Update in Hospice and Palliative Care

Summary and main findings

Psychosocial support is a core aspect of palliative care, and an increasingly important quality measure. ³ However, few psychosocial interventions have been rigorously studied. Dignity therapy is an individualized, short-term psychotherapy program developed for patients with life-limiting illness, which has the goal of relieving distress and enhancing patients' end-of-life experience. It involves patients reflecting on the aspects of their lives that matter most to them or for which they want to be remembered. The sessions result in a transcribed, edited document that patients can share or bequeath as they choose. The authors conducted a randomized trial of dignity therapy versus standard palliative care or client supportive care, a form of psychotherapy focused on current issues in a patient's life. The primary outcome was a reduction in symptoms or improvement in quality of life, as measured using standardized scales.

The study sample included 441 patients, a 50% enrollment rate, and the patients were equally allotted among the three treatment arms. More than 90% of the patients in each group had cancer, and their overall performance status was similar: spending most of their time in bed and requiring assistance with self care, though not bed bound. ⁴ The study found no significant differences among the groups on the main outcome measures, including physical symptoms, quality of life, or depression. However, patients in the dignity therapy arm were more likely than others to report that the therapy was helpful, and that it increased their sense of dignity, improved their quality of life, changed how their family saw them, and had been or would be helpful to their family.

Strengths and limitations

The study was a rigorous, three-armed blinded randomized controlled trial of seriously ill patients across three countries: United States, Canada, and Australia. Dignity therapy is a well-developed treatment that has been extensively studied. ⁵ The lack of effect of dignity therapy on the primary outcomes may be related to the low level of baseline distress across all groups at the start of the trial. Alternatively, dignity therapy may not improve the physical and psychosocial outcomes as measured by the standardized scales used in the study. Also, a post-hoc power calculation found that the study might have been underpowered to detect significant differences across treatment groups.

Clinical bottom line

While dignity therapy did not improve patient outcomes on standardized scales, it did improve patients' sense of dignity and self-rated quality of life, and may be helpful to family members.

Summary and main findings

Medical interpretation is a critical factor in the clinical care of seriously ill patients who do not speak the same language as their providers, yet little research has focused on its accuracy. This observational study was conducted in the outpatient practices of 10 medical oncologists associated with nine Australian hospitals. Eligible patients were newly diagnosed with incurable cancer. The first one to two consultations after the patient's diagnosis of metastatic disease were audiorecorded, transcribed, and translated into English by bilingual researchers. Coders categorized each unit of interpreted patient or physician speech as equivalent or not equivalent. For nonequivalent interpretations, they evaluated the impact of the change.

The study sample included 32 patients who had 49 encounters with their oncologist involving an interpreter, including family members (43%), professional in-person (39%), and professional telephone (14%). Patients' primary languages were those most commonly spoken in Australia: Mandarin or Cantonese (63%), Greek (22%), and Arabic (16%). On average, only 60% of the interpreted speech units were categorized as equivalent. Professional interpreters more frequently provided equivalent interpretations, compared to family members, 65% versus 50%, p=0.02. No difference was found in the rate of equivalent interpretations between professional in-person versus telephone interpreters. The coders judged many nonequivalent interpretations, such as omissions of insignificant information, to have no impact (37%). Some nonequivalent interpretations, such as those where the interpreter conveyed a more comforting message, or clarified or simplified information, were perceived to have a positive impact (31%). In 21% of nonequivalent interpretations, a potentially negative impact was assessed, such as the interpreter conveying more or less certainty than the speaker, using euphemisms, using a more authoritative or paternalistic tone, or giving incorrect information. Family and professional interpreters did not differ in the rate of nonequivalent interpretations with a potentially negative impact, 21% versus 26%, p=0.2.

Strengths and limitations

This is the first study in oncology to compare professional in-person, telephone, and family interpreters. Limitations include that the assessment of impact was subjective; the researchers minimized this with a systematic coding scheme and had high agreement within and across coders. Additionally, the study may have been underpowered to detect differences between in-person, telephone, and family interpreters. Though the study setting of Australia raises questions of generalizability, similar concerns have been documented in the United States in intensive care units (ICUs). ⁶ Finally, it will be important for future research to explore the burden that interpreting potentially upsetting information such as bad news places on families.

Clinical bottom line

In outpatient medical oncology consultations, nonequivalent interpretations were common, but only a minority were perceived to have a negative impact on communication. Nonequivalent interpretations with potentially negative impact occurred with similar frequency among professional and family interpreters.

Tulsky JA, Arnold RM, Alexander SC, Olsen MK, Jeffreys AS, Rodriguez KL, Skinner CS, Farrell D, Abernethy AP, Pollak KI: Enhancing communication between oncologists and patients with a computer-based training program: A randomized trial. Ann Intern Med 2011;155:593–601.

