#97383: Palliative Care and Pain Management at the End of Life

Palliative care service is rendered through several different models, including hospital-based inpatient programs, outpatient clinics (based in hospitals or private practices), and combined consultation services and inpatient programs. Hospice programs may provide a consultative service but generally assume direct responsibility for end-of-life palliative care rendered at home, hospital, or other hospice resident facility [15]. The Joint Commission began offering an advanced certification program for palliative care in September 2011 [18]. In an effort to enhance access to end-of-life care, models of care are being adapted for a variety of specific settings, such as rural communities, correctional facilities, long-term care facilities, children's hospitals, and intensive care units [19,20,21,22,23,24].

The delivery of comprehensive palliative care relies on a team of skilled providers with experience and training in the principles of palliative care (Figure 2). The team may be organized around a primary care clinician or a palliative care specialist who functions as consultant or principal provider [15,25,26,27]. Palliative care interventions have been shown to significantly improve patient outcomes, although the data are stronger for patients with cancer than for life-limiting diseases overall [28,29]. This is illustrated by results of a randomized controlled trial of early palliative care provided as an adjunct to standard oncologic care for patients with metastatic non-small cell lung cancer. In this study, involving 151 subjects enrolled from a single academic practice, patients in the intervention group were seen by a palliative care clinician regularly (once or more per month) for 12 weeks; in comparison to the group receiving only standard care, the intervention (palliative care) group had a measurably better quality of life, lower rate of depression, and improved survival by 2.7 months [30].

MEMBERS OF THE PALLIATIVE CARE TEAM

Figure 2

Source: [15,25,26,27]

The composition of a hospice care team is essentially the same as for a standard palliative care team. The primary service provided during hospice care is skilled nursing care and management of distressing symptoms, followed by bereavement services and medical social services (Figure 3) [31]. Clinical specialists (e.g., oncologists, cardiologists, pulmonologists) also become members of a palliative care team when they are involved with the care of their patients during the end of life. Family physicians and general internists typically provide primary palliative care; this level of care requires skill in core palliative care competencies (such as basic symptom assessment and management and knowledge of psychosocial and community services) [6,8].

SERVICES PROVIDED DURING HOSPICE CARE

Figure 3

Source: [31] Reprinted from Caffrey C, Sengupta M, Moss A, Harris-Kojetin L, Valverde R. Home health care and discharged hospice care patients: United States, 2000 and 2007. Natl Health Stat Report. 2011;(38):1-27.

The composition of the healthcare team and the roles of its members may differ across palliative care settings, and roles and responsibilities should be clear and well documented to help members work effectively as a team. Team members' roles should also be communicated to the patient and the family.

The primary care physician is usually responsible for referral to palliative care through hospice when the patient has a non-cancer diagnosis. In general, the primary care physician becomes the attending physician, assuming primary responsibility for the patient [32,33]. The primary care physician should be prepared to relinquish some autonomy in order to work effectively with the interdisciplinary team [32]. Home-based hospice care is organized around a team that includes the attending physician, registered nurse, social worker, and counselor. These team members are necessary for Medicare reimbursement [34]. The attending physician collaborates with other members of the hospice team to manage symptoms and fulfills other basic obligations, such as admission orders, medication prescriptions and refills, certification of hospice eligibility, and signing of the death certificate [33].

High-quality palliative care also requires special expertise in honest, compassionate communication. In addition to enhancing the patient's and family's experience, these skills help to establish trust and overcome barriers to adequate care and relief of symptoms. Several communication tasks are especially important: conveying accurate prognostic information while maintaining hope, eliciting information about symptoms, decision making about curative and palliative treatments, handling emotions, and dealing with requests from patients and families who have unrealistic goals [35,36,37]. The challenges of communicating effectively are discussed later in this course.

Cancer and human immunodeficiency virus (HIV)/AIDS were once the predominant diseases in hospice and palliative care programs, but as treatments for these diseases have improved, the number of individuals with the diseases in hospice programs has decreased while the number of individuals with chronic, progressive diseases has increased. Cancer still accounts for the highest proportion of a single disease entity (29.6%) in the hospice population [1]. In 2018, the four noncancer diseases that represented the highest percentages of hospice use were cardiac and circulatory disease (17.4%), dementia (15.6%), other (14.7%), and respiratory disease (11.0%) (Figure 4) [1].

DISTRIBUTION OF DIAGNOSES IN HOSPICE CARE, 2018

Figure 4

Source: [1]

Medical school and residency training programs emphasize disease recognition, diagnostic assessment, and treatment and management strategies that have restorative power, prolong life, and prevent death. The role of palliative care traditionally has not been sufficiently addressed [67]. Students who have participated in mandatory courses in palliative medicine have noted that they are better prepared to care for dying patients [68]. Efforts to enhance education have resulted in the development of more than 100 primary care residency programs that offer palliative medicine as part of the curriculum and 72 postgraduate medical fellowship programs in palliative care [69,70]. In addition, hospital-based palliative care programs have integrated the eight NCP domains into graduate courses and residencies for physicians and registered nurses, and certifications in palliative care have become available for physicians, nurses, and social workers [6]. Between 1996 and 2006, more than 2,100 physicians obtained certification in hospice and palliative medicine from the American Board of Hospice and Palliative Medicine [71]. As of January 2016, there were nearly 6,400 active certified hospice and palliative care physicians in the United States [72]. In 2006, subspecialty certification in hospice and palliative medicine was established for 10 Boards within the American Board of Medical Specialties. The first exam was held in 2008, and 1,274 physicians earned certification in hospice and palliative medicine. Since then, the number of physicians who have earned certification though the American Board of Medical Specialties has increased nearly fivefold, with physicians in internal medicine and family medicine accounting for 85% of the total (Table 3) [73].

Initiatives to enhance the knowledge and skills of nurses have included graduate nursing programs in palliative care, instructional resources for nursing educators, communication skills training for nurses, and educational programs for home care professionals, as well national certifications [6,74]. More than 18,000 nurses are Board-certified in hospice and palliative nursing [75].

Medicare reimbursement for hospice care became available when the Medicare Hospice Benefit was established in 1982, and reimbursement through private health insurances soon followed [34]. Reimbursement for hospice enabled more people with life-limiting disease to receive palliative care at home and in hospice units: the number of hospices in the United States has increased steadily, from 158 Medicare-certified hospices in 1985 to 4,639 in 2018 [1]. Despite the positive impact of the Medicare Hospice Benefit, the benefit has impeded timely referral because of its policy disallowing concurrent hospice care and life-prolonging treatment (Table 4). The eligibility requirements of the benefit explicitly state that the focus of hospice "is on caring, not on curing," and in order to receive reimbursement for hospice services, patients must sign a statement that they will forego curative treatment [34]. This requirement frightens some patients or their families, who subsequently view hospice as "giving up." Furthermore, the restriction does not account for palliative treatments that serve the dual purpose of alleviating symptoms while prolonging life. For example, therapeutic regimens and measures designed to optimally treat heart failure are the same as those used for palliative care of patients with heart failure [76]. At present, there are no Medicare regulations that specify which treatments are considered palliative, and this lack of clarity has led to variation in what treatments individual hospice programs offer. Hospice care may be denied to patients receiving palliative chemotherapy or radiotherapy, and this may result in many people not choosing hospice. Although oncology experts have noted that radiotherapy is an important component of palliative care for many people with metastatic cancer, only 3% of people receiving hospice care receive radiation therapy; expense and the need to transport patients were the primary barriers [77,78,79]. Other palliative interventions, such as chemotherapy, blood transfusions, total parenteral nutrition, and intravenous medications, may not be economically feasible for small hospice units but may be possible at larger ones [80,81,82].

Many have suggested that the hospice model should change to allow for integration of disease-directed therapy [83,84]. The Affordable Care Act of 2010 stipulates that the Centers for Medicare & Medicaid Services (CMS) implement a three-year demonstration project to evaluate concurrent hospice care and disease-directed treatment [84]. This project represents a significant change to the eligibility criteria and, while the change has the potential to improve access to hospice care, careful assessment of the effect of concurrent treatment on use of hospice as well as on quality of life, quality of care, survival, and costs is needed [84]. Phase 2 of the project, the Medicare Care Choices Model, began January 1, 2018, and became a five-year study involving 140 hospices and an estimated 150,000 eligible Medicare beneficiaries. An estimated 60% of the original participants remained in the model as of September 2019. As of June 2020, 82 hospices are participating. Also in June 2020, CMS extended the model for an additional year, through 2021. [85].

The Medicare Hospice Benefit criterion of a life expectancy of six months or less has also affected the timeliness of referral to hospice because of the aforementioned challenges in predicting prognosis. Many hospices were accused of fraud and were assessed financial penalties when government review found documentation of patients who received hospice care for longer than six months. As a result, many clinicians delayed hospice referral because of their lack of confidence in their ability to predict survival within six months. However, the six-month regulation has been revised, and a penalty is no longer assessed if a patient lives beyond six months if the disease runs its normal course [34].

Unfortunately, reimbursement for end-of-life care discussions is not as straightforward as for hospice care. An effort to establish government reimbursement for discussions of end-of-life care options, including hospice care and advance directives, sparked a political storm that led to the removal of the proposed reimbursement from the healthcare reform bill of 2011. However, beginning in January 2016, the CMS introduced two Current Procedural Terminology (CPT) reimbursement codes for advance care planning visits [86,87]. Despite a national increase in advance care planning claims, the overall claims rate remains low. Two-thirds of hospice and palliative medicine specialists did not use the new CPT codes in 2017, despite working with seriously ill patients [87].

To make appropriate referrals to hospice, clinicians must be able to determine accurate prognoses, at least within the six-month timeframe required for reimbursement. However, prognostication is a complex issue and is a primary barrier to hospice use [88,89,90,91,92,93,94,95]. Studies have found that physicians typically overestimate survival, and one study found that physicians overestimate prognosis both in determining it and in communicating it to the patient [93,96,97]. The difficulty in determining the risk of death within a specific time period not only affects the ability of clinicians to make appropriate referrals to hospice but also impedes the ability of patients and families to make necessary end-of-life decisions, with many patients not fully understanding the severity and progressive nature of the disease [98].

Several factors contribute to physicians' difficulty in prognostication, including a desire to meet the patient's needs (for a cure or prolongation of life) and a lack of reliable prognostic models [81,97,99]. Perhaps the most important factor contributing to prognostic difficulty is the variations in disease trajectories, which have been characterized as a short period of evident decline, long-term limitations with intermittent serious episodes, and a prolonged decline (Figure 5) [9,32,100].

TYPICAL DISEASE TRAJECTORIES AND SERVICE NEEDS OVER TIME FOR ELDERLY PATIENTS

Figure 5

Source: [9] Reprinted, with permission from Lynn J, Adamson DM. Living Well at the End of Life: Adapting Health Care to Serious Chronic Illness in Old Age. Santa Monica, CA: Rand; 2003.

How difficult it is to determine a prognosis depends on the disease trajectory. Determining a prognosis in the cancer setting was once clear-cut because of the short period of evident decline, but advances in cancer therapies have made it more difficult to estimate a prognosis. Studies have shown rates of accurate prognosis of 20%, with survival usually overestimated, up to a factor of five [93,96]. The unpredictable course of organ-failure diseases, with its long-term limitations and acute exacerbations has always made prognostication difficult [62,89,101,102]. In a survey of cardiologists, geriatricians, and internists/family practitioners, approximately 16% of respondents said they could predict death from heart failure "most of the time" or "always" [89]. Predicting survival for people with the third type of trajectory (prolonged decline) is extremely difficult because of the wide variation in progressive decline. The prognosis for dementia can range from 2 to 15 years, and the end-stage may last for 2 years or more [103,104].

To help facilitate more timely referrals to hospice, the NHPCO established guidelines for determining the need for hospice care, and these guidelines were adopted by the Health Care Finance Administration to determine eligibility for Medicare hospice benefits [76]. Other prognostic models have been developed, such as the SUPPORT model, the Palliative Prognostic (PaP) Score, and the Palliative Prognostic Index (PPI) [12,105,106,107]. Most were developed for use in the cancer setting and for hospitalized patients, and their value beyond those settings has not been validated [81,95]. In addition, the PaP and the PIP will identify most patients who are likely to die within weeks but are much less reliable for patients who have 6 to 12 months to live [95]. A systematic review showed that the NHPCO guidelines, as well as other generic and disease-specific prognostic models, were not adequately specific or sensitive to estimate survival of at least six months for older individuals with nonmalignant life-limiting disease, especially heart failure, COPD, and end-stage liver disease [99].

Most prognostic tools for organ-failure diseases are used to estimate the risk of dying and to select patients for treatment, not to determine when end-of-life care should be initiated. Several models have been established to determine prognosis for heart failure; the one used most often is the Seattle Heart Failure model, which represents the most comprehensive set of prognostic indicators to provide survival data for one, two, and five years [108,109]. Newer evidence-based recommendations for estimating survival in advanced cancer have been published, as has a nomogram; however, use of the nomogram for hospice referral is limited, as it estimates survival at 15, 30, and 60 days [81,110].

For estimating prognosis in advanced dementia—a condition with the most challenging disease trajectory—the Advanced Dementia Prognostic Tool (ADEPT) has been shown to be better than the NHPCO guidelines in identifying nursing home residents with advanced dementia at high risk of dying in six months [111]. However, the ability of ADEPT to identify these patients is modest [111]. Lastly, the Patient-Reported Outcome Mortality Prediction Tool (PROMPT) was developed to estimate six-month mortality for community-dwelling individuals 65 years or older with self-reported declining health over the past year; the model shows promise for making appropriate hospice referrals, but the model needs validation [112].

In addition to the low reliability of these models, another problem is that the clinician's prediction of survival remains integral, as it is one element in prognostic models, sometimes representing as much as half of a final score [110]. Other variables include performance status, laboratory data, and quality of life scales.

Researchers continue to evaluate prognostic variables to establish criteria for prognosis, especially disease-specific criteria. In its guidelines for palliative care, the Institute for Clinical Systems Improvement (ICSI) distinguished between the clinical indicators that should prompt palliative care discussions and those that should prompt hospice referral (Table 5) [113]. According to the ICSI, "all hospice is palliative care, but not all palliative care is hospice" [113].

This difficulty in determining prognosis can have a negative impact on the appropriate timing of hospice referral and the degree of benefit to be derived. Although the use of hospice has increased over the past decades, the timing of referral has not changed significantly since the mid-1980s [114]. The average length of hospice care is much lower than the six months allowed by the Medicare benefit; in 2018, the average length was 89.6 days, and the median duration (a more accurate reflection because it is not influenced by outliers) was 18 days [1]. In addition, approximately 27.9% of patients died (or were discharged) within only seven days [1]. Studies have indicated that the benefits of hospice increase as the duration of care increases, and such services as bereavement counseling, palliative care, and respite for caregivers is best when hospice care is provided for four to eight weeks, a longer period of time than the median stay [59,115].

Longer durations of hospice services are also linked to family members' perceptions of the quality of care. According to the findings of 106,514 surveys from 631 hospices in the United States, 11% of family members thought their loved one was referred "too late" to hospice; this perception was associated with more unmet needs, higher reported concerns, and lower satisfaction [116].

