Improving Care Experiences for Patients and Caregivers at End of Life: A Systematic Review

Abstract

Background.

End-of-life care is increasing as the US population ages. Approaches to providing high-quality end-of-life care varies across setting, diseases and populations. Several data collection tools measure patient and/or caregiver care experiences at end-of-life and can be used for quality improvement. Previous reviews examined palliative care improvements or available measures, but none explicitly on improving care experiences. We reviewed literature on improving patient and/or caregiver end-of-life care experiences.

Design.

We searched US English-language peer-reviewed and grey literature after 2000 on adult end-of-life care experiences. We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses for quantitative studies, Enhancing Transparency in Reporting the Synthesis of Qualitative Research for qualitative studies, and Assessment of Multiple Systematic Reviews for literature reviews.

Setting.

Palliative and hospice care.

Population.

Full-text abstraction of 84 articles, identifying 16 articles.

Measures.

Patient and/or caregiver end-of-life care experiences (captured through administrative data or direct report).

Results.

Articles examined palliative care experiences across settings; none studied hospice care experiences. Patients and/or caregivers assessed overall care experiences, clinician-staff interactions, provider communication, respect and trust, timeliness of care, spiritual support, caregiver knowledge of care plans, or bereavement support. Efforts aimed at improving end-of-life care experiences are limited and show mixed results.

Conclusions.

Literature on improving patient and/or caregiver end-of-life care experiences is emerging and focused on palliative care experiences. Evidence on improving hospice care experiences is lacking. Research on strategies for improving end-of-life care experiences should go beyond overall care experiences to include specific aspects of palliative and hospice care experiences.

Introduction

As the U.S. population ages, the use of palliative care (PC) and hospice care is rising.¹ Improving the quality of these end-of-life care experiences for patients and caregivers is important given the growing evidence that providing high-quality patient-centered aspects of PC or hospice care, such as pain or communication, for patients with advanced or serious illness vary widely² and the resulting care experiences also vary across setting, diseases, and populations.^3–5

Several data collection tools are designed to measure patient and/or caregiver end-of-life care experiences, including both palliative or hospice care experiences, and be used for quality improvement. These include: the Veterans Administration’s (VA) Bereaved Family Survey (BFS) that assesses families’ perceptions of the quality of care received by veterans in the last month of life;⁶ Family Evaluation of Palliative Care (FEPC), a post-death survey administered to bereaved family caregivers that assess families’ perceptions of the quality of PC received;⁷ the Toolkit of Instruments to Measure End-of-life care (TIME), that includes an after-death interview administered to the bereaved family member of adult persons who died of a chronic progressive illness and received services from a home health agency, nursing home, hospice or acute care hospital,⁸ and the CAHPS Hospice survey, completed by the primary caregiver of a hospice decedent, that asks about the decedent’s care experiences and the caregiver’s own experiences.⁹

Although the collection and measurement of patient and caregiver end-of-life care experiences is increasing, prior systematic reviews of end-of-life care have revealed a lack of research about efforts to improve the quality of patient and caregiver/family care experiences. Alexander and Herald, in reviewing the effectiveness of quality improvement (QI), found only 4% of nursing home studies explicitly examined the relationship between formal QI efforts or changes and quality-related outcomes, no evidence for improving care experiences.¹⁰ Aslakson et al¹¹ identified a variety of strategies to improve PC in the intensive care unit (ICU), without addressing PC experiences explicitly, and noted the need for more research on improvement strategies for end-of-life care experiences across settings. Fawole et al¹² reviewed communication quality interventions for patients with advanced or serious illness and found few studies measured patient/family experiences and almost none that found statistically significant improvements. A 2012 review from the Agency for Healthcare Research and Quality (AHRQ)² examined evidence on the effectiveness of interventions to improve outcomes of patients with advanced and serious illness, including PC and hospice care, identifying 17 studies that included measures of patient satisfaction/experience (2 hospice) and 4 with statistically significant improvements in PC communication across settings.² Lendon et al¹³ literature review identified several publicly available surveys and measures of patient, family or informal caregiver experiences and satisfaction with care at the end-of-life covering a variety of content that could be used for quality improvement and research.

We seek in this review to examine evidence on how to improve the quality of patient and/or caregiver end-of-life care experiences, explicitly including hospice experiences.

Methods

We reviewed articles on improving patient and caregiver experiences of end-of-life care across settings, explicitly including hospice care experiences. We adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines for quantitative studies, Enhancing Transparency in Reporting the Synthesis of Qualitative Research approach for qualitative studies, and Assessment of Multiple Systematic Reviews tool for literature reviews.^14–16

Search Strategy.