Summary and main findings

Provider empathy is linked to patient satisfaction, treatment adherence, and quality of life, yet health care providers universally struggle with responding to patients' emotions. Previously documented communication training-interventions effectively change providers' behavior, yet are costly and require large learner and facilitator time investments. The authors conducted a randomized controlled trial, called Studying Communication in Oncologist-Patient Encounters (SCOPE), which tested the ability of an individually tailored CD-ROM to enhance oncologists' empathic communication. Participants included medical, gynecologic, and radiation oncology physicians and their English-speaking patients with advanced cancer, recruited from oncology clinics at three medical centers in two states. In the pre-intervention phase of the study, all oncologists received a lecture on communication, and the investigators audiorecorded oncologists' encounters with their patients to capture their baseline communication. Oncologists were then randomly assigned to either the control or intervention arm. Intervention oncologists received an individually tailored CD-ROM that included modules on principles of effective communication, recognizing empathic opportunities, responding to empathic opportunities, conveying prognosis, and responding to difficult questions. Each module included audio clips from the oncologists' pre-intervention audiorecordings to illustrate their performance, along with individualized feedback. In the final phase of the study all oncologists had a new set of patient encounters audiorecorded; these were analyzed for empathic communication. Patients received a telephone survey about their perceptions of their oncologists.

The study sample included 48 oncologists who had 264 patient encounters recorded. Empathic statements, such as an oncologist saying, “I can see this is making you angry,” were used twice as frequently in the intervention compared to the control group (0.7 versus 0.4 per encounter, p=0.024). The investigators also measured how oncologists responded to empathic opportunities, such as a patient saying, “I have been really depressed lately.” Oncologists in the intervention group used continuer responses such as, “Tell me more about what is upsetting you” twice as frequently compared to controls (0.4 versus 0.2 per encounter, p=0.028). The patient survey showed that patients rated oncologists very highly in both the intervention and control groups. Patient trust was higher in the intervention compared to the control oncologists (4.7 versus 4.6 on a 5-point scale, p=0.036).

Strengths and limitations

This is a well-designed, novel, theory-based, multisite randomized controlled trial, based on years of research. One limitation is that the data cannot comment on how long the intervention effect will persist, given the one-month follow-up period. A limitation to generalizability is that only oncologists and cancer patients were included. Additionally, although the intervention was less intensive for oncologists than with previously reported training programs, it did require significant infrastructure to deliver. The authors note that with improvements and more widespread adoption of smart phone recording technology, this limitation may be partially overcome.

Clinical bottom line

Reviewing a CD-ROM created from oncologists' own recorded patient encounters increased their use of empathic communication.

Summary and main findings

For symptom management in serious illness, nonpharmacological interventions are appealing because of their lower side-effect profile. Small studies indicate that music improves anxiety and other outcomes. This systematic review evaluated randomized controlled trials of music interventions in cancer patients, compared to usual care, other therapies, and placebos such as headphones or white noise. The review included two intervention types: (1) music medicine, defined as interventions administered by health professionals; and (2) music therapy, defined as interventions implemented by trained music therapists using a systematic approach of assessment, tailoring to the patient, and evaluation. The authors reviewed primary studies with a broad range of psychological (e.g., depression, anxiety) and physical (e.g., pain, nausea) symptoms; physiological parameters (e.g., heart rate, blood pressure, cortisol); social and spiritual factors; and quality of life. Similarly designed studies were included in a meta-analysis, which compared music interventions to usual care.

The literature search identified 30 trials in 7 countries with 1891 participants that met the eligibility criteria. The study settings were general cancer (n=13), chemotherapy and radiation therapy (n=9), procedures and surgery (n=8), and pediatric (n=4). Thirteen studies assessed music medicine and 17 assessed music therapy. Sessions in the included trials lasted a mean of 30–45 minutes, with the number of sessions varying from 1–40. Meta-analysis of 13 trials involving 899 patients showed a medium to large effect of music interventions in reducing anxiety, compared to usual care. Similar analyses showed medium to large effects of music interventions, compared to usual care, in reducing pain and improving mood and quality of life.

Strengths and limitations

This review followed the highly regarded Cochrane methodology and had a particularly expansive study identification and data collection process. It found that music improved central outcomes in palliative care: anxiety, pain, and quality of life. Cochrane criteria rated the quality of evidence as low, because comparing music to usual care leaves open the possibility that some other aspect of the interventions, such as receiving attention, led to improved outcomes. Many music interventions are low in cost and side effect profile and easy to implement, so we may not need higher-quality evidence to implement them in our practice. Music medicine and music therapy interventions could not be compared, because not enough studies with the same outcome were identified. Thus whether interventions using trained music therapists are superior to those implemented by other health professionals remains a significant question.