In contrast to the restrictions on access to hospice, there are no restrictions on access to palliative care. Referrals for palliative care should be made on the basis of actual or anticipated needs at any time during the disease continuum; referrals should not be made on the basis of prognostic models [81,99]. Referrals for specialist palliative care should be made when treatment goals change from curative to palliative [117,118]. A consensus report from the Center to Advance Palliative Care provides guidance for identifying patients with a life-limiting illness who are at high risk of unmet palliative care needs [118]. The report includes criteria for referral for palliative care assessment at the time of hospital admission and during each hospital day (Table 6) [118]. Experts in nonmalignant life-limiting diseases are calling for earlier palliative care consultation. Such consultation before implantation of a left ventricular assist device as destination therapy is recommended, as it has been shown to improve the quality of care and advance care planning [119,120]. Guidelines for renal and respiratory diseases note that all patients with these diseases should be offered palliative care services, and the integration of palliative care specialists into liver transplantation teams has been suggested [121,122,123].

Physicians have reported that they delay discussions of palliative care and hospice because they fear the reaction of the patient and/or family [32,33]. Negative reactions are grounded in a lack of accurate knowledge about palliative care and hospice. According to two polls conducted in 2011 (800 adults in one poll and 1,000 adults in the other), 70% to 92% of respondents were not "too" or "at all" familiar with the term palliative care [124,125].

Enhancing the public's knowledge can improve access to palliative care: the 2011 polls showed that once palliative care was appropriately defined, 92% said they were likely (63% "very likely" and 29% "somewhat likely") to consider palliative care for a loved one who had a serious illness and 96% said that it was important for palliative (and end-of-life) care to be a top priority for the healthcare system [124,125].

Hospice is a more familiar concept to the general population. One of the polls showed that 86% of respondents were familiar with the term hospice care, and other studies have indicated that approximately half of patients with a life-limiting illness know what hospice is [125,126]. Although people may be familiar with the term, many believe several myths about hospice; for example, that hospice is only for old people, is only for people with cancer, is for people who do not need a high level of care, is used when there is no hope, and is expensive [127].

Several other factors contribute to negative feelings about hospice [80,88,128]:

When initiating a discussion about palliative care and hospice, clinicians should always first ask the patient if he or she has heard of either term and, if so, to describe his or her experience and knowledge [129]. Guidelines on communicating in the end-of-life setting note that clinicians must "clarify and correct misconceptions" about palliative care, especially emphasizing that such care is not limited to people who are imminently dying [117]. Clinicians should also address the factors that act as barriers to hospice by explaining that the goal of hospice is to die naturally—in the patient's own time, not sooner—and by ensuring that patients and families are fully informed about the prognosis, understand that the physician will be available for care, and know that routine care will continue [98,130].

Clinicians also need to evaluate their own attitudes about the use of curative therapies and hospice. Their interpretation of quality of life, a focus on longer survival rather than better quality of life, a fear of failure, and religious and cultural beliefs may influence their decision making about treatment options for patients near the end of life [131].

Among the most important factors for effective communication across all healthcare settings are knowledge of the language preference of the patient and family; an awareness of the patient's and family's health literacy levels; and an understanding of and respect for the patient's and family's cultural values, beliefs, and practices (referred to as cultural competency) [132,133,134]. These issues are significant, given the growing percentages of racial/ethnic populations. According to U.S. Census Bureau data from 2019, more than 67.8 million Americans speak a language other than English in the home, with more than 16.1 million of them (5.2% of the population) reporting that they speak English less than "very well" [135]. Clinicians should ask their patients what language is spoken at home and what language they prefer for their medical care information, as some patients prefer their native language even though they have said they can understand and discuss medical information in English [136]. When the healthcare professional and the patient speak different languages, a professional interpreter should be used. Studies have demonstrated that the use of professional interpreters rather than "ad hoc" interpreters (untrained staff members, family members, friends) facilitates a broader understanding and leads to better outcomes [137,138]. Using a family member as a translator confuses the role of that member in the family, may involve confidentiality issues, and may lead to a modified message to protect the patient. In addition, individuals with limited English language skills have indicated a preference for professional interpreters rather than family members [139]. Professional interpreters have recommended that clinicians can further enhance the quality of care by meeting with interpreters before discussions of bad news and by explicitly discussing with the interpreter whether strict interpretation or cultural brokering is expected [140].

Knowledge of the family's health literacy is important for achieving treatment goals and good outcomes, yet most individuals lack adequate health literacy. Studies have indicated that as many as 26% of patients have inadequate health literacy, which means they lack the ability to understand health information and make informed health decisions; an additional 20% have marginal health literacy [141,142,143]. Health literacy varies widely according to race/ethnicity, level of education, and gender, and clinicians are often unaware of the literacy level of their patients and family [134,144].

The Institute for Clinical Systems Improvement asserts that clinicians caring for patients with serious illness should examine their own cultural values and assumptions about what constitutes "good" care for patients nearing the end of life, recognizing not all patients will share these same values, and ensure goals and decisions remain centered around the patient's values/beliefs.

(https://www.icsi.org/wp-content/uploads/2020/01/PalliativeCare_6th-Ed_2020_v2.pdf Last Accessed: October 26, 2021)

Level of Evidence: Expert Opinion/Consensus Statement

Several instruments are available to test the health literacy level, and they vary in the amount of time needed to administer and the reliability in identifying low literacy. Among the most recent tools is the Newest Vital Sign (NVS), an instrument named to promote the assessment of health literacy as part of the overall routine patient evaluation [145]. The NVS takes fewer than three minutes to administer, has correlated well with more extensive literacy tests, and has performed moderately well at identifying limited literacy [134,144]. Two questions have also been found to perform moderately well in identifying patients with inadequate or marginal literacy: "How confident are you in filling out medical forms by yourself?" and "How often do you have someone help you read health information?" [134]. Clinicians should adapt their discussions and educational resources to the patient's and family's identified health literacy level and degree of language proficiency and should also provide culturally appropriate and translated educational materials when possible.

Cultural competency is essential for addressing healthcare disparities among minority groups [132]. Clinicians should ask the patient about his or her cultural beliefs, especially those related to health and dying, and should be sensitive to those beliefs [146]. In addition, information sharing and the role of decision maker vary across cultures, and the healthcare team must understand the family dynamics with respect to decision making [117]. Clinicians should not make assumptions about the preferences of the patient or family on the basis of cultural beliefs. Even within a single culture or ethnicity, the level of information desired, preferences for treatment, role of other family members in decision making, and goals of care differ among patients and families [40,117]. Clinicians should ask their patients about these issues, as well as other family and social factors and religious or spiritual views [40].

The Institute for Clinical Systems Improvement recommends that palliative care discussion or referral should be considered whenever a patient develops or presents with a serious or life-threatening illness, in all care settings.

(https://www.icsi.org/wp-content/uploads/2020/01/PalliativeCare_6th-Ed_2020_v2.pdf Last Accessed: October 26, 2021)

Level of Evidence/Strength of Recommendation: Low-Quality Evidence, Strong Recommendation

Patients and families have noted that communication about end-of-life care is one of the most important skills for clinicians to have [147]. Experts in end-of-life communication note that physicians have an obligation to discuss medical, psychosocial, and spiritual needs with seriously ill patients in a timely and sensitive manner [148]. In addition, communication guidelines developed by an Australian and New Zealand Expert Advisory Group recommend "all patients with advanced progressive life-limiting illnesses be given the opportunity to discuss prognosis…and end-of-life issues" [117].

Although the topic is emotionally and intellectually overwhelming for patients and families, they want information. In a systematic review (46 studies), Parker et al. found that patients with advanced life-limiting illnesses and their families have a high level of information needs at all stages of disease [149]. That study and others have shown that the end-of-life issues of most importance to patients are [62,149]:

Most patients want some discussion of end-of-life issues (including hospice care) at the time a life-limiting illness is diagnosed or shortly thereafter [62,149].

Although many physicians say they avoid discussing end-of-life issues because they are afraid the conversation will destroy the patient's hope, the discussion actually offers many benefits: it makes patients fully informed and thus better able to make decisions about treatment options and care goals; provides patients with an opportunity to achieve closure on life and family issues; allows patients to handle practical matters; and enables patients to carry out advance care planning [35,81,148,150]. As such, the discussion empowers patients, giving them a sense of control over choices [148,150]. Patients who discuss end-of-life issues and goals of care with their clinician also are more likely to receive care that is consistent with their preferences, to enroll in hospice, to complete advance directives, and are less likely to be intubated or to die in an intensive care unit [151,152].

Despite these benefits, studies have consistently shown that few clinicians and patients discuss end-of-life issues or discuss them in a timely manner. Overall, about 25% to 33% of physicians have noted that they did not discuss hospice or end-of-life care with their patients who have life-limiting diseases [128]. In a multiregional study of more than 1,500 people with stage IV lung cancer, 47% had not discussed hospice within four to seven months after diagnosis [153]. Discussions are particularly lacking among people with nonmalignant life-limiting diseases, with 66% to more than 90% of patients or clinicians reporting that they had not discussed end-of-life issues [62,126,154,155].

Even among clinicians who discuss end-of-life care with their patients, the timing is not optimal. Approximately 24% of physicians have noted that they provide hospice information at the time of diagnosis, the point at which this discussion is recommended [128,156]. In a national survey of clinicians caring for people with cancer, most respondents said they would wait until treatment options had been exhausted or symptoms had occurred before discussing end-of-life issues, and many said they would have the discussion only if the patient or family raised the issue [157].

Patients and clinicians should talk about end-of-life issues early to avoid discussing the topic during the stress of exacerbated disease or imminent death. The topic can then be framed as a component of care for all patients with a life-limiting illness [62,117]. According to published guidelines and expert recommendations, end-of-life issues should be discussed when the clinician would not be surprised if the patient died within six months to one year [6,117,148]. As other markers, an end-of-life discussion is generally recommended in the presence of moderate or severe COPD, during evaluation for liver transplantation, and in the presence of stage 4 or 5 chronic kidney disease or end-stage renal disease [62,121,123]. The 2009 ACCF/AHA guideline for the diagnosis and management of heart failure noted that end-of-life care options should be discussed when "severe symptoms in patients with refractory end-stage heart failure persist despite application of all recommended therapies;" the 2013 guideline for the management of heart failure is less clear about the timing of such a discussion [158,159]. The ACCP recommends discussing end-of-life care options when caring for patients with advanced lung cancer [54].

Other indications that should prompt a conversation about end-of-life care are a discussion of prognosis or of a treatment option with a low likelihood of success, a change in the patient's condition, patient and/or family requests or expectations that are inconsistent with the clinician's judgment, recent hospitalizations, and patient and/or family questions about hospice or palliative care [117,148].

Several patient-related and clinician-related factors contribute to the low rate of end-of-life discussions or their untimeliness. Most patients will not raise the issue for many reasons: they believe the physician should raise the topic without prompting, they do not want to take up clinical time with the conversation, they prefer to focus on living rather than death, and they are uncertain about continuity of care and fear abandonment [62,117,148,150]. Clinician-related factors include [81,147,148,160,161]:

Perhaps the greatest barrier to end-of-life care discussions is clinicians' lack of confidence in their ability to talk about end-of-life issues, and research has confirmed a low rate of effective communication skills among clinicians, especially with respect to delivering "bad news" [62,81,155,162].

The Australian/New Zealand communication guideline provides several evidence-based recommendations for discussing end-of-life issues, and other experts have offered practical guidance to help clinicians discuss bad news and end-of-life care more effectively [117,163,164,165]. These guidelines and expert recommendations emphasize communication behaviors that patients and families have noted to be most important, such as expression of empathy, acknowledgment and support of emotions, honesty, willingness to listen more than talk, and encouragement of questions [81,117,123,147,149,164].

The most commonly recommended communication approach is SPIKES, a six-step protocol that was developed for delivering bad news in the oncology setting and can be used in other settings [163,164]:

In establishing the setting, the clinician should ask the patient if he or she wishes to have a family member present for the conversation and should ensure that the discussion takes place in privacy [62,164]. The clinician should also introduce himself or herself to the patient and any others present [164]. With SPIKES, the setting also involves listening skills—the use of open-ended questions, clarification of points, and avoidance of distractions [164].

Bad news—even when delivered clearly and compassionately—can affect the ability of patients and family members to understand and retain information. To minimize misinterpretation, clinicians should use simple (jargon-free) language and open-ended questions and ask follow-up questions that include the patient's own words [117,164]. Clinicians should also check often to make sure the patient and/or family understands, as research has shown that clinicians tend to overestimate their patients' understanding of end-of-life issues [166]. The discussion should focus on the importance of relieving symptoms and enhancing the quality of life, to avoid having the patient and/or family think that the clinician is "giving up" or abandoning the patient [40,117]. Clinicians should also provide educational resources in a variety of formats (print, Web-based, video, etc.) to address different learning styles.

It was once thought that the ability to communicate effectively was innate and thus could not be taught [164]. However, multiday communication skills training programs have enhanced the skills and behaviors of beginning and experienced physicians and nurses. These programs have improved clinicians' use of more focused questions and open questions, expression of empathy, and appropriate responses to cues [167,168]. Patient-related interventions have also helped to enhance end-of-life discussions. A structured list of questions and the use of individualized feedback forms regarding end-of-life preferences have led more patients to ask their physicians about end-of-life care [169,170].

The vast majority of patients say that they want to know their prognosis, and most clinicians believe that patients and families should be told the truth about the prognosis [122,126,150]. However, discussions of prognosis are lacking among clinicians and patients with life-limiting diseases. Across studies and surveys, fewer than half of patients have had a truthful discussion of prognosis [81,108,150]. Many physicians have said they discuss prognosis only when asked by the patient or family [81].

In discussing prognosis, clinicians tend to be overly optimistic, and, although most clinicians believe that they should be truthful, they sometimes withhold the truth, often at the request of a family member [160]. Honesty about the prognosis, with acknowledgment of inherent uncertainty, is needed because patients who are aware of their prognosis are more likely to choose hospice rather than aggressive treatment and to carry out advance directives [31,81,171]. Conversely, patients who are not fully aware of their prognosis tend to overestimate their life expectancy, which can influence decision making about treatment options [129].

As with other end-of-life issues, the prognosis should be discussed when the clinician would not be surprised if the patient died within six months to one year [6,117,148]. For patients with cancer, it is recommended that the prognosis be discussed within one month after a new diagnosis of advanced cancer is made [161]. A guideline from the Renal Physicians Association notes that prognosis should be fully discussed with all patients who have stage 4 or 5 chronic kidney disease or end-stage renal disease [121]. A discussion about prognosis is also recommended before the initiation of such treatments as implantation of a left ventricular assist device, dialysis, and ventilator support [122,150,172,173,174].

Clinicians should carefully prepare for the discussion of prognosis by reviewing the patient's medical record and talking to other healthcare professionals involved in the care of the patient [117]. Because there is variation among patients with regard to their desire for information, clinicians should follow the "ask-tell-ask" approach: ask the patient if he or she is willing to discuss prognosis; if yes, discuss the prognosis and then ask the patient to confirm his or her understanding [62,164]. When discussing prognosis, quantitative estimates are more understandable for patients and family than qualitative ones (such as "poor"), and general timeframes for survival should be given [62,81,164,175]. In addition, clinicians should emphasize that prognosis is determined by looking at large groups of patients and that it is harder to predict survival for an individual [62,121,129,161]. The discussion of prognosis is often not documented in the patient's record but should be [117].