We used a structured search strategy on PubMed and PsycINFO to identify peer-reviewed studies, limiting our search to US English-language peer-reviewed articles published from January 2000 to March 2019 on adult end-of-life care. We identified articles using Medical subject headings (MeSH) and keywords with at least (1) one improving quality keyword, (2) one experience keyword, and (3) one end-of-life care keyword, or (4) a named end-of-life care experience survey (e.g., Bereaved Family Survey). Table 1 lists the search terms by category. We also searched 17 palliative or hospice organizational websites to identify grey literature. Table 2 lists the organizations and search terms used for the grey literature. We identified 1,317 articles.

Table 1.

Search strategy

Concept	Medical Subject Headings (MeSH)	Search Terms
Improving quality	Quality improvement	Quality improvement; performance improvement; process improvement; plan-do-study-act; six sigma; learning collaborative; best practices;
Patient and/or caregiver experience	Patient-centered care; patient satisfaction	Patient experience; patient centered care; patient satisfaction; bereaved family; bereaved caregiver
End-of-life care	Hospices; hospice care	Hospices; hospice care; nursing home; assisted living facilities; palliative care; end of life care; end of life experience survey

NOTE:

The search syntax was:

((((((English[Language])

AND

(“2000”[Date - Publication]: “3000”[Date - Publication])))

AND

(“quality improvement”[tiab] OR “performance improvement”[tiab] OR “process improvement”[tiab] OR “plan do study act”[tiab] OR “six sigma”[tiab] OR “learning collaborative”[tiab] OR benchmark*[tiab] OR compare[tiab] OR trend*[tiab] OR “information sharing”[tiab] OR transparency[tiab] OR “improving quality”[tiab] OR “improving care”[tiab] OR “quality measurement”[tiab] OR “quality improvement”[tiab] OR lean[tiab] OR “care improvement”[tiab] OR “improving performance”[tiab]

AND

“patient experience”[tiab] OR “patient experiences”[tiab] OR “patients’ experience”[tiab] OR “patient centered care”[tiab] OR “patient satisfaction”[tiab] OR (“bereaved family”[tiab] AND experience[tiab]) OR bereaved caregiver*[tiab] OR nurse practitioner*[tiab] OR physician*[tiab] OR clinical nurse*[tiab] OR clinician*[tiab] OR doctor*[tiab] OR “home health”[tiab] OR “chaplain” OR “social worker”[tiab] OR “navigator”[tiab]

AND

hospices[mesh] OR hospice care[mesh] OR nursing home*[tiab] OR assisted living facilit*[tiab] OR “hospice”[tiab] OR palliative[tiab] OR “end of life”[tiab] OR “family assessment of treatment at end of life”[tiab] OR FATE[tiab] OR “bereaved family survey”[tiab] OR “BFS”[tiab] OR ((“veterans administration”[tiab]) AND (hospices[mesh] OR hospice care[mesh] OR nursing home*[tiab] OR assisted living facilit*[tiab] OR “hospice”[tiab] OR palliative[tiab] OR “end of life”[tiab]))

OR

((“family evaluation”[tiab] AND “hospice care”[tiab]) OR FEHC[tiab]) OR (“CMS Hospice Item Set”[tiab] OR “hospice item set”[tiab] OR “CMS(HIS)”[tiab] OR “family assessment of treatment at end of life”[tiab] OR FATE[tiab] OR “bereaved family survey”[tiab] OR “BFS”[tiab])

Table 2:

