings--in either a hospital or nursing home. "Continuous care" provides 8-24 hours of licensed nursing care in the home for brief periods of time to manage complex problems or provide caregiver education. Hospice care is reimbursed by Medicare and most insurances at a specified daily rate, regardless of the medications, treatments or services provided.

Advance Directives and Risk of Sudden Death. SCD in a patient with compensated HF is a relatively common cause of death. Most SCDs occur outside the hospital, often at home or in the presence of a family member. Families commonly express the need to know how to respond in a cardiac emergency and report that this learning need is often unmet by health care professionals.204 Patients report wanting their families to know what to do in an emergency.204,205 A discussion with patients and families about the patient's wishes regarding resuscitation can include information about the effectiveness of resuscitation and its sequelae.206 Patients' wishes need to be clear to all healthcare providers and family care givers, and they should be documented in a written advance directive when possible. Discussions regarding patient and family preferences should be undertaken before an acute crisis develops.

Information on Cardiopulmonary Resuscitation. When patients and families make the choice to attempt resuscitation, family members can be advised how to obtain CPR training. Many clinicians express concern over the ability of families of high-risk cardiac patients to learn CPR, and the potential guilt they might feel if resuscitation fails. In fact, the majority of family members of patients at risk for SCD can successfully learn CPR, are not burdened by responsibility or guilt, and use CPR appropriately when the occasion arises.207,208

Choice to Allow Natural Death. When patients and families decide against resuscitation attempts, they need to be told what to do when death occurs outside the hospital. Without prior information, most people call 911 or a similar emergency medical system number. In some states, this action may end in unwanted resuscitation and prolonged life support efforts. Many states have statutes directing emergency response personnel to comply with written physician orders for resuscitation (such as the Physician Order for Life Sustaining Treatment originally developed in Oregon).209 A better option is to have a family member call a health care provider who knows the patient, has been informed of the patient's preference to not attempt resuscitation, and is willing to certify the cause of death.

As more patients with HF have ICDs implanted, it is important to plan what actions to take when patients are near the end of life. Defibrillation devices can be inactivated for those end-stage patients who do not desire resuscitation. A clear process for defibrillator deactivation should be identified to facilitate this step in advance of imminent death. A recent survey revealed that roughly 60% of cardiologists, 88% of geriatricians, and 95% of family physicians or internists had 2 or fewer conversations with patients and families about deactivation of implanted defibrillators.210 Kelley et al reported the results of a similar survey designed to assess physician management practices regarding ICD use near the end of life.181 Only 13% of the physician respondents accepted primary responsibility for discussions regarding device inactivation, 10% responded that another doctor should discuss, and 7% felt the patient or family should bring it up first.181 These data suggest that communication among patients and physicians regarding ICD therapy at the end of life is needed.

Table 8.4: Examples of Legal Advance Directives

Legal Advance Directive Description
Living Will This document uses standard language in the patient's state of residence, identifying whether specific or general life-prolonging interventions should be initiated (or continued) in the face of imminent death. Some states require 2 physicians to certify that the patient has a "terminal illness" for a living will to be enacted.
Durable Power of Attorney for Health Care (DPOA/HC) This document designates one or more individuals to make health care decisions on behalf of the person at a future time if the person is unable to speak independently. While the DPOA/HC does not typically identify specific interventions or approaches to care desired by an individual, patients should be encouraged to make their proxy aware of undesired states and/or generally preferred approaches to care. Patients with HF should be encouraged to appoint a DPOA/HC. Clinicians should discuss with patients with HF general preferences for care, including preferences for an attempt at resuscitation versus allowing natural death.