Treatment options and goals of care are other topics that are often avoided in the end-of-life setting. A discussion of the survival benefit of palliative chemotherapy is frequently vague or absent from discussions of treatment options for patients with cancer [176]. In another example, approximately 60% to 95% of physicians involved with the care of patients with heart failure have two or fewer conversations about deactivation of implantable cardioverter defibrillators, and the discussions are usually within the last few days of life [89,177].

Deciding when curative therapy should end is difficult because of the advances made in treatment and life-prolonging technology and the unpredictable course of disease, especially for organ-failure diseases. These factors have led many patients, as well as some clinicians, to have unrealistic expectations for survival [30,178]. Unrealistic expectations are a major contributor to an increased use of aggressive treatment at the end of life. Among more than 900 patients with cancer, those who thought they would live for at least six months were more likely to choose curative therapy than "comfort care" compared with patients who thought there was at least a 10% chance they would not survive for six months [179].

Many studies have demonstrated high rates of aggressive treatment within the last months to weeks of life, with increased rates of hospital admissions, stays in an intensive care unit, use of medical resources, and use of chemotherapy. Goodman et al. found that patients with severe chronic disease near the end of life spent a disproportionate number of days in an intensive care unit and received care from multiple physicians; more than half of the patients saw 10 or more physicians within the last six months of life [55]. Similarly, Sheffield et al. found high rates of admission to the intensive care unit among nearly 23,000 patients with pancreatic cancer, and Unroe et al. found that 80% of more than 229,000 people with heart failure were hospitalized in the last six months of life [180,181]. In the cancer setting, several researchers have reported increased rates of chemotherapy in the last two to four weeks of life [180,182,183]. However, studies to evaluate the benefit of high-intensity treatment near the end of life have consistently found that such treatment offers no survival benefit, decreases the quality of life, and delays the use of hospice [55,80,184,185].

Before discussing treatment options, the clinician should talk to the patient to gauge his or her level of understanding of the disease and prognosis and to explore the quality-of-life factors that are most important [186]. The clinician should frame the conversation to focus on active interventions rather than the end of curative therapy; should focus on the overall care goals; and should discuss options within the context of these goals (that is, does the patient wish to enroll in hospice, enroll in a phase I trial, or be present at a family event?) [81,117]. The discussion should include an explanation of the likelihood of achieving the patient's goals with each option and a comparison of the risks, benefits, and costs of each option, noting the overall lack of benefit of aggressive treatment [187,188]. It is also important to allow the patient and family enough time to express emotion and concerns and to ask questions [117,164,189]. Because frequent exacerbations in organ-failure diseases are usually predictive of a more rapid decline, hospitalizations for disease exacerbation should prompt discussions about changes in prognosis and treatment goals and advance care planning [101,190,191]. Admission to the hospital or intensive care unit should also prompt a discussion of goals and preferences with patients with cancer; this conversation should be documented within 48 hours after admission [161]. The ACCP recommends a discussion of the pros and cons of life-sustaining treatment when caring for patients with advanced lung cancer [54].

When the patient, family, and/or healthcare team do not agree on the benefit/utility of interventions, the clinician should consider consulting with social workers or pastoral care services to help with conflict resolution [187]. In addition, the clinician should explain to patients that the likelihood of insurance coverage for a treatment is low if it is not medically indicated [188].

Clinical guidelines have begun to address the use of aggressive treatment at the end of life. The ACCF/AHA guideline on the management of heart failure notes that it is not appropriate to carry out aggressive procedures in the last several months of life if they do not contribute to recovery or improve quality of life (including intubation and implantation of a cardiac defibrillator) [108]. In addition, discussion of device deactivation for patients with refractory heart failure is recommended [159]. The Renal Physicians Association recommends forgoing dialysis for patients with chronic kidney disease or end-stage renal disease who have "very poor prognosis" [121]. The increased use of chemotherapy near the end of life has led oncology experts to recommend more judicious use of chemotherapy. Oncologists have called for the discontinuation of chemotherapy when the chance of success is minimal, such as when disease progresses after three consecutive regimens [178,192].

Early discussion of preferences for life-sustaining measures is especially important. It is estimated that roughly 75% of patients will be unable to participate in some or all of the decisions pertaining to their care at the end of life [193,194]. Documentation of preferences helps to inform decision making by the physician and the patient's health care proxy (surrogate decision maker). Clinicians should encourage their patients to designate a healthcare proxy early in the course of a life-limiting disease [62,82,122]. Patients should be urged to clarify their wishes with their chosen health proxy, as a proxy often inaccurately predicts a patient's wishes or may have values that conflict with those of the patient [194].

Advance directives, designation of a healthcare proxy, do-not-resuscitate (DNR) orders, and living wills were developed as a way to ensure that patients received care that was consistent with their preferences and goals. Advance directives offer many benefits; they have been associated with a lower likelihood of in-hospital deaths, an increased use of hospice, and a significant reduction in costs [195]. Although early studies showed that advance directives did not always translate into patients receiving their preferred level of care, later studies have demonstrated that most patients with advance directives do receive care consistent with their preferences, especially if they want limited care (rather than "all possible" care) [194,196].

The American College of Physicians recommends that clinicians ensure that patients with "serious illness" engage in advance care planning, including the completion of advance directives [47]. Clinicians must emphasize the value of advance directives because most patients have not completed them. An estimated 20% of the population have written advance directives, with higher rates among the older population and nursing home residents and lower rates among minority populations and those with nonmalignant life-limiting diseases (compared with people with cancer) [197,198,199,200,201]. Other guidelines recommend that advance care planning be done early in the course of disease, to help avoid potential compromise of decision-making capacity near the end of life [62,108,121,122].

In preparing for a discussion about advance dir10ectives, clinicians should ask the patient if he or she wishes to have other family members present during the conversation. This is especially important for patients of some cultural backgrounds, as healthcare decisions are the responsibility of family members in many cultures [200]. Increased efforts should be aimed at obtaining advance directives from patients of minority races/ethnicities. Although the rate of advance directives is higher in the gay and lesbian community than in the general population, clinicians should emphasize the importance of these documents to gay and lesbian patients to ensure that the patient's wishes are carried out and to avoid legal consequences for the patient's partner [202].

DNR orders and living wills have limitations and have been open to interpretation, which has led to the development of medical order forms based on patients' preferences. Developed at the Oregon Health & Science University, the Physicians Orders for Life-Sustaining Treatment (POLST) program is designed to ensure that a patient's preferences regarding cardiopulmonary resuscitation, scope of treatment, artificial nutrition by tube, and use of antibiotics (in some states) can be followed, regardless of where the patient receives care [203]. Nearly all states have an endorsed POLST program or one in development [203]. A POLST does not replace a traditional advance directive, and when available, an advance directive should accompany the POLST form [203]. Early studies have demonstrated that the use of POLST has led to higher rates of meeting patients' preferences [204,205].

Legal issues related to advance directives and POLSTs vary according to state, and clinicians should be familiar with the law in the state in which they practice [197].

The inadequate management of pain is the result of several factors related to both patients and clinicians. In a survey of oncologists, patient reluctance to take opioids or to report pain were two of the most important barriers to effective pain relief [226]. This reluctance is related to a variety of attitudes and beliefs [222,226]:

Education and open communication are the keys to overcoming these barriers. Every member of the healthcare team should reinforce accurate information about pain management with patients and families. The clinician should initiate conversations about pain management, especially regarding the use of opioids, as few patients will raise the issue themselves or even express their concerns unless they are specifically asked [227]. It is important to acknowledge patients' fears individually and provide information to help them differentiate fact from fiction. For example, when discussing opioids with a patient who fears addiction, the clinician should explain that the risk of addiction is low [222]. It is also helpful to note the difference between addiction and physical dependence.

There are several other ways clinicians can allay patients' fears about pain medication:

Encouraging patients to be honest about pain and other symptoms is also vital. Clinicians should ensure that patients understand that pain is multidimensional and emphasize the importance of talking to a member of the healthcare team about possible causes of pain, such as emotional or spiritual distress. The healthcare team and patient should explore psychosocial and cultural factors that may affect self-reporting of pain, such as concern about the cost of medication.

Clinicians' attitudes, beliefs, and experiences also influence pain management, with addiction, tolerance, side effects, and regulations being the most important concerns [222,223,226,228,229,230]. A lack of appropriate education and training in the assessment and management of pain has been noted to be a substantial contributor to ineffective pain management [226,228,230,231]. As a result, many clinicians, especially primary care physicians, do not feel confident about their ability to manage pain in their patients [226,228].

Clinicians require a clear understanding of available medications to relieve pain, including appropriate dosing, safety profiles, and side effects. If necessary, clinicians should consult with pain specialists to develop an effective approach.

Legal and Ethical Issues Related to the Treatment of Pain

Fear of license suspension for inappropriate prescribing of controlled substances is also prevalent, and a better understanding of pain medication will enable physicians to prescribe accurately, alleviating concern about regulatory oversight. Physicians must balance a fine line; on one side, strict federal regulations regarding the prescription of schedule II opioids (morphine, oxycodone, methadone, hydromorphone) raise fear of Drug Enforcement Administration investigation, criminal charges, and civil lawsuits [222,232]. Careful documentation on the patient's medical record regarding the rationale for opioid treatment is essential [232]. On the other side, clinicians must adhere to the American Medical Association's Code of Ethics, which states that failure to treat pain is unethical. The code states, in part: "Physicians have an obligation to relieve pain and suffering and to promote the dignity and autonomy of dying patients in their care. This includes providing effective palliative treatment even though it may foreseeably hasten death" [233]. In addition, the American Medical Association Statement on End-of-Life Care requires that physicians "reassure the patient and/or surrogate that all other medically appropriate care will be provided, including aggressive palliative care and appropriate symptom management, if that is what the patient wishes" [234].

Physicians should consider the legal ramifications of inadequate pain management and understand the liability risks associated with both inadequate treatment and treatment in excess. The undertreatment of pain carries a risk of malpractice liability, and this risk is set to increase as the general population becomes better educated about the availability of effective approaches to pain management at the end of life. Establishing malpractice requires evidence of breach of duty and proof of injury and damages. Before the development of various guidelines for pain management, it was difficult to establish a breach of duty, as this principle is defined by nonadherence to the standard of care in a designated specialty. With such standards now in existence, expert medical testimony can be used to demonstrate that a practitioner did not meet established standards of care for pain management. Another change in the analysis of malpractice liability involves injury and damages. Because pain management can be considered as separate from disease treatment and because untreated pain can lead to long-term physical and emotional damage, claims can be made for pain and suffering alone, without wrongful death or some other harm to the patient [235].

The proper storage and disposal of prescription pain medications should also be considered. Taking steps to ensure that medications are stored and destroyed securely and safely can help prevent unintended overdose and substance abuse. In 2010, the U.S. Senate passed the Secure and Responsible Drug Disposal Act, which amended the Controlled Substances Act to permit the take-back disposal of medications by authorized persons (rather than the patient with the prescription) [236]. As such, healthcare professionals may be required to dispose of drugs returned by patients in addition to drug samples that have expired or are not being dispensed. For best practice guidelines on the disposal of medications by patients or healthcare professionals, please visit the Drug Enforcement Administration Office of Diversion Control at https://www.deadiversion.usdoj.gov/drug_disposal [237].

Patients with History of Substance Abuse

The population of people with a history of substance abuse presents challenges to the effective use of pain medication, with issues related to trust, the appropriate use of pain medications, interactions between illicit drugs and treatment, and compliance with treatment. The issues differ depending on whether substance abuse is a current or past behavior.

With active substance abusers, it is difficult to know if patients' self-reports of pain are valid or are drug-seeking behaviors. It has been recommended that, as with other patients at the end of life, self-reports of pain should be believed [67,227]. A multidisciplinary approach, involving15 psychiatric professionals, addiction specialists, and, perhaps, a pain specialist, is necessary. To decrease the potential for the patient to seek illicit drugs for pain, an appropriate pain management plan should be implemented and the patient should be reassured that pain can be managed effectively [67,227]. When planning treatment, the patient's tolerance should be considered; higher doses may be needed initially, and doses can be reduced once acute pain is under control. Long-acting pain medications are preferred for active substance abusers, and the use of nonopioids and co-analgesics can help minimize the use of opioids. Setting limits as well as realistic goals is essential and requires establishing trust and rapport with the patient and caregivers.

Establishing trust is also essential for patients with former substance abuse behavior, who often must be encouraged to adhere to a pain management program because of their fears of addiction. Involving the patient's drug counselor is beneficial, and other psychologic clinicians may be helpful in assuring the patient that pain can be relieved without addiction. Recurrence of addiction is low, especially among people with cancer, but monitoring for signs of renewed abuse should be ongoing [227].

Patients who are following a methadone maintenance program may also fear effective pain management as a risk for recurrent abuse. Two approaches may be followed for these patients: they may receive an increased dose of methadone as the pain reliever or they may be given other opioids along with the same methadone dose, with the dose of the opioid titrated for effective pain relief [67,227]. Again, involvement of the drug counselor is important.

The prevalence of pain at the end of life has been reported to range from 8% to 96%, occurring at higher rates among people with cancer than among adults with other life-limiting diseases [211,238].

Pain can be caused by a multitude of factors and is usually multidimensional, with pain frequently being exacerbated by other physical symptoms and by psychosocial factors, such as anxiety or depression [223].

Pain should be assessed routinely, and frequent assessment has become the standard of care [223]. Pain is a subjective experience, and multidimensional in nature, and although patients' self-reporting of pain does not always correlate with objective functional measures, the patient's self-report of pain is the most reliable indicator [239,240]. Research has shown that pain is underestimated by healthcare professionals and overestimated by family members [223,241]. Therefore, it is essential to obtain a pain history directly from the patient, when possible, as a first step toward determining the cause of the pain and selecting appropriate treatment strategies. When the patient is unable to communicate verbally, other strategies should be used to determine the characteristics of the pain, as will be discussed.

Questions should be asked to elicit descriptions of the pain characteristics, including its location, distribution, quality, temporal aspect, and intensity. In addition, the patient should be asked about aggravating or alleviating factors. Pain is often felt in more than one area, and physicians should attempt to discern if the pain is focal, multifocal, or generalized. Focal or multifocal pain usually indicates an underlying tissue injury or lesion, whereas generalized pain could be associated with damage to the central nervous system. Pain can also be referred, usually an indicator of visceral pain.

The quality of the pain refers to the sensation experienced by the patient, and it often suggests the pathophysiology of the pain [223]. Pain that is well localized and described as aching, throbbing, sharp, or pressure-like is most likely somatic nociceptive pain. This type of pain is usually related to damage to bones and soft tissues. Diffuse pain that is described as squeezing, cramping, or gnawing is usually visceral nociceptive pain. Pain that is described as burning, tingling, shooting, or shock-like is neuropathic pain, which is generally a result of a lesion affecting the nervous system.