Grey literature search

Organization	Search Terms
Institute for Healthcare Improvement	improving hospice; hospice quality; hospice improvement; hospice measure; palliative measure; palliative improvement; palliative quality; improving palliative; “learning collaborative” hospice; “information sharing” hospice; “information sharing” palliative care; performance hospice
National Hospice and Palliative Care Organization	“quality improvement” OR “quality measurement” OR “performance improvement” OR “process improvement” OR “plan do study act” OR “six sigma” OR “learning collaborative” OR benchmark* OR “information sharing” OR transparency OR “improving quality” OR “improving care” OR lean OR “care improvement” OR “improving performance”
Centers for Medicare and Medicaid Services	Advanced Google: allintitle: hospice “measure” site:cms.gov; allintitle: palliative “measure” site:cms.gov
American Academy of Home Care Medicine	hospice, palliative care
American Academy of Hospice and Palliative Medicine	improving quality; measure hospice; improving hospice; “palliative measure” “six sigma” “information sharing” “white paper” “quality measure” “improving performance”
American Health Care Association	“hospice measure” “improving hospice” “hospice quality” “palliative measure” “Palliative improvement” “six sigma” “lean” “plan do study act” “improving performance”
American Medical Group Association	“hospice measure” “improving hospice” “hospice quality” “palliative measure” “Palliative improvement” “hospice” “palliative care”
Association of Professional Chaplains	“hospice measure” “improving hospice” “hospice quality” “palliative measure” “Palliative improvement” “hospice” “palliative care”
Center to Advance Palliative Care	“hospice measure” “improving hospice” “hospice quality” “palliative measure” “Palliative improvement” “six sigma” “lean” “plan do study act” “improving performance”
Healthcare Chaplaincy Network	palliative care measure; hospice measure; improving hospice; improving palliative care
Long-Term Quality Alliance	hospice; palliative
National Association for Home Care and Hospice	“quality improvement” OR “quality measurement” OR “performance improvement” OR “process improvement” OR “plan do study act” OR “six sigma” OR “learning collaborative” OR benchmark* OR “information sharing” OR transparency OR “improving quality” OR “improving care” OR lean OR “care improvement” OR “improving performance”
National Palliative Care Research Center	“quality improvement” OR “quality measurement” OR “performance improvement” OR “process improvement” OR “plan do study act” OR “six sigma” OR “learning collaborative” OR benchmark* OR “information sharing” OR transparency OR “improving quality” OR “improving care” OR lean OR “care improvement” OR “improving performance”
National Quality Forum	hospice; palliative care
Physicians Assistants in Hospice and Palliative Care Medicine	“quality improvement” OR “quality measurement” OR “performance improvement” OR “process improvement” OR “plan do study act” OR “six sigma” OR “learning collaborative” OR benchmark* OR “information sharing” OR transparency OR “improving quality” OR “improving care” OR lean OR “care improvement” OR “improving performance”
Social Work Hospice and Palliative Care Network	“quality improvement” OR “quality measurement” OR “performance improvement” OR “process improvement” OR “plan do study act” OR “six sigma” OR “learning collaborative” OR benchmark* OR “information sharing” OR transparency OR “improving quality” OR “improving care” OR lean OR “care improvement” OR “improving performance”
Veterans Affairs	improving quality; measure hospice; improving hospice; “palliative measure”; “hospice performance”; “hospice measure”; “bereaved family survey”

Article Screening.

We reviewed titles and abstracts of identified articles. After an initial period of double-coding to establish consistency across reviewers (DQ, SM), remaining abstracts were independently reviewed to determine eligibility. Reviewers discussed discrepancies during meetings and resolved disagreement to reach consensus on inclusion. All abstracts marked for inclusion were then reviewed by both authors.

As shown in Figure 1, articles were excluded if they: did not study efforts to improve end-of-life care experiences (N=443); did not contain measures of patient and/or caregiver experiences (N=234); were not about end-of-life care (N=174); were not empirical studies (e.g., commentaries; N=337 articles); or involved only pediatric populations (N=45 articles).

Figure 1: PRISMA Flow Diagram.

Abstraction.

We undertook a full review of 84 articles. Two researchers reviewed the full article and abstracted specific information into a form: country, study aims, care setting, care type, design, timeframe, methods, statistical approach, control variables, sample size, identification of experience measures, other outcomes, improvement focus (e.g., overall care experience), and results concerning improvement activities and experience/outcome measures. Care settings included hospital, intensive care unit, hospital, inpatient hospice unit, nursing home, assisted living facility, VA, home health and other. Experience/outcome measures included patient/caregiver reported, patient/caregiver experience captured through administrative data, clinical measures and other. To ensure both reviewers employed a similar approach, a small number of articles were selected for double-review and discussion. Once agreement was reached, each reviewer abstracted half the remaining articles. After abstraction, each article was reviewed by the other reviewer to ensure accuracy of abstracted content and discussed if needed to gain consensus achieving 100% interrater agreement.

We determined that the available tools for rating study quality were too time and labor consuming to employ given we were not conducting a meta-analysis of effect sizes, so we developed a general approach to rate study quality. Both reviewers independently performed a global rating of study quality (poor, adequate or good), using the following criteria: (1) Sample size (e.g., <20=automatic “poor” rating); (2) Study design (e.g., a retrospective study would receive an “adequate” rating if all other rating categories were “good”); and (3) Statistical approach (e.g., appropriate test for study design, tested and reported statistical significance of finding, controlled for appropriate variables; if one of these categories is not met, maximum rating =”adequate”). Studies with sample size less than 20 or weaknesses in two or more categories automatically received a rating of “poor.” We discussed our ratings to gain consensus. Thirteen articles were excluded as “poor”. Half (n=8) were rated as “adequate” and the other half as “good”.

