End of Life Care

I. Heart Failure—End of Life Care: What every physician needs to know.

While the course and prognosis for individual patients with heart failure (HF) is difficult to predict, “end of life” in HF patients is characterized either by refractory symptoms, hypotension or evidence of end-organ damage despite optimized HF therapy, or by marked frailty and/or cognitive impairment or other advanced illness in the presence of symptomatic HF. The later description is far more common and applies to the vast majority of elderly patients who die with HF.

II. Diagnostic Confirmation: Are you sure your patient has Heart Failure?

Identifying the end of life

As “end of life” can be difficult to recognize prospectively, a more important step is to identify goals of care for patients. When HF patients have significant other impairments in addition to their HF, or when they have advanced HF that has not benefited from available reasonable interventions, then care should focus on palliation and planning for the end of life. Some patients will want to focus also on extending life, and it is possible to both prepare for the end of life, and work to extend life as long as possible.

A. History Part I: Pattern Recognition:

Heart failure (HF) is a clinical syndrome characterized by symptoms of fatigue and/or shortness of breath or decreased exertional capacity, and evidence of both volume overload and structural heart disease. The clinical assessment should include a careful physical examination to document cardiac pathology, a pulmonary examination to document congestion (or a chest x-ray for the same purpose) and the exclusion of other volume-overload states, such as cor pulmonale, cirrhosis, or nephrotic syndrome.

Cardiac function should be assessed to identify left ventricular ejection fraction, valvular disease, or hypertrophic cardiomyopathy. Typically this assessment is performed by echocardiography; however, cardiac CT, cardiac MRI, and angiography with left ventriculogram also may be used.

Heart failure with preserved or normal ejection fraction is more commonly seen in elderly patients, particularly those with a prior history of hypertension and/or diabetes. The diagnosis of heart failure in young individuals is fairly straight-forward, but in the elderly it is more difficult because fatigue and limited exertion are common in the very old and may not represent HF.

B. History Part 2: Prevalence:

Heart failure (HF) increases in prevalence with increasing age. As more persons are surviving with hypertension and ischemic heart disease, increasing numbers of persons in their 60s and 70s and older develop HF as a consequence of their underlying cardiovascular disease. Changes in the aging heart also increase the prevalence of HF in the elderly who do not have known cardiovascular disease.

Various estimates suggest about 10% of HF patients have AHA-ACC Stage D disease. Approximately 60,000 individuals die annually in the United States with HF as the underlying diagnosis on their death certificate; however, many more(260,000 or more) die from other processes each year with HF listed as a contributor to death. Thus the end of life for HF patients most commonly occurs as the result of frailty, injuries, and other acute and chronic illnesses.

C. History Part 3: Competing diagnoses that can mimic disease Heart Failure.

Advanced heart failure

In elderly patients, physical frailty overlaps with HF. The underlying skeletal myopathy in HF is very similar to if not the same as that in “sarcopenia” of the elderly. Marked limitations in function, characterized by slow gait speed are hallmarks of both HF and frailty. The challenge in the elderly is to identify HF, and this requires evidence of volume overload in the setting of cardiac dysfunction (either systolic or diastolic).

D. Physical Examination Findings.

Heart failure at the end of life can be characterized by hypotension and/or volume overload with or without other evidence of end-organ failure. Based on studies of HF patients managed by HF experts, volume overload and congestion is not usual at the end of life for HF patients managed by HF experts..

Patients with pulmonary congestion identified by rales on auscultation should receive aggressive management to relieve congestion. Slow gait speed, difficulty with transfers, or getting up from a seated position characterize late-stage HF in frail elders.

E. What diagnostic tests should be performed?

As mentioned above, the end of life with HF is characterized by profound frailty. The Short Physical Performance Battery (SPPB), a 4-item test of lower extremity function, is one test that identifies frailty (www.grc.nia.nih.gov/branches/ledb/sppb/index.htm). The SPPB requires 10 to 12 minutes to administer.

In one group of elderly hospitalized HF patients, failure to complete any of the SPPB tasks was associated with a 50%, 6-month mortality. Very slow gait speed (requiring more than 6 seconds to walk 5 m) may also identify frail patients at high risk of death soon.

Patients with advanced HF may be bed-bound or markedly limited in exertion.