Temporal aspects of pain refer to its onset: acute, chronic, or "breakthrough." A recent onset characterizes acute pain, and there are accompanying signs of generalized hyperactivity of the sympathetic nervous system (diaphoresis and increased blood pressure and heart rate). Acute pain usually has an identifiable, precipitating cause, and appropriate treatment with analgesic agents will relieve the pain. When acute pain develops over several days with increasing intensity, it is said to be subacute. Episodic, or intermittent, pain occurs during defined periods of time, on a regular or irregular basis. Chronic pain is defined as pain that persists for at least three months beyond the usual course of an acute illness or injury. Such pain is not accompanied by overt pain behaviors (grimacing, moaning) or evidence of sympathetic hyperactivity.

"Breakthrough" is the term used to describe transitory exacerbations of severe pain over a baseline of moderate pain [242]. Breakthrough pain can be incident pain or pain that is precipitated by a voluntary act (such as movement or coughing), or can occur without a precipitating event. Breakthrough pain occurs in as many as 90% of people with cancer or in hospice settings and is often a consequence of inadequate pain management [222].

Documentation of pain intensity is key, as several treatment decisions depend on the intensity of the pain. For example, severe, intense pain requires urgent relief, which affects the choice of drug and the route of administration [4,223]. The numeric rating scale is the tool used most often to assess pain; with this tool, patients rate pain on a scale of 0 to 10 [223]. Visual analogue scales (patients rate pain on a line from 0 to 10) and verbal rating scales, which enable the patient to describe the pain as "mild," "moderate," or "severe," have also been found to be effective. Some patients, however, may have difficulty rating pain using even the simple scales. In an unpublished study involving 11 adults with cancer, the Wong-Baker FACES scale, developed for use in the pediatric setting, was found to be the easiest to use among three pain assessment tools that include faces to assess pain [243].

Functional assessment is important. The healthcare team should observe the patient to see how pain limits movements and should ask the patient or family how the pain interferes with normal activities. Determining functional limitations can help enhance patient compliance in reporting pain and adhering to pain-relieving measures, as clinicians can discuss compliance in terms of achieving established functional goals [227]. The Memorial Pain Assessment Card can be used to evaluate both the severity of pain and the effect of pain on function [223,244].

Physical examination can be valuable in determining an underlying cause of pain. Examination of painful areas can detect evidence of trauma, skin breakdown, or changes in osseous structures. Auscultation can detect abnormal breath or bowel sounds; percussion can detect fluid accumulation; and palpation can reveal tenderness. A neurologic examination should also be carried out to evaluate sensory and/or motor loss and changes in reflexes. During the examination, the clinician should watch closely for nonverbal cues that suggest pain, such as moaning, grimacing, and protective movements. These cues are especially important when examining patients who are unable to verbally communicate about pain.

Strong evidence supports pain management approaches for people with cancer, but the evidence base for management of pain in people with other life-limiting diseases is weak [47,54,108,187,212,214]. Effective pain management involves a multidimensional approach involving pharmacologic and nonpharmacologic interventions that are individualized to the patient's specific situation [223].

Pharmacologic Interventions

The WHO analgesic ladder, introduced in 1986 and disseminated worldwide, remains recognized as a useful educational tool but not as a strict protocol for the treatment of pain. It is intended to be used only as a general guide to pain management [245]. The three-step analgesic ladder designates the type of analgesic agent based on the severity of pain (Figure 6) [245]. Step 1 of the WHO ladder involves the use of nonopioid analgesics, with or without an adjuvant (co-analgesic) agent, for mild pain (pain that is rated 1 to 3 on a 10-point scale). Step 2 treatment, recommended for moderate pain (score of 4 to 6), calls for a weak opioid, which may be used in combination with a step 1 nonopioid analgesic for unrelieved pain. Step 3 treatment is reserved for severe pain (score of 7 to 10) or pain that persists after Step 2 treatment. Strong opioids are the optimum choice of drug at Step 3. At any step, nonopioids and/or adjuvant drugs may be helpful.

THE WORLD HEALTH ORGANIZATION'S THREE-STEP LADDER OF ANALGESIA

Figure 6

Source: [245]

The WHO ladder is also accompanied by five guiding principles [245]:

The pharmacologic treatment of pain involves selecting the right drug(s) at the right dose, frequency, and route, and managing side effects [223]. A decision pathway was developed for use in the cancer setting and can be applied to other settings (Figure 7) [223].

DECISION PATHWAY FOR PAIN MANAGEMENT

Figure 7

Source: [223] Reprinted, with permission from Dalal S, Bruera E. Assessment and management of pain in the terminally ill. Prim Care Clin Office Pract. 2011;38:195-223.

Nonopioid analgesics, such as aspirin, acetaminophen (Tylenol), and nonsteroidal anti-inflammatory drugs (NSAIDs), are primarily used for mild pain (Step 1 of the WHO ladder) and may also be helpful as co-analgesics at Steps 2 and 3. Acetaminophen is among the safest of analgesic agents, but it has essentially no anti-inflammatory effect. Toxicity is a concern at high doses, and the maximum recommended dose is 3–4 g per day [223]. Acetaminophen should be avoided or given at lower doses in people with a history of alcohol abuse or renal or hepatic insufficiency [223].

NSAIDs are most effective for pain associated with inflammation. Among the commonly used NSAIDs are ibuprofen (Motrin, Advil), naproxen (Aleve, Naprosyn), and indomethacin (Indocin). There are several classes of NSAIDs, and the response differs among patients; trials of drugs for an individual patient may be necessary to determine which drug is most effective [67]. NSAIDs inhibit platelet aggregation, increasing the risk of bleeding, and also can damage the mucosal lining of the stomach, leading to gastrointestinal bleeding. There is a ceiling effect to the nonopioid analgesics; that is, there is a dose beyond which there is no further analgesic effect. In addition, many side effects of nonopioids can be severe and may limit their use or dosing.

Moderate pain (Step 2) has often been treated with analgesic agents that are combinations of acetaminophen and an opioid, such as codeine, oxycodone, or hydrocodone. However, it is now recommended that these combination drugs be avoided, as limits on the maximum dose of acetaminophen limits the use of a combination drug [206,223]. Individual drugs in combination is preferred, allowing for increases in the dose of the opioid without increasing the dose of the co-analgesic.

Strong opioids are used for severe pain (Step 3) [108,187,214,223]. Morphine, buprenorphine, oxycodone, hydromorphone, fentanyl, and methadone are the most widely used Step 3 opioids in the United States. Unlike nonopioids, opioids do not have a ceiling effect, and the dose can be titrated until pain is relieved or side effects become unmanageable. For an opioid-naïve patient or a patient who has been receiving low doses of a weak opioid, the initial dose of a Step 3 opioid should be low, and, if pain persists, the dose may be titrated up daily until pain is controlled. Typical starting doses for patients who are opioid-naïve have been noted, but these doses should be used only as a guide, and the initial dose, as well as titrated dosing, should be done on an individual basis (Table 8). Guidelines suggest that the most appropriate dose is the one that relieves the patient's pain throughout the dosing interval without causing unmanageable side effects [187,206,246].

More than one route of opioid administration will be needed by many patients during end-of-life care, but in general, opioids should be given orally, as this route is the most convenient and least expensive. The transdermal route is preferred to the parenteral route, although dosing with a transdermal patch is less flexible and so may not be appropriate for patients with unstable pain [223]. Intramuscular injections should be avoided because injections are painful, drug absorption is unreliable, and the time to peak concentration is long [223].

Morphine is considered to be the first-line treatment for a Step 3 opioid [206]. Morphine is available in both immediate-release and sustained-release forms, and the latter form can enhance patient compliance. The sustained-release tablets should not be cut, crushed, or chewed, as this counteracts the sustained-release properties. Morphine should be avoided in patients with severe renal failure [214].

Buprenorphine (Butrans) has the general structure of morphine but differs from it in several ways [250]. The transdermal formulation of the drug was approved in 2010 for moderate-to-severe chronic pain in patients requiring an around-the-clock opioid for an extended period [223]. It may be used for people with renal impairment but is contraindicated in patients who have substantial respiratory depression [247,250].

The sustained-release form of oxycodone (OxyContin) has been shown to be as safe and effective as morphine for cancer-related pain, and it may be associated with less common side effects, especially hallucinations and delirium [251]. Oxycodone is also available in an immediate-release form (Roxicodone). Oxycodone should be used in people with advanced chronic kidney disease only if alternative options are not available [214]. If the drug must be used, the intervals between doses should be increased, and the patient should be monitored closely [214].

Hydromorphone and fentanyl are the most potent opioids; neither drug should be given to an opioid-naïve patient. Hydromorphone, which is four times as potent as morphine, is available in immediate- and extended-release forms [252]. Fentanyl is the strongest opioid (approximately 80 times the potency of morphine) and is available as a transdermal drug-delivery system (Duragesic; Ionsys), buccal film (Onsolis), tablet (Fentora), nasal spray (Lazanda), sublingual spray (Subsys), sublingual tablet (Abstral), and lozenge (Actiq) [247,253]. Fentanyl preparations have a more rapid onset than other opioids given nonparenterally [223]. Because of its potency, fentanyl must be used with extreme care, as deaths have been associated with its use. Physicians must emphasize to patients and their families the importance of following prescribing information closely, and members of the healthcare team should monitor the use of the drug. Fentanyl, administered subcutaneously, is the recommended choice for patients with advanced chronic kidney disease [214].

The use of methadone to relieve pain has increased substantially over the past few years, moving from a second-line or third-line drug to a first-line medication for severe pain in people with life-limiting diseases [254]. A systematic review showed that methadone had efficacy similar to that of morphine. However, the authors' conclusions were based on low-quality evidence. Other opioids (e.g., morphine, fentanyl) are easier to manage but may be more expensive than methadone in many economies [255]. Physicians must be well educated about the pharmacologic properties of methadone, as the risk for serious adverse events, including death, is high when the drug is not administered appropriately [255,256]. If the dose of methadone is increased too rapidly or administered too frequently, toxic accumulation of the drug can cause respiratory depression and death. Because of the unique nature of methadone, and its long and variable half-life, extreme care must be taken when titrating the drug, and frequent and careful evaluation of the patient is required. Practitioners are advised to consult with a pain or palliative care specialist if they are unfamiliar with methadone prescribing or if individual patient considerations necessitate rapid switching to or from methadone [187].

Meperidine (Demerol) should not be used in the palliative care setting because of limited efficacy and potential for severe toxicity. Agonist-antagonist opioids (nalbuphine [Nubain], butorphanol [Stadol], and pentazocine [Talwin]) are not recommended for use with pure opioids, as they compete with them, leading to possible withdrawal symptoms.

Tapentadol (Nucynta) is a short-acting opioid approved for moderate to severe pain in adults; an extended-release formulation (Nucynta ER) was approved in 2011 for moderate-to-severe chronic pain when an around-the-clock opioid is needed [257]. The drug is associated with a lower incidence of adverse effects than other opioids, and it has been shown to be highly effective for chronic pain conditions but has not been extensively studied in cancer-related pain or the palliative care setting [258]. A 2014 study of 123 patients that had previously received long-term analgesia for cancer-related pain showed tapentadol significantly reduced pain scores and was generally well tolerated; concomitant use of pain medications was also reduced [259].

According to the Institute for Clinical Systems Improvement, there needs to be shared decision-making with the patient about reducing or eliminating opioids to avoid unnecessary complications from long-term opioid use. This involves following and re-evaluating the patient closely, with dose reduction or discontinuation as needed.

(https://www.icsi.org/wp-content/uploads/2020/01/PalliativeCare_6th-Ed_2020_v2.pdf Last Accessed: October 26, 2021)

Level of Evidence: Expert Opinion/Consensus Statement

The most appropriate option for breakthrough pain is an immediate-release opioid taken in addition to the around-the-clock regimen [223]. The fentanyl buccal tablet has been shown to be effective and safe for relieving breakthrough pain in people who are opioid tolerant [187,260,261]. Between January 2011 and January 2012, three forms of fentanyl were approved for breakthrough pain in people with cancer: fentanyl sublingual tablet (Abstral), fentanyl nasal spray (Lazanda), and fentanyl sublingual spray (Subsys). Abstral and Lazanda have since been discontinued [247,252]. As of 2021, the fentanyl lozenge (Actiq) and buccal tablet (Fentora) are also approved for breakthrough cancer pain [252]. For each formula, the initial dose may be repeated once if pain is not relieved adequately after 30 minutes. Patients should wait at least two hours before using the sublingual tablet, buccal film, or the nasal spray for another breakthrough pain episode; the interval is four hours for the sublingual spray, lozenge, or buccal tablet [247,252].

When pain responds poorly to escalated doses of an opioid, other approaches should be considered, including alternative routes of administration, use of alternate opioids (termed opioid rotation or opioid switching), use of adjuvant analgesics, and nonpharmacologic approaches. A process for opioid switching has been established (Figure 8); the first step is to calculate the equianalgesic dose of the new drug (Table 9) [187,206,223]. Additional care is needed when switching to methadone, and conversion ratios have been established (Table 10) [187]. Evidence suggests that the traditionally recommended equianalgesic doses for the fentanyl transdermal patch are subtherapeutic for patients with chronic cancer-related pain, and more aggressive approaches may be warranted (Table 11) [187,223,262].

PROCESS FOR OPIOID SWITCHING

aReduce by a greater percentage if the patient is older, frail, or has significant organ dysfunction. If changing to methadone, reduce the dose by 75%. If changing to transdermal fentanyl, do not reduce the dose and continue the current opioid for 12 to 48 hours.

Figure 8

Source: [187,206,223]

Another approach that has been used for pain management in the cancer setting is combination opioid therapy, or the concurrent use of two strong opioids. The effectiveness of this approach has been evaluated in only two studies, and the combination was morphine and oxycodone or morphine with fentanyl or methadone [263]. The evidence to support a recommendation of combination opioid therapy is weak, and the side effects most likely outweigh the benefit [263].

Opioids are associated with many side effects, the most notable of which is constipation, occurring in nearly 100% of patients. The universality of this side effect mandates that once extended treatment with an opioid begins, prophylactic treatment with laxatives must also be initiated. Tolerance to other side effects, such as nausea and sedation, usually develops within three to seven days. Some patients may state that they are "allergic" to an opioid. It is important for the physician to explore what the patient experienced when the drug was taken in the past, as many patients misinterpret side effects as an allergy. True allergy to an opioid is rare [223]. Opioid rotation may also be done to reduce adverse events.

When opioids are prescribed, careful documentation of the patient's history, examinations, treatments, progress, and plan of care are especially important from a legal perspective. This documentation must provide evidence that the patient is functionally better off with the medication than without [67]. In addition, physicians must note evidence of any dysfunction or abuse.

Adjuvant agents are often used in conjunction with opioids and are usually considered after the use of opioids has been optimized [67]. The primary indication for these drugs is adjunctive because they can provide relief in specific situations, especially neuropathic pain. Examples of adjuvant drugs are tricyclic antidepressants, anticonvulsants, muscle relaxants, and corticosteroids (Table 12) [187,223]. A systematic review found that there was limited evidence to support the use of selective serotonin reuptake inhibitors (SSRIs) for neuropathic pain, but one serotonin-norepinephrine reuptake inhibitor, venlafaxine (Effexor), was found to be effective [264].