Clinical bottom line

Music interventions improved multiple outcomes including anxiety, pain, and quality of life in diverse patient populations and study settings.

Summary and main findings

Research about pediatric palliative care consultation is limited. Given the fundamentally different nature of childhood illness, extrapolating data from the adult literature is not appropriate. ⁷ The objective of this study was to describe characteristics and outcomes of pediatric patients who received hospital-based palliative care. The authors conducted a chart review of a cohort of pediatric patients who received consults between January and March of 2008 at six hospitals, five across the Unites States and one in Canada.

Of 515 patients whose charts were reviewed, 17% were <1 year old (including fetal consults), 68% were between 1 and 18 years, and 16% were older than 18. The majority of patients were male (54%); and diagnoses included genetic/congenital conditions (41%), neuromuscular conditions (39%), and cancer (20%). Main indications for the consult included: symptom management (58%), assistance with communication (49%), support with decision making (42%), and coordination of care (35%). The main signs/symptoms encountered during pediatric consults were cognitive impairment, speech impairment, and fatigue or sleep problems. One-third of patients experienced pain. Of patients who received a consult, 12% died within 30 days of the consult; however nearly 75% of the cohort was still alive at one year.

Strengths and limitations

This study helps clinicians better understand the landscape of pediatric palliative care and provides a clear sign of the maturity of the field. The study was a multisite project, which adds to the generalizability of the results. However, it was a retrospective chart review, which can raise issues of validity. Important next steps include describing the types of care provided beyond the initial consult and assessing the effect of palliative care on patient and family outcomes.

Clinical bottom line

The key features of pediatric palliative care include a broad spectrum of diseases, unique pediatric symptomatology, and the potential for a longer relationship with patients and their families.

Summary and main findings

The prevalence of dementia is increasing, and agitation and aggression are common in patients living with severe dementia. ⁸ Antipsychotics, while often used in this population, have significant adverse effects. ⁹ Providers are often reluctant to use opioid medications in patients with dementia. This multicenter clustered, single-blinded, randomized controlled trial tested a stepwise protocol for the pharmacologic management of pain with behavioral disturbances in nursing home residents with dementia. The study took place in 60 nursing homes in 5 towns in Norway. Patients were age 65 or older, had at least moderate dementia and a history of clinically relevant behavioral disturbances. Patients in the nursing home units randomized to the intervention received analgesia for eight weeks as per a standardized protocol based on the American Geriatrics Society guidelines for pain in older adults. ¹⁰ The protocol included (1) beginning standing nonopioid medications in patients who were on no medications or on low-dose nonopioid medications; (2) using a short-acting opioid for patients who were on full-dose nonopioid medications or low-dose opioids; (3) using long-acting, transdermal opioids for patients who were on low-dose long-acting opioids or could not swallow medications; and (4) starting an adjuvant medication for patients with neuropathic pain. Change in agitation was measured using nurses' ratings of a series of behaviors using a standardized scale. ¹¹ Secondary outcomes included changes in aggression, pain, cognition, and activities of daily living.

Across all sites, 352 residents were enrolled in 60 clusters. The majority of patients were in their mid eighties; nearly 75% were women. The level of agitation was similar and clinically significant in both groups at baseline. Patients treated with the pain medication algorithm had decreases in their overall aggression and pain scores. These scores returned to a level similar to their counterparts in the control group four weeks after the study ended, demonstrating that the effect was most likely related to the pain medications. There was also a relationship between having a lower pain score and less aggression. At the end of the treatment period there were no significant differences between the two groups in terms of cognition or activities of daily living.

Strengths and limitations

This was a rigorous cluster-randomized controlled trial. A quarter of the patients received opioids, yet only three were excluded due to adverse events, making it likely that the interventions' effect on behavioral disturbances resulted from pain control as opposed to sedation. However, the population was outside of North America, so there may be barriers to implementing such a trial in the United States, especially given cultural concerns about using opioids in patients with dementia. The study also does not detail the types of staff education that were necessary to accomplish the intervention.

Clinical bottom line

A systematic approach to the management of pain significantly reduced agitation in nursing home residents with moderate to severe dementia.

Summary and main findings

Though increasing evidence documents its positive effects, important questions remain about the best model of inpatient palliative care. This cross-sectional survey study compared family ratings of care for patients who died in Veterans Affairs (VA) hospitals and nursing homes after either receiving no palliative care, a palliative care consultation, or care in a dedicated palliative care unit. The study included all patients who died between July 2008 and December 2009 in the 77 VA medical centers that had both a palliative care consultation service and a dedicated palliative care unit. Patients who died within 24 hours of admission were excluded. One month after patient death, the researchers telephoned English- and Spanish-speaking next of kin to administer the short version of the Family Assessment of Treatment at End-of-Life. This validated 10-item measure assesses overall care provided in the last month of life as well as specific aspects of patient and family care, such as pain control and provision of information and emotional support. ¹² The authors conducted two comparisons among families reporting the highest possible rating for each item (e.g., excellent): (1) comparison of patients who received usual care with those who received a palliative care consult and (2) comparison of patients who received a palliative care consult with those cared for in a dedicated palliative care unit.