Patients nearing the end of life may have declining renal function; however, once the end of life is recognized, diagnostic tests to confirm laboratory abnormalities are not typically performed.

1. What laboratory studies (if any) should be ordered to help establish the diagnosis? How should the results be interpreted?

Progressively declining renal function, particularly with signs of volume overload suggest advanced HF. Thus elevated BUN and CR in combination with elevated BNP or NT-proBNP are seen in advanced disease.

Persistent elevation of BNP can be seen in patients with advanced HF; however, this is not specific and some patients live for years with elevated BNP. Similarly, hyponatremia, probably a marker of neuroendocrine disarray, is a marker of poor prognosis. In elderly patients with HF dying from other conditions, the laboratory abnormalities may not be as significant.

Several online calculators, including the Seattle Heart Failure Model (depts.washington.edu/shfm/about.php) and the Enhanced Feedback for Effective Cardiac Treatment (EFFECT) score (www.ccort.ca/CHFriskmodel.aspx) may help identify patients with advanced heart failure or high likelihood of dying within 30 days or a year. Each of these models has limitations, and at best are based on large populations of patients, so they do not specifically predict outcomes for individual patients.

2. What imaging studies (if any) should be ordered to help establish the diagnosis? How should the results be interpreted?

Once a diagnosis of HF is confirmed, additional imaging does not particularly help in identifying patients nearing the end of life.

III. Management.

Palliative care should be provided throughout the course of HF care. This means that symptoms should be assessed and treated, patient and family stress and burden should be assessed, and support provided to reduce sources of anxiety and burden. Reinforcement about diet (especially sodium), exercise, and self-care should be provided.

Specific counseling should address what to do in an emergency, and how to manage subacute decompensation.

Patients and their families should be made aware that death is possible with HF early in the course of treatment; however, patients might not perceive the life-shortening nature of HF. Simply allowing patients to know that HF shortens lives may help them prepare with a later time when they are nearing the end of life or must choose a course of care.

A. Immediate management.

When a decision point or negative turn in status precipitates a conversation about goals of care, it is appropriate to begin by inquiring about the patient and family’s understanding, hopes and specific goals. The physician should acknowledge these, correct misunderstanding and identify what is reasonable medically.

It may be necessary to identify what fits the category of “hopes for the best” and what plans (“for the worst”) should be made to correspond with potential progressive decline and death. Some patients will welcome a chance to plan for reduced interventions and a focus on symptomatic management, and others will want to continue efforts to prolong life.

While it is never appropriate to provide futile interventions, providing care that potentially prolongs life is appropriate for some patients even when clinicians believe prognosis is poor. It is worth setting goals in this situation and identifying a time period after which goals and the approach to care will be reevaluated.

At the same time, make plans for under what circumstances life-prolonging efforts would be curtailed. Planning permits the family and physicians to reassess in the event of undesired outcomes, and then shift the focus of care to allow death.

Patient or family distress can be significant with chronic HF, as well as at times when status worsens. Providing psychosocial and/or spiritual support can be helpful at these times, but clinicians can help by simply recognizing distress and acknowledging it with patients and families. (“This is stressful for most patients”; or “I can see that you are upset” are examples of language that helps acknowledge distress.)

Symptom management should include: assessment and management of dyspnea, pain, sleep disturbance, anxiety, and fatigue at a minimum. One helpful tool to do this can be the Edmonton Symptom Assessment Scale that uses a 0 to 10 scale (or modified as a 0 to 4 scale) to rank symptom severity for 9 common symptoms. (See Table 1.) Alternatively asking patients to rate their symptom severity (from 1 to 4, with 4 being the worst possible and 1 being not bothersome or not present) is a way to identify what interventions to target for symptom management.

Any way that symptoms are assessed, the main issue is to have an approach to manage distress. In the face of obvious volume overload, diuresis improves the patient’s dyspnea, exertional fatigue and anxiety, and establishing near normal volume status improves symptoms. Aggressive efforts to normalize volume are the mainstay of symptom management; however, symptoms of dyspnea and fatigue persist despite euvolemia for many patients.

Opioids clearly decrease dyspnea and fatigue associated with exertion in patients with NYHA Class II or chronic stable HF. Opioids are probably also effective at the end of life, though only one study has documented their safety and efficacy in patients with advanced HF.