Nonpharmacologic Interventions

Several nonpharmacologic approaches are therapeutic complements to pain-relieving medication, lessening the need for higher doses and perhaps minimizing side effects. These interventions can help decrease pain or distress that may be contributing to the pain sensation. Approaches include palliative radiotherapy, complementary/alternative methods, manipulative and body-based methods, and cognitive/behavioral techniques. The choice of a specific nonpharmacologic intervention is based on the patient's preference, which, in turn, is usually based on a successful experience in the past.

Palliative radiotherapy is effective for managing cancer-related pain, especially bone metastases [47,265,266]. Bone metastases are the most frequent cause of cancer-related pain; 50% to 75% of patients with bone metastases will have pain and impaired mobility [265]. External-beam radiotherapy is the mainstay of treatment for pain related to bone metastases. At least some response occurs in 70% to 80% of patients, and the median duration of pain relief has been reported to be 11 to 24 weeks [265]. It takes one to four weeks for optimal therapeutic results [265,266].

However, palliative radiotherapy has become a controversial issue. Although the benefits of palliative radiotherapy are well documented and most hospice and oncology professionals believe that palliative radiotherapy is important, this treatment approach is offered at approximately 24% of Medicare-certified freestanding hospices, with less than 3% of hospice patients being treated [77,78,79]. As previously noted, reimbursement issues present a primary barrier to the use of palliative radiotherapy [77,78,79]. Among other barriers are short life expectancy, transportation issues, patient inconvenience, and lack of knowledge about the benefits of palliative radiotherapy in the primary care community [77,78,266,267].

One study found that more than half (54%) of people use complementary/alternative medicine therapies at the end of life [268]. The most commonly used therapies are massage, music, relaxation techniques, and acupuncture [268,269,270,271,272].

Massage, which can be broadly defined as stroking, compression, or percussion, has led to significant and immediate improvement in pain in the hospice setting [273]. Both massage and vibration are primarily effective for muscle spasms related to tension or nerve injury, and massage can be carried out with simultaneous application of heat or cold. Massage may be harmful for patients with coagulation abnormalities or thrombophlebitis [227].

Focused relaxation and breathing can help decrease pain by easing muscle tension. Progressive muscle relaxation, in which patients follow a sequence of tensing and relaxing muscle groups, has enabled patients to feel more in control and to experience less pain and can also help provide distraction from pain. [227]. This technique should be avoided if the muscle tensing will be too painful.

Acupuncture typically provides pain relief 15 to 40 minutes after stimulation. Relief seems to be related to the release of endorphins and a susceptibility to hypnosis [227]. The efficacy of acupuncture for relieving pain has not been proven, as study samples have been small. However, acupuncture has been found to be of some benefit for cancer-related pain when the therapy is given in conjunction with analgesic therapy [274].

Other nonpharmacologic interventions that have been helpful for some patients but lack a strong evidence base include manipulative and body-based methods (such as application of cold or heat, and positioning), yoga, distraction, and music or art therapy. The application of cold and heat are particularly useful for localized pain and have been found to be effective for cancer-related pain caused by bone metastases or nerve involvement, as well as for prevention of breakthrough incident pain [227]. Alternating application of heat and cold can be soothing for some patients, and it is often combined with other nonpharmacologic interventions.

Cold can be applied through wraps, gel packs, ice bags, and menthol. It provides relief for pain related to skeletal muscle spasms induced by nerve injury and inflamed joints. Cold application should not be used for patients with peripheral vascular disease. Heat can be applied as dry (heating pad) or moist (hot wrap, tub of water) and should be applied for no more than 20 minutes at a time, to avoid burning the skin. Heat should not be applied to areas of decreased sensation or with inadequate vascular supply, or for patients with bleeding disorders.

Changing the patient's position in the bed or chair may help relieve pain and also helps minimize complications such as decubitus ulcers, contractures, and frozen joints. Members of the healthcare team as well as family members and other informal caregivers can help reposition the patient for comfort and also perform range-of-motion exercises. Physical and occupational therapists can recommend materials, such as cushions, pillows, mattresses, splints, or support devices.

Hatha yoga is the branch of yoga most often used in the medical context, and it has been shown to provide pain relief for patients who have osteoarthritis and carpal tunnel syndrome but it has not been studied in patients at the end of life. Yoga may help relieve pain indirectly in some patients through its effects on reducing anxiety, increasing strength and flexibility, and enhancing breathing [275]. Yoga also helps patients feel a sense of control.

Methods to provide distraction from pain come in a wide variety of methods, including reciting poetry, meditating with a calm phrase, watching television or movies, playing cards, visiting with friends, or participating in crafts.

Music therapy and art therapy are also becoming more widely used as nonpharmacologic options for pain management. Listening to music has been shown to decrease the intensity of pain and reduce the amount of opioids needed, but the magnitude of the benefit was small [276]. Research suggests that art therapy contributes to a patient's sense of well-being [277]. Creating art helps patients and families to explore thoughts and fears during the end of life. An art therapist can help the creators reflect on the implications of the art work. Art therapy is especially helpful for patients who have difficulty expressing feelings with words, for physical or emotional reasons.

A sense of fatigue and weakness is one of the most common symptoms near the end of life, and patients often consider this symptom to be more troublesome than pain [287,288]. The prevalence of fatigue has been reported to range from 12% to 97% of patients with life-limiting diseases, and the prevalence is fairly consistent across disease settings [211,279].

Among the most common contributors to fatigue in people with advanced life-limiting diseases are medications, anemia, dehydration, direct tumor effects on energy consumption and supply, infection, metabolic disturbances, fluid and electrolyte imbalance, dyspnea, sleep apnea, depression, and loss of skeletal muscle due to cachexia [108,278,279].

Ensuring adequate management of symptoms related to fatigue may help in preventing the condition. Clinicians should advise the patient to conserve energy as much as possible, to follow a normal sleep cycle, and to engage in aerobic exercise [108,278,279,286].

Assessing fatigue can be a challenge, but as with pain, the patient's report of how he or she is feeling is the gold standard in the assessment. For patients who speak a language other than English, questions about fatigue should include such words and phrases as "tired," "weak," and "lack of energy," as the word "fatigue" may translate differently in some languages [279]. Several tools are available to assess fatigue, but because it usually occurs in a cluster of symptoms, many of these tools are multidimensional instruments, often involving several questions, which can be impractical [279,284]. In assessing patients for fatigue, the clinician should ask such questions as "Do you feel unusually tired or weak?" or "How tired/weak are you?" [279].

An easy-to-use instrument is the Brief Fatigue Inventory, which includes nine items that ask the patient to rate the severity of fatigue on a scale of 0 (no fatigue) to 10 ("as bad as you can imagine") [289]. The patient is asked to consider the current level of fatigue as well as fatigue experienced within the past 24 hours and to indicate the degree to which fatigue has interfered with activities, mood, walking ability, relations with other people, and enjoyment of life.

Assessment should also include a physical examination to detect an underlying cause of fatigue, a focused history-taking, and laboratory tests, as appropriate, to rule out suspected causes (Figure 9) [279].

ALGORITHM FOR THE DIAGNOSIS OF FATIGUE IN PATIENTS RECEIVING PALLIATIVE CARE

ACTH = adrenocorticotropic hormone, Ca = calcium, CRP = C-reactive protein, Mg = magnesium, NRS = numerical rating scale, Phos = phosphate, TSH = thyroid-stimulating hormone.

Figure 9

Source: [279] Reprinted, with permission from Radbruch L, Strasser F, Elsner F, et al. Fatigue in palliative care patients—an EAPC approach. Palliat Med. 2008;22(1):13-32.

Little evidence is available to support guidelines for the management of fatigue during the end of life. Most of the research on nonpharmacologic and pharmacologic treatment options has been conducted with subjects receiving active cancer treatment or long-term follow-up care after cancer treatment. Fatigue in the palliative care setting is addressed specifically by the European Association for Palliative Care (EAPC) (all settings) and the NCCN (cancer setting) and is noted in guidelines for palliative care for advanced heart failure [108,187,279]. In addition, the Agency for Healthcare Research and Quality has addressed fatigue in the cancer setting, and systematic reviews have been done to help determine effective pharmacologic and nonpharmacologic interventions [281,290,291,292,293]. Management of fatigue should include treatment of an underlying cause, if one can be identified, but symptomatic relief should also be provided (Figure 10) [108,187,279].

ALGORITHM FOR THE MANAGEMENT OF FATIGUE IN PATIENTS RECEIVING PALLIATIVE CARE

Figure 10

Source: [279] Reprinted, with permission from Radbruch L, Strasser F, Elsner F, et al. Fatigue in palliative care patients—an EAPC approach. Palliat Med. 2008;22(1):13-32.

When medications are the underlying cause of the fatigue, nonessential medications should be discontinued, and changing medications or the time of dosing may reduce tiredness during the day. Appropriate management of infection, cachexia, depression, and insomnia may also help reduce fatigue [279,287]. The patient's life expectancy and preferences should be considered before carrying out treatment of an underlying cause [279]. Fatigue may provide a protective effect for patients in the last days or hours of life [279]. As such, the patient may be more comfortable without aggressive treatment of fatigue during that period [279].

The treatment of anemia as an underlying cause of fatigue (and other symptoms) is a complex issue. Many studies have provided evidence to recommend the use of erythropoiesis-stimulating agents (erythropoietin [Epogen], darbepoetin [Procrit]) for anemia in people with cancer, HIV/AIDS, chronic kidney disease, and heart failure because of benefit in increasing the hemoglobin level, improving exercise tolerance, reducing symptoms, and decreasing the need for blood transfusions [279; 290; 294; 295; 296]. However, safety concerns led the U.S. Food and Drug Administration (FDA) to require a boxed warning on the label of erythropoiesis-stimulating agents regarding the increased risk of several adverse events (death, myocardial infarction, stroke, venous thromboembolism, thrombosis of vascular access, and tumor progression or recurrence) among people with chronic kidney disease or cancer [297]. The FDA recommends using the lowest dose sufficient to avoid red blood cell transfusion [297]. Recommendations for these agents in these populations have been withdrawn or revised [87; 291; 298; 299]. A 2010 systematic review and meta-analysis (11 studies, 794 subjects) demonstrated benefit of erythropoiesis-stimulating agents among people with heart failure and mild anemia (>10 g/dL) with no increase in adverse events [296].

Most patients will try to manage fatigue by resting and/or sleeping more often, and many healthcare professionals will also recommend this strategy. However, additional rest and/or sleep usually does not restore energy in patients who have fatigue related to a life-limiting disease; continued lack of exercise may even promote fatigue [279]. Regular aerobic exercise and strength training has been found to alleviate fatigue, although much of the research in this area has been conducted with cancer survivors [284]. For example, a meta-analysis (28 studies, 2,083 subjects) demonstrated a significant effect of exercise in the treatment of fatigue during and after cancer treatment [293]. An update to this review and meta-analysis supported the benefit of aerobic exercise for individuals with cancer-related fatigue and recommended further research to determine the optimal type, intensity, and timing of an exercise intervention [300]. Some small studies of fatigue have been done in the palliative care setting, and exercise was found to be beneficial [301,302,303].

Although an exercise program is recommended, decreasing activity to conserve energy is also encouraged [108,187,279]. Clinicians should talk to the patient and family about the importance of the patient conserving energy by adjusting daily activities to correspond to times of peak energy, setting priorities for activities, following a normal wake-sleep cycle, and using assistive devices, and delegating less important tasks [187,279]. Encouraging adequate nutrition, stress reduction through meditation or relaxation techniques, and engagement in enjoyable activities can help restore energy [187,279]. Counseling about setting realistic goals for activities and function may also help patients and family members adapt to new daily routines.

Pharmacologic treatment of fatigue should be undertaken only after potential causes of fatigue have been ruled out [187,279]. The EAPC and the ASCO note that methylphenidate (Ritalin) and modafinil (Provigil) may reduce fatigue [279,284]. The NCCN does not recommend modafinil due to limited evidence of benefit and recommends that methylphenidate be used cautiously and should be considered only after other causes of fatigue have been ruled out [278]. The recommendations from the EAPC and the ASCO are based on systematic reviews showing a significant effect of methylphenidate for the treatment of fatigue in people with cancer or HIV/AIDS or for opioid-induced sedation [279,290,291,292,304]. An optimal dose of methylphenidate has not been defined, but an initial dose of 5–10 mg (given in the morning) has been used, with the dose titrated to 40–60 mg per day (given once in the morning and once at midday) [279]. Among the side effects are nervousness, jitteriness, agitation, arrhythmia, and tachycardia [279]. The initial recommended dose of modafinil is 200 mg per day [279]. Major side effects have included agitation, nervousness, sleep disturbances, nausea, and diarrhea. Since the publication of these recommendations, researchers conducting a systematic review concluded that the evidence was insufficient to recommend a specific drug for the treatment of fatigue in the palliative care setting [292]. The ASCO notes that evidence for their use is weak; others argue that an improper (too low) dose and short study duration leads to suboptimal effect in trials and that individual response to central nervous system stimulants is highly variable [284,305].

Corticosteroids (prednisone and dexamethasone) have been used frequently to treat fatigue in the palliative care setting, but no research on their effectiveness is available [292]. These agents have provided short-term relief of fatigue and improved quality of life among people with cancer, but because of the toxicity associated with long-term use, they should be considered only at the end life or to alleviate fatigue for a well-defined goal (such as allowing the patient to attend a special event) [187,279].

The prevalence of dyspnea among adults with life-limiting disease ranges from 10% to 95%, with the highest rates among people with COPD, lung cancer, and heart failure, especially in the last week of life [54,207,209,211,213,306,307].

Physical causes of dyspnea vary according to the life-limiting disease and/or comorbid conditions and include pleural effusion, airway obstruction, pulmonary embolism, pericardial effusion, and asthma [54]. Pain and psychologic conditions such as anxiety and depression may augment the severity or prolong the duration of dyspnea [54,213].

Measures to reduce anxiety can help to prevent dyspnea or reduce its severity. In addition, patients with heart failure or lung diseases should be advised to conserve energy.

Practice guidelines recommend that clinicians regularly assess dyspnea in patients receiving end-of-life care [47,54,213]. Assessment should involve asking the patient to note the severity and/or distress related to dyspnea, as objective testing, such as respiratory rate, arterial blood gas levels, and pulse oximetry, do not always correlate with a patient's experience of shortness of breath [213]. Tools for patient-reported dyspnea include a modified Borg scale, a visual analog scale, or a numerical scale [213,308]. In addition to asking about the severity of breathlessness, the clinician should ask about other symptoms, especially concurrent chest pain, and about the activities that cause dyspnea. Patients with dyspnea often modify their activities to avoid dyspnea, so the clinician should ask the patient if he or she has changed or stopped any activities because of dyspnea [54]. Because of the link between psychologic factors and dyspnea, the clinician should also evaluate the patient's psychosocial status.

Physical assessment of the patient should include evaluation of breath sounds, heart rate, respiratory rate, jugular pressure, and functional status. Testing should be done to identify a suspected underlying cause of dyspnea [54,122].