During the study period, 10,633 patients died at the eligible medical centers; 55% of their families completed the survey. Median patient age was 70 years, 98% were male, and 70% were white. Life-limiting diagnoses included cancer, chronic lung disease, dementia, and heart and renal disease. Families of patients who died after receiving a palliative care consult were more likely to rate the overall care provided in the last month of life as “excellent” compared to families of patients who received usual care, with adjusted proportions of 51% versus 46%, p=0.04. Furthermore, families of patients who died after receiving care in a dedicated palliative care unit were more likely to rate overall care provided in the last month of life as “excellent” compared to families of patients who received a palliative care consult with adjusted proportions of 63% versus 53%, p<0.001. Similar results were found for the survey items assessing the specific aspects of care.

Strengths and limitations

This large study adds to the mounting evidence that palliative care improves family ratings of care provided near the end of life. Further, it is the first to investigate whether the outcomes of palliative care provided as a consult differ from those of palliative care in a dedicated inpatient unit. Though only 55% of patients' families were surveyed, the authors adjusted statistically for bias that this may have introduced. The authors also used statistical methods to adjust for the likelihood of receiving a consult or care in a palliative care unit. Typical of the VA medical centers where the study was conducted, the study population was predominantly male and white. The large number of hospitals (n=77) included in the study increases the generalizability of the findings.

Clinical bottom line

Palliative care consults were associated with higher ratings of care by bereaved family members of patients who died in VA hospitals. Receiving care in a dedicated palliative care unit was associated with even higher ratings.

Morrison RS, Dietrich J, Ladwig S, Quill T, Sacco J, Tangeman J, Meier DE: Palliative care consultation teams cut hospital costs for Medicaid beneficiaries. Health Aff (Millwood) 2011;30:454–463.

Summary and main findings

Medicaid spending is experiencing unprecedented levels of growth, threatening enrollees and the stability of our national budget. Palliative care may be an effective way of improving the quality of care for seriously ill Medicaid patients while decreasing costs. This retrospective case control study, conducted between 2004 and 2007 in four New York State hospitals, compared costs for seriously ill patients who received or did not receive palliative care consults. Eligible patients were admitted to one of the participating hospitals, had Medicaid as their primary and sole insurer, and had a range of advanced illnesses, including cancer and noncancer diagnoses. Based on administrative and billing records, the researchers classified patients as receiving a palliative care consult or not. Patients who received a palliative care consult were matched to patients who did not receive a consult using propensity scoring, which adjusts for the likelihood of a given patient receiving a consult. Daily and total hospital costs, as well as care received, were compared between the groups. Separate analyses were conducted for patients who were discharged alive and for those who died in the hospital.

Of 2212 eligible patients identified during the study period, 1717 were discharged alive; 296 of them received a palliative care consult. Through propensity scoring, 98% were matched to 1427 usual care patients who were discharged alive. Of 495 patients who died in the hospital, 189 received a palliative care consult; 98% of these patients were matched to usual care patients who died in the hospital. For patients who were discharged alive, palliative care consultation was associated with a net savings of $4098 over the course of the hospitalization compared to usual care. The savings was $7563 for patients who died during the admission. Average daily savings associated with palliative care was $316 beginning on the day of the consult. Patients who died in the hospital and received palliative care were less likely to die in the ICU, compared to usual care patients, 34% versus 58%, p=0.04. Patients discharged alive were more likely to receive hospice, 30% versus 1%, p<0.001.

Strengths and limitations

This study adds to the growing body of evidence that palliative care consult services decrease costs and improve quality, and the study highlights the important effect this can have at the national level. While this study's observational design is a limitation, the propensity score matching guards against bias, and the temporal relationship between the consults and savings strengthens the argument that the association is causal. Though patient and family outcomes were not measured directly, there is ample evidence that palliative care improves these. The New York location of all the hospitals and the mature status of their palliative care consult services limits generalizability. The services all received training through the Center to Advance Palliative Care ¹³ and thus were more likely to have followed National Quality Forum preferred practices. ¹⁴ The authors estimate that with this training, a hospital starting a new palliative care service could realize these savings within five years.

Clinical bottom line

Palliative care is associated with decreased hospitalization costs for seriously ill Medicaid patients. Consistent with the goals of a majority of patients and their families, palliative care recipients spent less time in the ICU and were more likely to be enrolled in hospice.