Thigh muscle strengthening reduces dyspnea and fatigue in chronic stable HF, but may also benefit frail elderly patients or patients with advanced HF. Stimulants, including caffeine, target chemoreceptors and reduce fatigue in patients with chronic HF, and may be appropriate at the end of life.

Symptoms of depression are common in HF. Depression can be identified through the use of the PHQ-9 or short form Geriatric Depression Scale (GDS), both of which are self-administered, although asking a patient whether they are feeling “down or blue most of the time” is also a reasonable screen for depression.

Clinical depression should be treated with a trial of antidepressants, in addition to counseling. All antidepressants have potential adverse effects.

The tricyclic antidepressants (TCAs) fell out of vogue when selective serotonin reuptake inhibitors (SSRIs) became available; however, the less anticholinergic TCAs nortriptyline and desipramine are seeing increased use as the adverse effects of SSRIs are better known.

SSRIs still have an advantage in relatively faster onset of action and faster titration than TCAs. Both SSRIs and TCAs cause orthostatic hypotension, and SSRIs can worsen fluid retention and cause hyponatremia, particularly when renal function is impaired. Treating sleep-disordered breathing (with either continuous positive airway pressure [CPAP] or nocturnal oxygen supplementation) improves depression and anxiety, and increases well-being.

Pain in HF patients should be evaluated, and then treated with local therapies or low dose opioids. Opioid therapy should begin with short acting medications, and avoid drugs with neurotoxic renally excreted metabolites, such as morphine.

Advanced HF is accompanied by both sarcopenia (loss of muscle tissue and strength) and by weight loss and cachexia. Cachexia is identified by loss of 6% of “dry” body weight over 6 months, and is accompanied by anorexia.

It is important to note that cachexia should be identified in advanced HF, but that sarcopenia is present in about one third of women and one half of men over age 80, even in the absence of HF. Clear evidence exists for treating sarcopenia and cachexia with ACE inhibitors.

Other agents that block the RAAS and neurohormonal alterations of HF may also be beneficial. Exercise is also an important therapy in sarcopenic or cachectic persons, though the focus should be on thigh muscle strengthening (as that benefits dyspnea and fatigue) and on starting with multiple short periods of exercise daily. Patients that are severely deconditioned will need encouragement to exercise.

B. Physical Examination Tips to Guide Management.

Physical examination should address the presence of pulmonary congestion (left-sided HF) and of elevated central venous pressure and edema (right-sided HF). Significant right sided HF suggests pulmonary hypertension, which should be evaluated by a minimum overnight oximetry or a sleep study to identify sleep-disordered breathing and/or nocturnal hypoxia.

Frailty can be identified by a slow gait speed (more than 6 seconds to walk 5 m), and sarcopenia can be screened for by testing the patient’s ability to stand from a chair without using the arms. The normal person should be able to stand from a chair 3 to 5 times in rapid succession without using his or her arms, whereas frail and sarcopenic persons will often not be able to stand at all without pulling or pushing themselves up.

C. Laboratory Tests to Monitor Response To, and Adjustments in, Management.

No specific laboratory tests are indicated; however, a patient report of symptom severity and interference with activities should be periodically reobtained. Albumin is low in advanced HF patients and an indicator of poor prognosis. Reduced functional status over the past 6 to 12 months also suggests poor prognosis.

In general, following albumin is not helpful. If you want to test an aggressive nutritional program, than a prealbumin is a more sensitive measure of change in albumin synthesis; however, it is also appropriate to follow weight and physical function or gait speed. Reduction in gait speed over time or reduction in physical function predict a shortened length of life.

The depression screens, such as PHQ9 and GDS, may be administered serially to assess response to an antidepressant.

D. Long-term management.

Long-term management of HF should continue through the end of life. Medications for left ventricular systolic dysfunction that alter the neurohumoral disarray (ACE-inhibitors or ARBs, beta-blocker, and aldosterone-blockers) are important to managing the symptoms of HF. Diuretics should be adjusted to maintain as close to normal volume status as possible.

Symptoms can be further treated with opioids, thigh muscle strengthening, nocturnal oxygen supplementation and so on.

E. Common Pitfalls and Side-Effects of Management

Opioids must be carefully titrated. Morphine and codeine both have active metabolites that cause delirium, myoclonus, and anxiety. All patients with advanced HF have renal dysfunction. Thus these opioids should be avoided except for limited short-term use (Table I).