The American College of Physicians, the American Thoracic Society, the Canadian Thoracic Society (endorsed by the ACCP), and the NCCN have developed evidence-based guidelines for the management of dyspnea [47,54,122,213,309,310]. In addition, evidence-based recommendations for managing dyspnea in people with advanced heart failure are available [108]. A stepwise approach to managing dyspnea should be taken, with the first step being treatment of the underlying cause, if one can be identified [54]. Nonpharmacologic interventions should be used first; if the response is inadequate, pharmacologic interventions may be added.

Supplemental oxygen is commonly used to treat dyspnea. Strong evidence supports the use of oxygen and pulmonary rehabilitation for dyspnea, and supplemental oxygen may provide relief of dyspnea for people with advanced lung or heart disease who have hypoxemia at rest or with minimal activity [47,54,212,213,309,310]. However, data suggest that oxygen offers no benefit to patients who do not have hypoxemia [108].

A variety of nonpharmacologic interventions have been suggested in several practice guidelines, although the evidence base varies (Table 13) [122,213,309,310]. In a systematic review of nonpharmacologic interventions and an update of that review for dyspnea in people with advanced malignant and nonmalignant diseases, there was strong evidence for chest wall vibration and neuroelectrical muscle stimulation and moderate evidence for walking aids and breathing training [311,312]. The updated review found low strength of evidence for acupuncture/acupressure, no evidence for the use of music, and insufficient evidence to recommend the use of a fan, music, relaxation, counseling and support, and psychotherapy [311,312]. A subsequent small randomized controlled trial demonstrated that a handheld fan directed at the face reduced breathlessness [313].

Opioids represent the primary recommended pharmacologic intervention for severe dyspnea in people with advanced cancer and lung disease [47,213,309]. A systematic review and meta-analysis (18 randomized controlled trials) demonstrated a significant positive effect of opioids on breathlessness [315]. Guidelines recommend that oral or parenteral opioids be considered for all patients with severe and unrelieved dyspnea; nebulized opioids have not had an effect when compared with placebo [47,212,213,309]. Oral morphine is the most commonly prescribed opioid, but other opioids, such as diamorphine, dihydrocodeine, fentanyl, hydromorphone, and oxycodone, may be used [213]. The dose should be selected and titrated according to such factors as renal, hepatic, and pulmonary function and past use of opioids [213]. An oral dose of morphine of 2.5–10 mg every four hours as needed (1–5 mg intravenously) has been recommended for opioid-naïve patients [122]. Although respiratory depression is a side effect associated with opioids, especially morphine, this effect has not been found with doses used to relieve dyspnea [122,316]. Evidence-based recommendations for palliative care for people with heart failure note that diuretics represent the cornerstone of treatment of dyspnea [108]. Nitrates may also provide relief, and inotropes may be appropriate in select patients [108]. The recommendations also include the use of low-dose opioids [108].

Anxiolytics are often a recommended option for relief of breathlessness because of the association between anxiety and dyspnea. However, anxiolytic agents have not been found to be effective for the management of dyspnea. A systematic review published in 2010 (seven studies, 200 subjects) showed that benzodiazepines had no beneficial effect on breathlessness in people with advanced cancer or COPD [317]. An update of the review included one additional study, but it did not alter the authors' original conclusion that benzodiazepines have no beneficial effect on breathlessness in these patient populations [318]. Bronchodilators and systemic corticosteroids may be helpful in relieving dyspnea in people with lung cancer and underlying obstructive airway disease [314]. In addition, analgesics may help relieve dyspnea associated with pain.

The prevalence of constipation among adults with life-limiting disease ranges from 8% to 70%, and constipation occurs in almost all patients taking opioids [211,227,307]. The prevalence of constipation in palliative care settings is even higher, at 30% to 90% [321].

Opioids are the primary factor in constipation in the palliative care setting, and many other prescribed drugs can contribute to constipation, including tricyclic antidepressants, antacids, antiepileptic drugs, anticholinergic agents, and antihypertensives [319]. Additional factors that may contribute to constipation are diverticuli, inflammatory bowel disease, metabolic conditions (hypercalcemia, hypokalemia, hypothyroidism, uremia), cerebral tumors, dehydration, and radiation fibrosis [319]. For patients with cancer, constipation may be directly due to tumor involvement that causes intestinal obstruction. A diet low in fiber and decreased physical activity also increase the likelihood of constipation.

Prevention of constipation is key, as prophylaxis is more effective than treatment after constipation has been identified. As such, all treatment guidelines strongly recommend that a prophylactic bowel regimen be initiated when treatment with opioids (or other constipation-causing drugs) begins [310,319,320,321]. The recommended prophylaxis is an osmotic and/or a stimulant laxative [310,319,320]. Many nonpharmacologic approaches are recommended, and patients should be encouraged to plan a diet with adequate fiber, to increase fluid intake, and to engage in physical activity, as appropriate [310,319,320,321]. Family members should be asked to help the patient comply with these measures. Ensuring that the patient has sufficient privacy and comfort with toileting is also recommended [319,320,321].

Issues of personal privacy often lead to a reluctance of patients to discuss constipation, so clinicians and other healthcare professionals must initiate the discussion and talk honestly about what to expect and measures to prevent and manage the symptom. The assessment tools used most often are the Bristol Stool Form Scale and the Constipation Assessment Scale [319,320]. Assessment should include a review of the list of medications, a history of bowel habits, and abdominal and rectal examination. In addition to checking the list of prescribed medications to determine if constipation is a side effect, the physician should ask the patient about over-the-counter drugs and herbal remedies, as constipation can be a consequence of aluminum-containing antacids, ibuprofen, iron supplements, antidiarrhea drugs, antihistamines, mulberry, and flax. A detailed history of bowel habits helps to establish what is considered normal for the individual patient. The patient should be asked about frequency of stool, the appearance and consistency of stools, use of bowel medications, and previous occurrence of constipation. In general, physical examination of the abdomen for tenderness, distention, and bowel sounds can rule out intestinal obstruction as the cause of constipation. A rectal examination can identify the presence of stool, fecal impaction, or tumor. Imaging of the abdomen (by plain x-ray or computerized tomography) may be appropriate to confirm the presence of obstruction. Consideration of the patient's prognosis and preferences for care should be factored into a decision to carry out diagnostic testing. As with assessment of all symptoms, constipation should be reassessed frequently; assessment at least every three days is recommended [320].

The goal of treatment should be relief of symptoms related to constipation and re-establishment of bowel habits to the patient's comfort and satisfaction; some recommend a goal of one nonforced bowel movement every one to two days, or at least three times per week [310,320,321]. Systematic reviews have demonstrated that data are insufficient to support one laxative or combination of laxatives over others [319,322,323,324].

Many laxatives are FDA approved for occasional constipation, and much of the evidence on their efficacy has come from studies of chronic constipation, not patients with life-limiting disease. In its guidelines for the management of chronic constipation, the American College of Gastroenterology notes the following [323]:

The results of a systematic review of studies in the palliative care setting also demonstrated insufficient data for recommendations because of a lack of direct comparisons of laxatives [322]. An update to this review includes evidence that subcutaneous methylnaltrexone is effective in inducing laxation in palliative care in patients with opioid-induced constipation where conventional laxatives have failed [324]. Researchers have concluded that the choice of a laxative should be made on an individual basis, with considerations of patient preferences and the side-effect profile [319,322]. For all patients, oral formulations are recommended over rectal suppositories [319,320]. Rectal suppositories and/or enemas should be used with caution in patients receiving chemotherapy due to the increased risk of the rectal route of administration in the setting of cytopenia [310].

European and Canadian consensus groups and the NCCN have developed practice guidelines for constipation in the palliative care setting on the basis of the available data and expert opinion (Figure 11) [310,319,320]. First-line recommended treatment is a stimulant laxative plus a stool softener (PEG or lactulose) [310,319,322]. A small study of senna with and without docusate for hospitalized patients with cancer showed no significant benefit to the addition of docusate; docusate is specifically not recommended in the Canadian consensus recommendations [320,325]. If constipation persists, other options are bisacodyl, magnesium hydroxide, or sorbitol [310]. Methylnaltrexone (Relistor) was approved by the FDA in 2008 for the treatment of opioid-induced constipation [326]. A systematic review indicated that the subcutaneous drug is effective in the palliative care setting, and is especially useful for patients with constipation refractory to conventional laxatives [324]. In 2014, the FDA approved naloxegol, an oral agent for the management of opioid-associated constipation [327]. However, this drug's approval is limited to patients with chronic non-cancer pain. Practice recommendations note that methylnaltrexone or naloxegol should be considered for patients taking opioids after failure of other laxatives [310,319,320]. Withdrawal of opioids should never be a strategy to manage constipation.

ALGORITHM FOR THE MANAGEMENT OF CONSTIPATION IN PATIENTS RECEIVING PALLIATIVE CARE

Figure 11

Source: [319] Reprinted, with permission from Larkin PJ, Sykes NP, Centeno C, et al. The management of constipation in palliative care: clinical practice recommendations. Palliat Med. 2008;22(7):796-807.

Nonpharmacologic interventions are important adjuncts to laxatives, and the interventions used as prophylaxis are recommended for ongoing management [319,320].

Nausea alone affects approximately 6% to 68% of adults with life-limiting disease, and vomiting affects 40% [211]. The rate of nausea and vomiting is highest among patients with cancer [211].

The potential causes of nausea and vomiting near the end of life vary according to life-limiting disease [206,214,246,328,329]:

The causes also differ according to the pathway stimulated (Table 14) [328,329]. Most often the cause is multifactorial, but sometimes no cause can be determined.

The prevention of nausea and vomiting has focused on prophylactic treatment for patients receiving chemotherapy or radiation therapy for cancer. Although most patients at the end of life do not receive anticancer treatment, chemotherapy may be given as part of palliative care. ASCO classifies chemotherapy drugs according to their emetogenic potential: high (>90% incidence of emesis without an antiemetic), moderate (30% to 90% incidence), low (10% to 30% incidence), and minimal (<10% incidence) [330]. According to ASCO guidelines, a 5-hydroxytryptamine type 3 (5-HT3) antagonist, dexamethasone (Decadron), and a neurokinin 1 (NK1) receptor antagonist (such as aprepitant [Emend]) should be used as prophylaxis for a highly emetic chemotherapy agent or combination (such as an anthracycline and cyclophosphamide) [330]. Palonosetron (Aloxi), in combination with dexamethasone or netupitant, is recommended for chemotherapy agents with moderate emetic risk, and dexamethasone is recommended before the first dose of chemotherapy with a low emetic risk. A 5-mg dose of olanzapine has been shown to be safe and effective when used in combination with a 5-HT3 receptor antagonist, an NK1 receptor antagonist, and dexamethasone in patients with high emetic risk [330]. For nausea and vomiting not related to chemotherapy, treatment with regular dosing of an antiemetic will help prevent subsequent episodes of the symptoms.

A detailed history, physical examination, and review of the medication list are essential for planning effective management of nausea and vomiting. In talking with the patient, the clinician should ensure that the patient is actually experiencing nausea, as patients have used the term nausea to describe other feelings, such as pain, distention, abdominal discomfort, and early satiety [67,331]. The clinician should ask about the onset of the nausea, how frequently it occurs, if there are precipitating factors, and if there is a relationship to food intake. It may be helpful to ask the patient to rate the intensity of nausea on a scale similar to a pain scale (a 10-point numerical scale). Because the cause of nausea and vomiting is often multifactorial, a multidimensional assessment is beneficial, with particular attention paid to such other symptoms as pain, appetite, fatigue, depression, and anxiety. The physical examination should include evaluation for signs of cachexia or malnutrition, assessment of the abdomen for evidence of bowel obstruction, increased bowel sounds, and abdominal distention. A plain x-ray of the abdomen (e.g., kidney, ureters, bladder) may distinguish constipation with stool seen in the bowel from malignant bowel obstruction [331]. In addition, a neurologic examination should be done to determine if there are signs of increased intracranial pressure, papilledema, or autonomic insufficiency [67,331]. Diagnostic testing may include laboratory studies to rule out metabolic disorders, renal impairment, or liver failure, or radiographs of the abdomen to determine if there is obstruction.

Nausea and vomiting assessment often leads to no clear etiology or several possible causes [331]. Nausea is often not reported; patients should be asked if they have experienced nausea even if they have not vomited [246].

Evidence-based guidelines for the management of nausea and vomiting unrelated to chemotherapy and radiation are lacking [332]. In addition, most studies of these symptoms and recommendations are related to the cancer setting. In general, experts have recommended that antiemetics be selected on the basis of the emetic pathway and the etiology of the nausea and/or vomiting, but systematic reviews have found that the evidence for recommendations is weak to moderate at best [206,328,329,332,333,334]. One systematic review found no evidence that the choice of antiemetic according to etiology or multiple antiemetics was better than a single antiemetic [332].

Several classes of pharmacologic agents can be used to manage nausea and vomiting; the main classes used in the end-of-life setting are prokinetic agents, dopamine receptor antagonists, antihistamines, anticholinergics, 5-HT3 receptors, and corticosteroids (Table 15) [328,329,334]. A first antiemetic should be scheduled and titrated to efficacy, maximum recommended dose, or dose-limiting side effects. If a first drug does not adequately control nausea and vomiting, a second (and perhaps subsequent agents) with different receptor binding can be added in a stepwise manner [331]. Oral antiemetics are recommended unless the patient is vomiting or has symptomatic gastric stasis. Drugs that block several receptors (e.g., olanzapine) may be advantageous if the nausea and vomiting seems refractory [331].

The prokinetic agent metoclopramide (Reglan) has been recommended as a first-line treatment because of its central and peripheral actions and its effectiveness for many chemical and undetermined causes of nausea [206,246,328,331]. The drug should be used with caution in patients with heart failure, diabetes, and kidney or liver disease; the dose should be reduced by 50% for older patients and those with moderate-to-severe renal impairment [329,331]. Chronic use of prokinetic agents and dopamine receptor antagonists may be associated with the development of tardive dyskinesia, especially in frail, elderly patients [187]. Octreotide (Sandostatin), dexamethasone, and hyoscine hydrobromide (Scopolamine) are recommended for bowel obstruction [92,328,329,331,332]. Ondansetron (Zofran) has been suggested for chronic nausea, but in September 2011, the FDA issued a safety announcement about the drug, noting that it may increase the risk of QT prolongation on electrocardiogram. [329,335]. In 2012, the FDA updated the safety information specifically for the 32-mg IV dose of the drug, and the manufacturer subsequently announced changes to the drug label removing this dose [336]. Haloperidol (Haldol) is recommended for uremia-induced nausea in people with end-stage chronic kidney disease [214]. Dexamethasone is used for nausea and vomiting related to increased intracranial pressure and, although the evidence is limited, it is also used as second-line treatment for intractable nausea and vomiting and as an adjuvant antiemetic [246,328,329]. Olanzapine (Zyprexa), an atypical antipsychotic, has also been effective for nausea that has been resistant to other traditional antiemetics, as well as for opioid-induced nausea [337]. A benzodiazepine (such as lorazepam [Ativan]) may be of benefit if anxiety is thought to be contributing to nausea or vomiting [329].

