Kavalieratos, D., Corbelli, J., Zhang, D., Dionne-Odom, J.N., Ernecoff, N.C., Hanmer, J., . . . Schenker, Y. (2016). Association between palliative care and patient and caregiver outcomes: A systematic review and meta-analysis. JAMA, 316, 2104–2114. 

DOI Link

Purpose

STUDY PURPOSE: To complete a systematic review of palliative care interventions in randomized, controlled trials (RCTs) involving adults with life-limiting illness and meta-analysis to identify the relationship of those intervention with quality of life, symptom burden, and survival of those adults and their caregivers

TYPE OF STUDY: Systematic review of palliative care RCTs

Search Strategy

DATABASES USED: MEDLINE, EMBASE, CINAHL, and CENTRAL (inception to 2016)
 
ARTICLE TOPIC INCLUSION CRITERIA: Adults aged 18 years or older with a life-threatening illness diagnosis; patient report of a minimum of one of nine patient-level outcomes: mood, advance care planning, place of death, survival, use of resources, healthcare expenditures, care satisfaction, symptom burden, or quality of life (QOL). Interventions included at least two of eight possible domains of palliative care defined by the National Consensus Project for Quality Palliative Care. The systematic review included RCTs that featured usual care, waitlist, or attention control comparisons.
 
ARTICLE TOPIC EXCLUSION CRITERIA: Pediatrics, non–English-language, single symptom interventions (e.g., opioids for dyspnea), focus on only one palliative care domain (e.g., advance care planning only), non-focus on patients with life-threatening illness, caregiver-only intervention
 
Most of meta-analysis trials (72%) occurred in the United States and in ambulatory (32.5%) or home settings (41.8%). Forty-two trials reported subjective outcomes (those reported by patients), but over half were deemed high risk for bias or unclear bias (total of 81.4%). Of the 43 trials, 32 reported objective (survival) outcomes, but over 67% were deemed as high risk or unclear risk of bias. Trial interventions focused on a median of five interventions related to eight palliative care dimensions. Forty-two trials delineated physical dimensions of care, whereas 39 focused on psychological care, although no trial explicitly addressed cultural influences on interventions. Of 15 trial interventions evaluating caregiver outcomes, only five collected caregiver data only without an explicit intervention for caregivers. A variety of research designs existed in studies used for the meta-analysis. 
 
 
 

Literature Evaluated

TOTAL REFERENCES RETRIEVED: 6,158
 
EVALUATION METHOD AND COMMENTS ON LITERATURE USED: Conduct of the systematic review and meta-analysis occurred in concordance with the Cochrane Handbook for Systematic Reviews of Interventions. Articles met inclusion criteria from four database searches (inception to 2016). To evaluate the quality of studies in the systematic review, two reviewers independently evaluated each study for meeting inclusion criteria. If two reviewers’ evaluations varied, two different authors collaborated with the initial two reviewers to reach consensus. In addition, two of the four authors reviewed each trial according to a systematic and customized form to elicit information from study primary and secondary reports. Two investigators independently rated study risk of bias using the Cochrane Collaboration tool. A narrative synthesis included all trials (N = 43, 30 of patients with cancer) and considered quality of life, survival outcomes, and symptom burden premeta-analysis. Various statistical techniques controlled for trial heterogeneity, variation in study endpoint times, influence of study risk, intervention strength, study setting, and patient disease that may have influenced meta-analysis results and outcome measurement of patient symptom burden, survival, and quality of life, and their influence on caregivers.

Sample Characteristics

  • FINAL NUMBER STUDIES INCLUDED = 56 (30 with cancer)
  • TOTAL PATIENTS INCLUDED IN REVIEW = 12,731 patients, 2,479 caregivers 
  • SAMPLE RANGE ACROSS STUDIES: The mean patient age was 67 years; no caregiver age data were available. Sample sizes ranged from 27–434 participants.
  • KEY SAMPLE CHARACTERISTICS: All patients had life-limiting illnesses. Patients were in hospital, home, and ambulatory settings. Subjective outcomes: Half of reviewed trials had a high level of bias (55.8%), and 16% had a low level of bias; the rest was unknown. Objective outcomes: 44% had a high level of bias, and 6.9% had a low level of bias; the rest was unknown or not measured. Study interventions included physical and psychological aspects of care, with cultural sensitivity underused and underrepresented. Interventions addressed a median of five of eight palliative care components.

Phase of Care and Clinical Applications

PHASE OF CARE: Active cancer care
 
APPLICATIONS: Palliative care

Results

The reviewers looked at nine domains: patient quality of life, physical symptoms, survival, patient mood, advanced care planning, site of death, resource utilization and expenditures, satisfaction with care, and caregiver outcomes resulting from patient symptom burden, survival, and quality of life. Patient quality of life was assessed in 24 studies (4,576 patients). Twelve of those studies had high risk of bias and seven were low risk of bias (five were unknown). Of the seven low-risk-of-bias studies, five reported improved quality of life. In fifteen trials, quality of care was associated with a statistically significant improvement in quality of life. High bias and heterogeneity were significant issues in the analysis. Physical symptoms were reviewed in 29 trials (10,105 people). Seventeen of 29 trials looked at physical symptoms. Of the seven that were low risk bias, after sensitivity analyses, palliative care was not associated with change in symptom burden in four trials at the 1- to 3-month follow-up because of heterogeneity. The reviewers went on to describe that, because of high risk of bias and heterogeneity, no association existed between palliative care and improved survival, patient mood, advanced care planning, site of death, and resource use.

Conclusions

In this review, the evidence suggests that palliative care intervention improves symptom burden and patient quality of life in those who have been diagnosed with an advanced cancer or with a serious illness. The review was not able to establish if palliative care improved caregiver quality of life. Significant issues existed with assessing the association of palliative care with quality of life, symptom burden, and adult survival because of the problematic quality and rigor of RCTs used in the systematic review and meta-analysis.

Limitations

  • Mostly low quality/high risk of bias studies
  • High heterogeneity
  • The authors reported multiple limitations of this review, including the wide spectrum and diversity in palliative care interventions. This diversity contributed to heterogeneity to the meta-analysis. Missing data, the exclusion of quasiexperimental studies, and trial statistical analysis may have contributed to underpowered studies in the meta-analysis. Across trials, clinical trial data were not uniformly reported, and information about points in time when patient and caregiver may have received palliative care was lacking. The risk of study bias (a subjective interpretation) and use of the Cochrane Risk of Bias tool may not have defined specifics of behavioral interventions found in many of the trial studies used in the meta-analysis and systematic review.
  • No specific information provided on evaluation rating forms
  • No caregiver-only studies included in meta-analysis to measure caregiver outcomes
  • No trial focused on influence of culture on intervention assessment or effect.

Nursing Implications

Multiple RCTs have established that palliative care improves patient experience and quality of care. Insufficient evidence exists to assess whether that is true for patient and caregiver dyads or for caregivers assessed separately from patients. More studies reflecting methodological rigor, cultural sensitivity, and quality to identify aspects of effective palliative care for both patients and caregivers remain a priority.

Legacy ID

6438