Table I.

Drug	Duration	Half-life	Route	Starting dose and range
Fentanyl	1-2 hr	1.5-3 hr	IM, SQ	0.02-0.1 mg
Hydrocodone	4-8 hr	3.3-4.5 hr	PO	5-20 mg
Hydromorphone	4-5 hr	2-3 hr	PO	1-7.5 mg
Methadone	4-6 hr	15-30 hr	PO	2.5-20 mg
Oxycodone	4-6 hr	NA	PO	15-30 mg

Nonsteroidal antiinflammatory agents are contraindicated in HF patients, and should be avoided.

Nonsystemic approaches to pain management, such as topical salicylates or other topical analgesics, and physical therapy are appropriate for musculoskeletal pain in HF patients.

Opioids that do not have renally excreted metabolites include fentanyl and methadone, although physicians familiar with their prescription should initiate both of these. Fentanyl is not FDA-approved for noncancer pain, and methadone has complex distribution, such that it accumulates in fat with an effectively long half-life over time.

At doses in the range of 100 mg/day, methadone can cause QTc prolongation. Oxycodone and hydromorphone are generally tolerated in patients with renal dysfunction, though not all patients respond to oxycodone. Thus hydromorphone is an appropriate first-line, short-acting opioid in HF patients.

IV. Management with Co-Morbidities

The vast majority of HF patients die from a comorbid problem. The HF might complicate management of the underlying cause of death (such as with advanced lung disease, pneumonia, or hip fracture). It is also possible that HF simply coexists with cancer, stroke, or dementia or other causes of death.

Prognostic counseling should not be forced on patients and their families. It should be integrated into decisions about care and appropriate interventions. Many patients will not want to know their possible length of life; however, physicians should be prepared to estimate this if patients or families request.

Prognosis should be provided to patients and families when they request it, or when it is important to guide choices about care. All patients with HF should be made aware that HF shortens the length of life in general. Physicians should be prepared to answer questions from patients or families when they arise.

Length of life estimates should be given as a range (for example, months to a year or two), and should be told with the proviso that some patients live longer and others shorter than we expect. It is appropriate to acknowledge the uncertainty associated with the course of HF.

Prognostic counseling should also integrate a discussion of goals of care. This would reasonably start by asking the patient or family what is important at this point in the patient’s life, and acknowledging what can be hoped for. Goals should be a statement of what can be reasonably obtained, and an acknowledgement both of what the patient hopes for and ways in which the hopes can be achieved.

The discussion of goals of care should also include a discussion of plans that should be made. Often the statement of such plans can be acknowledged as “if things do not go as we hope.” This “hope for the best-plan for the worst” dichotomy is effective at addressing and maintaining hope, while acknowledging the reality of eventual death.

A discussion about goals or prognosis can begin by identifying for the patient that you’ve reached a decision point, or a turning point in their illness. Starting out such a conversation by asking the patient and family how they believe things are going recently, then sharing your impression (for example, “your heart and kidney function are worse and we’ve reached a point where we might not have good treatments to make them better”, or “… we’ve reached a point where we need to rethink the goals of your care”) begins the conversation.

While it is uncomfortable to tell a patient they will die soon, it is important to be honest with patients. Phrases such as “I worry that you may not have long to live” or “I am sad to say that we might be nearing the last phase of your life” can frame the discussion honestly and kindly.

Hospice care is appropriate when the clinician estimates the patient has about 6 months life expectancy, and when the patient prefers to not be rehospitalized for their HF. The HF clinician should document the likely prognosis and be prepared to provide subsequent documentation at specific intervals when hospice eligibility is reviewed. As long as the patient has a likely 6-month prognosis, hospice care can be recertified repeatedly until the end of life.

The Medicare hospice benefit requires the patient or a surrogate to enroll in the hospice benefit. Specifically, the patient must agree to care focused on comfort and to using the hospice to manage exacerbations and distress, rather than the acute hospital. Patients may revoke the hospice benefit at any time and return to the acute care system.

Care in a hospice varies among agencies, so it is wise to either identify hospice agencies with expertise in heart failure, or to work directly with the hospice providers to manage the patient’s medications. Larger agencies and those with specific heart failure programs are most likely to provide expensive intravenous or other therapies to manage heart failure symptoms.