In addition to pharmacologic management of nausea and vomiting, other supportive approaches include maintenance of oral hygiene, regular baths to reduce unpleasant odors, and small meals at regular intervals [206,246]. Cold foods may be better tolerated than hot foods because of decreased smells.

Anorexia occurs in 21% to 92% of adults with life-limiting disease, with the highest rates found among patients with cancer [207,211,307]. Cachexia has been reported in 16% to 57%, again with the highest rates found among people with cancer [341].

Across life-limiting diseases, anorexia can occur as a result of several other symptoms, such as fatigue, constipation, xerostomia, dysphagia, mucositis, and nausea. Endocrine abnormalities may also be the cause, and psychologic, social, and spiritual distress can affect the desire to eat [187,345]. Changes in taste sensations (leading to food aversions), altered sense of smell, and early satiety have been common among people with cancer and anorexia [345,346].

Studies have shown that multiple factors contribute to cachexia. Abnormal metabolism is thought to lead to a negative protein and energy balance, with subsequent loss of muscle mass [201,339,342,343]. Inflammation, increased neurohormonal activity, insulin resistance, and increased muscle protein breakdown are often associated with cachexia [339,343,347]. The role these factors play in the development of cachexia may differ according to the underlying chronic condition.

Preventive measures for anorexia include effective management of symptoms that are known to have a potential impact on the desire and/or ability to eat. No appropriate measures to prevent cachexia are available.

Guidelines for assessing anorexia and cachexia have been developed for the cancer and HIV settings [310,348]. According to NCCN guidelines, assessment of anorexia and cachexia in patients with cancer include the following [310]:

The guidelines for the assessment of HIV-related wasting recommend the following [348]:

Few evidence-based guidelines for the treatment of anorexia and cachexia are available, primarily because of the lack of studies on these under-recognized conditions and the still-emerging understanding of the causes of cachexia. The first step in managing anorexia is to treat symptoms that interfere with appetite and/or the ability to eat. In addition, nonpharmacologic interventions should be directed at improving enjoyment of food, increasing the sense of well-being, and enhancing a sense of normalcy in daily activities. The patient should be encouraged to try favorite foods, to eat small frequent meals, and to drink high-calorie nutritional supplements [310,347,349,350]. Other interventions include an exercise program, consultation with a nutritionist, swallowing evaluation, and psychiatric consultation (if it is determined that the patient has an eating disorder) [310]. For people with end-stage liver disease and an inadequate caloric intake, protein restriction (to prevent hepatic encephalopathy) should be avoided [201].

Two drugs are FDA approved as appetite stimulants for anorexia associated with life-limiting disease (Table 17). Megestrol acetate is FDA approved for the treatment of anorexia, cachexia, or unexplained weight loss in patients with AIDS [351]. It has become the most widely used drug for these indications for people with other life-limiting diseases, and a meta-analysis of data from studies (involving people with a variety of life-limiting illnesses) demonstrated that megestrol acetate was beneficial, especially with respect to improving appetite and weight gain in people with cancer [351]. Meta-analysis showed a benefit of megestrol acetate compared with placebo, particularly with regard to appetite improvement and weight gain in cancer, AIDS, and other underlying conditions. There was insufficient information to define the optimal dose, but higher doses were more related to weight improvement than lower doses. Side effects (e.g., edema, thromboembolic phenomena) and deaths were more frequent in patients treated with megestrol acetate compared with placebo [351]. Today, use of megestrol is limited due to the increased risk for thromboembolism. Dronabinol (Marinol), an oral cannabinoid, is FDA approved for anorexia associated with weight loss in people with AIDS [326]. Because of its effects, dronabinol should be used with caution for people with cardiac disorders, depression, or a history of substance abuse; people taking concomitant sedatives or hypnotics; and older individuals [326].

In addition to appetite stimulants, metoclopramide, a drug approved for treatment of nausea and vomiting, is recommended for anorexia related to early satiety in people with cancer [310,349].

The treatment of cachexia is more challenging because its pathophysiology is poorly understood and because treatments may differ according to the life-limiting disease. According to the guidelines for cachexia related to cancer and HIV infection, management includes improving nutritional intake, treating disease-related causes of cachexia, treating anorexia, and addressing psychosocial or lifestyle issues [310,348].

Currently, there is no one treatment or combination of treatments that is effective for all patients with cachexia [350]. Increasing oral intake alone is not sufficient, and reversal of wasting may not always be possible; the goal should be to prevent or delay further wasting and functional decline [310,348,350]. As noted, the use of megestrol acetate is effective in increasing weight, but increased nutrition and weight are not sufficient to effectively manage cachexia, and more research is needed to identify agents to increase body mass and to define a multimodal strategy to stop and/or reverse wasting. These strategies may differ according to the underlying chronic disease.

Studies have indicated that recombinant human growth hormone (rhGH) significantly increases lean body mass and improved physical endurance and quality of life in people with HIV [352,353]. In addition, rhGH has shown benefit in cachexia related to pulmonary and cardiac disease [354]. Recombinant somatropin (Serostim) is approved for the treatment of people with HIV with wasting or cachexia; concomitant antiretroviral therapy is necessary [252]. The drug is contraindicated in active neoplasia [252].

The anabolic steroid oxandrolone (Oxandrin) is FDA approved as adjunctive therapy to promote weight gain after weight loss following extensive surgery, chronic infection, or severe trauma and for some patients without a definitive pathophysiologic cause of weight loss [252]. The drug has shown benefit in increasing body weight and lean body mass in cachexia related to HIV and COPD [355,356]. The drug is safe and well tolerated, but more studies are needed to determine its risk-benefit ratio before it can be used more widely [357].

Ghrelin has been evaluated for the treatment of cachexia, and its anti-inflammatory properties may address the proposed role of inflammation in the development of cachexia [358]. The results of small studies have demonstrated that ghrelin increases lean body mass in people with end-stage renal disease, COPD, and heart failure [354]. A 2018 systematic review found insufficient evidence to either support or refute the use of ghrelin for the treatment of cachexia [359]. Again, more research is needed before this agent can become part of clinical practice.

For people with a limited life expectancy (weeks to days), the clinician should provide education and emotional support to the patient, family, and/or caregiver regarding the natural history of anorexia and cachexia, including the following points [310]:

The prevalence of diarrhea among adults with life-limiting disease varies widely, ranging from 3% to 90%, with the highest rates reported among people with HIV infection or AIDS [211].

The most frequent cause of diarrhea in patients receiving palliative care is overuse of laxatives and leakage around a fecal impaction [360]. Other causes are infectious disease or underlying disease in people with HIV/AIDS or metastatic colorectal cancer. Diarrhea is also a side effect of many drugs, including antihypertensives, antacids containing magnesium, some NSAIDs, potassium supplements, quinidine, thiazide diuretics, retroviral agents, prokinetic agents (metoclopramide), and antibiotics [310,360].

No appropriate measures to prevent diarrhea are available.

A detailed history is the cornerstone of assessing patients for diarrhea. The condition is distressful, yet embarrassing, and direct questions should be asked because the patient may not be forthcoming about the symptom. The clinician should ask the patient about the onset of diarrhea, dietary habits, food intolerances, timing of diarrhea in relation to eating, and medications [360,361]. The patient should also describe bowel movements in terms of frequency, color, and consistency. If possible, a stool specimen should be evaluated.

If infectious diarrhea is suspected, a stool sample should be evaluated to identify the causative organism [361,362].

The American Gastroenterological Association developed guidelines for the treatment of chronic disease in the general clinical setting, but no specific guidelines are available for the management of diarrhea in palliative care [362].

Treatment of an underlying condition is the optimal approach to managing diarrhea. The clinician should review the medication list and discontinue or reduce the dose of any medication that may be the cause [360,361].

Nonpharmacologic approaches to managing diarrhea include avoiding gas-forming and bulky foods, hot spices, fats, alcohol, and milk until diarrhea is controlled. The patient should be encouraged to drink plenty of fluids to avoid dehydration; beverages with added electrolytes, such as sports drinks, can help maintain proper electrolyte balance.

Pharmacologic management includes the use of bulk-forming agents, adsorbents, and opioids [361]. Kaolin and pectin (Kaopectate), available over the counter, is a combination of adsorbent and bulk-forming agents. However, it provides modest relief and it may take up to 48 hours to be effective [361]. Loperamide (Imodium) is the drug of choice for diarrhea because its side effect profile is better than that for codeine or diphenoxylate (Lomotil) [361]. The initial dose of loperamide is 4 mg, with an additional 2 mg after each loose stool [361]. The package insert for loperamide notes that the maximum daily dose in a 24-hour period is 16 mg, but doses of up to 54 mg a day have been used as part of palliative care with few adverse events [361]. Octreotide has been effective for profuse secretory diarrhea associated with HIV infection and can be used to treat refractory diarrhea [361]. The use of octreotide for diarrhea in the palliative setting is usually off-label, as the drug is FDA approved for the treatment of diarrhea and flushing associated with metastatic carcinoid tumors [247]. Octreotide is administered as a continuous subcutaneous infusion at a rate of 10–80 mcg/hr until improvement of symptoms [361]. Infectious diarrhea should be treated with an appropriate antibiotic. A systematic review found probiotic agents to be of benefit in the management of acute infectious diarrhea [363]. An update to that review could not confirm whether probiotic agents reduce the duration of diarrhea [364].

Insomnia is common among the general population, and rates reported for adults with life-limiting disease are even higher, ranging from 9% to 83% [201,207,211,367]. The highest rates have been found among patients with end-stage renal disease [211].

The primary difference between insomnia in the general population and in people with life-limiting diseases is that insomnia in the latter group is usually secondary to the life-limiting disease or its symptoms [366]. Overall, uncontrolled pain is the most common contributor to the inability to sleep well [366,367]. Other common physical symptoms such as dyspnea, nocturnal hypoxia, nausea and vomiting, pruritus, and hot flashes are also causes of insomnia. Restless legs syndrome may be a substantial contributor to the disruption of sleep in persons with end-stage renal disease [210,310,368,369].

In addition, many psychologic conditions associated with a life-limiting disease can cause insomnia; depression, anxiety, delirium, spiritual distress, and grief can make it difficult to fall or remain asleep [366]. Insomnia is a side effect of many drugs, most notably corticosteroids, antidepressants, decongestants, opioids, and some antiemetics [310,365,370]. Patients also may have difficulty sleeping because of disruptions in the normal sleep-wake cycle that result from inactivity and napping during the day. Lastly, stimulants, such as caffeine, and alcohol may keep patients from falling asleep easily.

Adequate relief of pain and other symptoms is the mainstay of preventing insomnia. The most effective preventive measure is limiting the amount of time in bed during the day and restricting the amount of daytime sleep [366]. Encouraging patients to increase activity during the day, as tolerated; to adhere to a regular schedule with limited naps; and to avoid caffeine and alcohol in the afternoon and evening can help lead to more healthy sleep patterns.

Few patients with life-limiting diseases report insomnia, and few clinicians pursue sleep symptoms in their patients [366]. Clinicians should obtain a sleep history from all patients, following guidelines developed by the American Academy of Sleep (Table 18) [365]. The Epworth Sleepiness Scale has been recommended as an assessment tool [310,371].

Clinicians should evaluate patients physically as well as psychologically for signs and symptoms that have been identified as contributors to sleep disturbances.

The American Academy of Sleep has developed an evidence-based guideline for the evaluation and management of chronic insomnia in adults and a practice parameter for the psychologic and behavioral treatment of insomnia, but neither offers specific guidelines for managing insomnia at the end of life [365,372]. Nonpharmacologic interventions should be implemented first, with pharmacologic therapy added to the treatment plan if these interventions are not effective [366]. Optimizing sleep habits can be useful, especially if they are begun early in the course of the disease.

The nonpharmacologic approaches used to prevent insomnia are also the primary management strategies. Among the recommended behavioral strategies are the following [365,372]:

Cognitive behavioral therapy has also been reported to be effective when used in combination with behavioral interventions [365,372]. No nonpharmacologic strategy has been found to be superior to another [372]. These interventions are effective and recommended for older individuals and can also be effective for people with life-limiting disease when strategies are individualized according to the patient [366,372].

Several drugs have been approved by the FDA for the treatment of insomnia; the classes of these drugs are sedative-hypnotics and benzodiazepines (Table 19). In addition, antidepressants and antihistamines are often used for insomnia, but this use is off-label.

Among sedative-hypnotics, zolpidem (Ambien) is a short- to intermediate-acting drug used primarily for sleep-onset insomnia [365,373]. Zolpidem is recommended by the NCCN for insomnia as part of palliative care for people with cancer [310]. Another sedative-hypnotic, eszopiclone (Lunesta), is intermediate-acting and is one of only three insomnia medications approved by the FDA for long-term use [373].

There is a limited number of studies regarding the use of benzodiazepines in palliative care [374]. However drugs in this class are the most commonly used drugs for the treatment of short-term insomnia in people with life-limiting disease [366]. Benzodiazepines are effective in decreasing the time needed to fall asleep as well as the likelihood of waking up during the night [366,373]. Their use should be short term, as their long-term efficacy has not been clearly defined, although this issue is not as important for patients with a limited life expectancy [366]. Lorazepam (Ativan) is a recommended drug for insomnia in people with cancer [310]. The long-acting effect of flurazepam (Dalmane) may be of benefit for some patients [366].

The antidepressant trazodone (Desyrel) is the preferred antidepressant for insomnia (although it is not FDA approved for this indication) [252,366]. It is the drug of choice among tricyclic antidepressants because of its shorter half-life and its milder anticholinergic side effects [366]. Antidepressants are especially useful for people who have anxiety or depression.

The most recently (2019) FDA-approved drug for insomnia is lemborexant (DayVigo), an orexin receptor antagonist [252]. The first approved drug of this class, suvorexant (Belsomra), received FDA approval in 2014 [252]. These agents support sleep through inhibition of orexin A and B, which are neuropeptides that promote wakefulness [252,375]. Another drug is ramelteon (Rozerem), a melatonin receptor agonist. This drug is short acting and used primarily for sleep-onset insomnia [252,373]. Ramelteon is FDA approved for long-term use [365].

For insomnia related to restless legs, a systematic review showed that dopamine agonists are effective, with cabergoline (Dostinex) and pramipexole (Mirapex) often having a greater efficacy than levodopa (L-Dopa) [376].

Barbiturates are not recommended for insomnia because of the rapid development of tolerance [366]. Two supplements promoted for sleep enhancement—melatonin and valerian—have not been shown to be effective for managing insomnia [366,377].

Several factors must be considered when treating older patients with insomnia. For example, it has been recommended that benzodiazepines be avoided in older individuals because of side effects such as increased risk for falls, confusion, and "hangover" [373]. However, these side effects must be considered in light of an individual's particular situation and weighed against the benefits [366,373]. Eszopiclone and ramelteon have been studied in older individuals and have a favorable side-effect profile for that population [373]. Lower doses are often recommended for older individuals [365].

The prevalence of delirium among adults in hospice or receiving palliative care ranges from 28% to 80%, occurring most frequently among patients with cancer [382,383]. Terminal delirium is a distinct entity that occurs within the last days or hours of life, and it is estimated to occur in 80% of dying patients [384].

Many factors may cause delirium, and although the cause is usually multifactorial, often no cause is found [385]. In one comprehensive review, the primary contributor to delirium was unrelieved pain [386]. Delirium is also often caused by medications, including several that are used in the end-of-life setting, such as opioids, corticosteroids, benzodiazepines, and NSAIDs, or the sudden withdrawal from alcohol or drugs (particularly benzodiazepines) upon admittance to a hospital or hospice [310,383,384]. In addition, age, cognitive deficits, impaired vision/hearing, emotional stress, depression, and comorbidities are predisposing factors of delirium [381,384].

Because of the substantial influence of unrelieved pain, adequate pain management can help prevent delirium. Prevention strategies are directed at minimizing precipitating factors, which include a high number of medications (more than six), dehydration, decreased sensory input, psychotropic medications, and a change in environment.

The diagnosis of delirium relies on identifying its two features: cognitive impairment and deficits in attention; these features can be assessed with the Mini-Mental State Examination [384]. The Confusion Assessment Method (CAM) is considered to be the gold standard for distinguishing between delirium from other causes of altered mental status, and other tools to evaluate delirium include the Delirium Rating Scale, the Delirium Symptom Interview, and the Memorial Delirium Assessment Scale [384]. Communication with the healthcare team and family is vital in assessing the patient to help determine the onset and course of delirium as well as signs indicative of the syndrome. Some specific ways to help determine if a patient has delirium include [380]:

Clinical assessment and physical examination should also be directed at ruling out underlying causes, such as infection or metabolic abnormalities, and the medication list should be reviewed carefully [310,384].

The treatment of an underlying cause, if identified, is a key step in managing delirium. Whether delirium can be reversed depends on the cause. Delirium caused by psychotropic medications, dehydration, or hypercalcemia is more likely to be reversible than delirium caused by hypoxia, metabolic abnormalities, or nonrespiratory infections [380,387].

Several nonpharmacologic interventions have been successful in preventing and managing delirium (Table 20) [310,380,382,384]. If delirium is refractory to nonpharmacologic measures, medications may be prescribed. Level 1 evidence supports the use of haloperidol and chlorpromazine (Thorazine) (typical antipsychotics), and these drugs have the advantage of being available in formulations that allow for multiple routes of administration and of being the most cost-effective [310,380]. Several systematic reviews have been done to determine the efficacy of antipsychotics for delirium, and although each review has identified only a few well-designed trials, the results have supported the continued use of these drugs (Table 21) [380,384,388,389,390]. One of these reviews focused on patients with terminal illness; the review identified only one small study (30 subjects) eligible for analysis; haloperidol and chlorpromazine were equally effective, but the risk for cognitive impairment was slightly greater with chlorpromazine [388]. A 2020 update to this review included four studies with 399 participants, most with advanced cancer or advanced AIDS and mild- to moderate-severity delirium [391]. The reviewers found no high-quality evidence to either support or refute the use of drug therapy for delirium symptoms in terminally ill adults. Low-quality evidence indicates that risperidone or haloperidol may slightly worsen delirium symptoms compared with placebo, and moderate- to low-quality evidence indicates that these two agents may slightly increase extrapyramidal adverse events for patients with mild- to moderate-severity delirium [391]. In the other reviews, the efficacy of haloperidol was found to be similar to that of olanzapine, risperidone (Risperdal), and quetiapine (Seroquel) (atypical antipsychotics) [389,390]. In two small nonrandomized studies—one involving hospitalized patients with cancer—aripiprazole (Abilify) was safe and effective for the treatment of delirium, especially the hypoactive subtype [392,393]. Mild-to-moderate delirium can be managed with low oral doses of antipsychotics, titrating the dose to optimum relief; higher doses can be used for severe delirium [310,384]. For older patients and those with multiple comorbidities, treatment should begin with lower doses and titration should be slow [380]. Factors to consider when selecting a drug include the side-effect profile, the patient's age and baseline mental status, the time to response, and the subtype of delirium [380]. There is no recommendation regarding the use of other drug classes for delirium in palliative care (e.g., α-2 receptors agonists, cholinesterase inhibitors, melatonergic drugs, psychostimulants) [382].

The goal of treatment is to reach patients' baseline mental state, not to sedate them, and patients should be reassessed frequently until this goal is met [384]. If agitation is refractory to high doses of haloperidol, the antipsychotic lorazepam may be helpful [310,384]. The management of delirium also includes providing support to family, to help them cope with the condition [310,384]. The management of terminal delirium will be discussed later in this course.

Severe anxiety varies widely among adults with life-limiting disease, ranging from 8% to 79%, with the highest rate among patients with cancer [211].

One of the primary causes of anxiety is inadequate pain relief. Anxiety may also be the result of a patient's overwhelming concern about his or her illness, the burden of the illness on the family, and the prospect of death. In addition, anxiety is a potential side effect of many medications, including corticosteroids, metoclopramide, theophylline, albuterol, antihypertensives, neuroleptics, psychostimulants, antiparkinsonian medications, and anticholinergics. Lastly, withdrawal from opiates, alcohol, caffeine, cannabis, and sedatives can result in anxiety, particularly in the first few days of admission [404].

Effective pain management is the best way to prevent anxiety. Also, educating the patient and the family about what to expect over the course of the illness and providing adequate psychologic and spiritual support can help comfort the patient, thereby preventing anxiety.

Family members and friends may be able to provide information about the level of anxiety experienced by the patient currently and in past situations. All members of the healthcare team should evaluate the patient and the clinical record for reversible causes of anxiety, such as those caused by medications or withdrawal syndromes, and should try to distinguish anxiety from delirium, depression, or bipolar disorder [405,406].

Anxiety manifests itself through physical as well as psychologic and cognitive signs and symptoms. These signs and symptoms include dyspnea, paresthesia, tachycardia, chest pain, urinary frequency, pallor, restlessness, agitation, hyperventilation, insomnia, tremors, excessive worrying, and difficulty concentrating.

According to the Institute for Clinical Systems Improvement, short-term psychotherapy modalities (e.g., dignity therapy) can provide reduction in depression and anxiety symptoms at the end of life.

(https://www.icsi.org/wp-content/uploads/2020/01/PalliativeCare_6th-Ed_2020_v2.pdf Last Accessed: October 26, 2021)

Level of Evidence: Expert Opinion/Consensus Statement

Nonpharmacologic approaches are essential for managing anxiety, and the addition of pharmacologic treatment depends on the severity of the anxiety [67,407]. Effective management of pain and other distressing symptoms, such as constipation, dyspnea, and nausea, will also help to relieve anxiety. If the anxiety is thought to be caused by medications, they should be replaced by alternate drugs. Other strategies include psychologic support that allows the patient to explore fears and concerns and to discuss practical issues with appropriate healthcare team members. Relaxation and guided imagery may also be of benefit [408]. A consult for psychologic therapy may be needed for patients with severe anxiety.

When pharmacologic management is deemed necessary, benzodiazepines are generally preferred, and administration on an as-needed basis is usually sufficient [67]. Neuroleptics and tricyclic antidepressants may also be effective (Table 22). For all medications, the initial dose should be low and subsequently titrated to produce the desired effect within the level of tolerance. Benzodiazepines should be given with caution in older patients, as these drugs may harm memory or cause confusion and agitation in patients who have cognitive impairment [409].

The prevalence of depression varies widely among adults with life-limiting diseases, ranging from 3% to 82%, with the highest rate among patients with HIV/AIDS and end-stage liver disease [201,211].

Unrelieved pain is one of the primary risk factors for depression. Other causes within the physical domain include metabolic disorders (hyponatremia or hypercalcemia), lesions in the brain, insomnia, or side effects of medications (corticosteroids or opioids). Many patients with heart failure have comorbidities and polypharmacy, both of which can increase the risk of depression [412,413]. Psychosocial causes include despair about progressive physical impairment and loss of independence, financial stress, family concerns, lack of social support, and spiritual distress.

Adequate management of pain, attention to psychosocial and spiritual well-being, and early referral for mental health or pastoral counseling are the best strategies to prevent depression.

It is important to differentiate grief from depression. Grieving can be an appropriate response to loss, but if the symptoms persist, the Institute for Clinical Systems Improvement recommends that depression be considered.

(https://www.icsi.org/wp-content/uploads/2020/01/PalliativeCare_6th-Ed_2020_v2.pdf Last Accessed: October 26, 2021)

Level of Evidence: Expert Opinion/Consensus Statement

The diagnosis of depression is complicated, as the usual somatic signs of depression—anorexia, sleep disturbances, weight loss, and fatigue—are often symptoms related to the underlying disease or part of the constellation of symptoms experienced by patients with life-limiting disease [227]. Because of this, assessment should focus on psychologic and cognitive symptoms, such as:

A diagnosis of depression requires the presence of at least five depression-related symptoms within the same two-week period, and the symptoms must represent a change from a previous level of functioning [378]. A simple screening tool that has been found to be effective is to ask the patient, "Are you depressed?" or, "Do you feel depressed most of the time?" [227,414,415]. The physician should also discuss the patient's mood and behavior with other members of the healthcare team and family to help determine a diagnosis. Patients who have thoughts of suicide must be assessed carefully. The physician should differentiate between depression and a desire to hasten death because of uncontrolled symptoms [67]. Psychologic counseling should be sought, as well as measures to enhance the management of symptoms.

The effective management of depression requires a multimodal approach, incorporating supportive psychotherapy, cognitive strategies, behavioral techniques, and antidepressant medications [47]. Patients with depression should be referred to mental health services for evaluation, and resultant approaches may include formal therapy sessions with psychiatrists or psychologists or counseling from social workers or pastoral advisors. In addition, physicians can help by having discussions with the patient to enhance his or her understanding of the disease, treatments, and outcomes, and to explore expectations, fears, and goals. Behavioral interventions, such as relaxation techniques, distraction therapy, and pleasant imagery have been effective for patients with mild-to-moderate depression [47].

Strong evidence supports the use of tricyclic antidepressants or SSRIs, along with psychosocial interventions, for the management of depression in patients with cancer [47,212]. Evidence to support the use of specific pharmacologic agents to treat depression in patients with noncancer diagnoses is not as strong, but psychostimulants may also be helpful [47,67,227,416]. The choice of medication depends on the time available for treatment. The most immediate effect (within days) is achieved with a rapid-acting psychostimulant (e.g., dextroamphetamine, methylphenidate); longer times to therapeutic effect are associated with SSRIs (two to four weeks) and tricyclic antidepressants (three to six weeks).

Palliative sedation may be considered when an imminently dying patient is experiencing suffering (physical, psychologic, and/or spiritual) that is refractory to the best palliative care efforts. Terminal restlessness and dyspnea have been the most common indications for palliative sedation, and thiopental and midazolam are the typical sedatives used [310,434,435]. For patients who have advanced kidney disease, midazolam is recommended, but the dose should be reduced because more unbound drug becomes available [214,310]. Before beginning palliative sedation, the clinician should consult with a psychiatrist and pastoral services (if appropriate) and talk to the patient, family members, and other members of the healthcare team about the medical, emotional, and ethical issues surrounding the decision [67,227,310,436,437]. Formal informed consent should be obtained from the patient or from the healthcare proxy.

Physician-assisted death, or hastened death, is defined as active euthanasia (direct administration of a lethal agent with a merciful intent) or assisted suicide (aiding a patient in ending his or her life at the request of the patient) [67]. The following are not considered to be physician-assisted death: carrying out a patient's wishes to refuse treatment, withdrawal of treatment, and the use of high-dose opioids with the intent to relieve pain. The American Medical Association Code of Ethics explicitly states, "Physician-assisted suicide is fundamentally incompatible with the physician's role as healer, would be difficult or impossible to control, and would pose serious societal risks" [438]. Position statements against the use of physician-assisted death have been issued by many other professional organizations, including the NHPCO and the AAHPM [439,440]. The AAHPM states that their position is one of "studied neutrality" [439]. The basis for these declarations is that appropriate hospice care is an effective choice for providing comfort to dying patients.

In 2010, in a first-of-its-kind comprehensive consensus statement, the Heart Rhythm Society in collaboration with the major cardiology, geriatrics, and palliative care societies, emphasized that deactivation of implantable cardioverter-defibrillators is neither euthanasia nor physician-assisted death [441]. The organizations urged clinicians to respect the right of patients to request deactivation.

The NCCN guidelines recommend that physicians explore requests for assisted death and explain to the patient the distinctions among assisted suicide, treatment withdrawal, and aggressive symptom management [310]. Some states have enacted assisted death statutes. State laws vary, and knowledge of your local statutes is necessary.

Grief comprises a range of feelings, thoughts, and behaviors that fall in the realm of the physical, emotional, and social domains [67]. Individuals may have trouble sleeping, changes in appetite, or other physical symptoms or illness. Emotions can include sadness, anxiety, guilt, and anger. Return to work, activities with friends, and taking care of family can be beneficial.

The British Columbia Medical Services Commission asserts that the relationship between the physician and the patient is one of the most potent therapeutic tools for assisting patients who are dealing with grief. Reassurance about the normal pattern of grief and a commitment to supporting the patient in an ongoing way is the mainstay of care. It may involve scheduled follow-up visits as necessary.

(https://www2.gov.bc.ca/assets/gov/health/practitioner-pro/bc-guidelines/palliative3.pdf Last Accessed: October 26, 2021)

Level of Evidence: Expert Opinion/Consensus Statement

Grief counseling for the family and patient should begin when the patient is alive, with a focus on life meaning and the contributions from the patient's family. An understanding of the mediators of the grief response can help physicians and other members of the healthcare team recognize the family members who may be at increased risk for adapting poorly to the loss [443]. These mediators are:

Clinical assessment should be carried out for individuals at risk of complicated grief. Distinguishing between grief and depression can be challenging, as many signs and symptoms are similar. However, the hallmarks of depression are constant and unremitting feelings of worthlessness, hopelessness, helplessness, anhedonia, and suicidal ideation [206].

Satisfactory adaptation to loss depends on "tasks" of mourning [443]. Previous research referred to "stages" of mourning, but the term "task" is now used because the stages were not clear-cut and were not always followed in the same order. The tasks include:

After the patient's death, members of the palliative care team should encourage the family to talk about the patient, as this promotes acceptance of the death. Explaining that a wide range of emotions is normal during the mourning process can help family members understand that experiencing these emotions is a necessary aspect of grieving. Frequent contact with family members after the loved one's death can ensure that the family is adjusting to the loss. Referrals for psychosocial and spiritual interventions should be made as early as possible to optimize their efficacy.

Bereavement support should begin immediately with a handwritten condolence note from the clinician. Such notes have been found to provide comfort to the family [444,445]. The physician should emphasize the personal strengths of the family that will help them cope with the loss and should offer help with specific issues. Attendance at the patient's funeral, if possible, is also appropriate.

How bereavement services are provided through a hospice/palliative care program vary. Programs usually involve contacting the family at regular intervals to provide resources on grieving, coping strategies, professional services, and support groups [227,310]. When notes are sent, family members should be invited to contact the physician or other members of the healthcare team with questions. Notes are especially beneficial at the time of the first holidays without the patient, significant days for the family (patient's birthday, spouse's birthday), and the anniversary of the patient's death. Bereavement services should extend for at least one year after the patient's death, but a longer period may be necessary [6,227].