Plans for the end of life should include a discussion of where care should be provided (hospital, home, another location), and whom will provide care (nurses, family, hospice providers). Doctors should define how they will maintain contact with the patient through the end of life.

Depending on where the patient receives care, this could include daily visits in the hospital or weekly phone calls with nursing home staff or hospice clinicians. Physicians can also write a brief note to the patient or family to let them know they are thinking of them.

This effort to maintain continuity requires minimal time for physicians and their staff, yet can be very meaningful to patients and their families. After death, it is appropriate to either call or write the family and express condolences.

Plans with patients and families should anticipate events such as edema and weight gain, increased dyspnea, syncope, and clear instructions, and appropriate medications should be provided.

All treatments should be reviewed with the patient in light of new goals, and those that are no longer achieving the goals should be discontinued. When the goal is to allow natural death, treatments including medications such as statins, and implantable electrical devices should be discontinued or deactivated. Deactivation of electrical devices should be conducted electively and all centers that implant devices should have a protocol for deactivation.

V. Patient Safety and Quality Measures

A. Appropriate Prophylaxis and Other Measures to Prevent Readmission.

Quality of care at the end of life includes communication that addresses patient and family needs, maintenance of continuity as noted above, and management of symptoms to the level desired by the patient. Some patients elect to receive medications to manage dyspnea or pain even though they are sedated, and others chose to remain alert as long as possible and accept some discomfort to do so.

Documenting the patient’s preferred level of comfort and assessment of symptoms is important.

It is also important to document communication with the patient and family.

Edmonton scale

(see Figure 1)

B. What's the Evidence for specific management and treatment recommendations?

Goodlin, SJ. “Palliative Care in Congestive Heart Failure. (Solicited State of the Art Review)”. J Amer Coll Cardiol. vol. 54. 2009. pp. 386-96. (This article is a review of pathophysiology of symptoms in heart failure and recommendations about integration of palliative care.)

Goodlin, SJ, Kutner, J, Connor, S. “Hospice care for heart failure patients”. J Pain Symptom Manage. vol. 29. 2005. pp. 525-8. (This survey identifies hospice care provided to patients with heart failure.)

Lampert, R, Hayes, DL, Annas, GJ. “HRS expert consensus statement on the management of cardiovascular implantable electronic devices (CIEDs) in patients nearing end of life or requesting withdrawal of therapy”. Heart Rhythm. vol. 7. 2010. pp. 1008-26. (Statement from the Heart Rhythm Society about end of life care.)

Padeletti, L, Arnar, DO, Boncinelli, L. “EHRA expert consensus statement on the management of cardiovascular implantable electronic devices in patients nearing end of life or requesting withdrawal of therapy”. Europace. vol. 12. 2010. pp. 1480-9. (Statement from European Heart Rhythm Association about end of life care.)

Zile, MR, Gaasch, WH, Anand, IS. “Mode of death in patients with heart failure and a preserved ejection fraction: results from the Irbesartan in Heart Failure With Preserved Ejection Fraction Study (I-Preserve) trial”. Circulation. vol. 121. 2010. pp. 1393-1405. (The article discusses a course to death for patients with HFPEF.)

Goodlin, SJ, Quill, TE, Arnold, RM. “Communication and decision-making about prognosis in heart failure care”. J Card Failure. vol. 14. 2008. pp. 106-113. (Article discusses approach to communication and decision making in HF.)

C. DRG Codes and Expected Length of Stay.

Hospital DRG codes reflect the primary illness as well as secondary conditions. There is no specific DRG code in use by Medicare for “end of life.” A modifier (“V-code”) for palliative care was developed in the late 1990s; however, this was abandoned and is not tracked by Medicare.

Patients receiving hospice care can continue to receive specialty HF care. This care is billed as usual through Medicare B. Even when patients elect to have the hospice medical director assume their primary care in hospice, the HF care should be managed collaboratively with the hospice staff.

It is important to maintain contact with patients and their families to express condolences regarding their decline, or to let the patient know you are thinking of them. Clinicians should also write or call families with condolences after a patient’s death.

No sponsor or advertiser has participated in, approved or paid for the content provided by Decision Support in Medicine LLC. The Licensed Content is the property of and copyrighted by DSM.

Jump to Section