During abstraction, we excluded 25 qualitative studies about patient/caregiver care experiences that did not examine efforts to improve such care experiences; 15 non-empirical studies; 13 non-US studies; and two without patient or caregiver experience measures. Altogether, 16 studies were included in the review^17–32 (Table 3).

Table 3.

Included Articles

First Author and Year	Setting	Design and Sample Size *rated “good”	Type of Care	Types of Included Measures **improved ∼mixed +no improvement
Allen, SR 2017	ICU	Pre-post Pre: 412 family members, 49 Nurses, 5 Physicians; Post: 427 family members, 47 Nurses, 6 Physicians	Palliative	Patient/Caregiver-reported experience**; Patient/Caregiver experience (via administrative data); Clinical
Amro, OW 2016	Other (dialysis center)	Longitudinal with comparison group 50 patients in intervention group, 151 comparison group, 9 Nephrologists 8 Physicians in 2 outpatient dialysis facilities	Palliative	Patient/Caregiver experience (via administrative data)**
Armstrong, B 2013	Hospital	Longitudinal 25 patients in a rural Appalachian community hospital	Palliative	Patient/Caregiver-reported experience; Clinical (including symptom management)**
Bookbinder, M 2005	Hospital	Pre-post with comparison group* 101 patients who died across 5 units compared to156 patients who died 9 months prior across same units: 3 hospital units (Geriatrics, Oncology, inpatient palliative care unit) compared to 2 general medical units	Palliative	Patient/Caregiver experience (via administrative data); Clinical (including symptom management)**
Brumley, RD 2003	Home health	Longitudinal with comparison group* 162 patients enrolled in Palliative Care Program who died compared to 139 in the comparison group	Palliative	Patient/Caregiver-reported experience**; Clinical
Brumley, R 2007	Home health	Random control trial* 298 homebound, terminally ill patients with a prognosis of approximately 1 year or less to live plus one or more hospital or emergency department visits in the previous 12 months in 2 health maintenance organizations in 2 states	Palliative	Patient/Caregiver-reported experience**; Clinical (including symptom management)
Campion, FX 2011	Other (cancer center)	Longitudinal* 273 oncology practices	Palliative	Patient/Caregiver experience (via administrative data); Clinical (including symptom management)**
Curtis, JR 2008	ICU	Pre-post* Pre: 253 patients who died in the ICU Post: 337 patients who died in the ICU Of which, 496 had identified family member and 275 family members responded, 523 Nurses	Palliative	Patient/Caregiver-reported experience+; Clinical
Curtis, JR 2011	ICU	Cluster Randomized Trial* Sample A – Pre: 239 patients and their families Post: 182 patients with their families in 6 intervention hospitals; Pre:187 patients and their families Post: 214 patients and their families in 6 control hospitals Sample B – Pre: 285 patients and their nurses Post: 167 patients with their nurses in 6 intervention hospitals; Pre:113 patients and their nurses Post: 213 patients and their nurses in 6 control hospitals	Palliative	Patient/Caregiver-reported experience+; Patient/Caregiver experience (via administrative data); Clinical
Deptola, AZ 2019	Hospital	Pre-post Pre: 151 patients Post: 132 patients	Palliative	Patient/Caregiver experience (via administrative data)**; Clinical
Hoverman, JR 2018	Other (cancer center)	Longitudinal with comparison group* 509 patients with 9 common cancer diagnosis in study group with a matched cohort of 900 patients	Palliative	Patient/Caregiver-reported experience+; Clinical
Jacobs, LG 2002	Hospital	Pre-post Pre: 194 hospitalized patients in critical care or geriatrics, 133 family, Post: 46 Physicians	Palliative	Patient/Caregiver-reported experience+; Clinical (including symptom management)**
Kaufer, M 2008	ICU	Pre-post Pre: 43 family members of patients who died in medical ICU (MICU) Post: 45 family members of patients who died in MICU	Palliative	Patient/Caregiver-reported experience∼
Keay, TJ 2003	Nursing home	Longitudinal with control group 203 dying patients, 12 physicians in 4 intervention nursing homes before and after educational program, 18 physicians in 1 control nursing home	Palliative	Patient/Caregiver experience (via administrative data)**; Clinical (including symptom management)**
McCormick, AJ 2010	ICU	Longitudinal 35 social workers, 283 patients, 275 families completed surveys one month after patient death	Palliative	Patient/Caregiver-reported experience∼
Scaife, R 2017	Home health	Retrospective review with control group* 154 intervention patients, 146 patients with usual care	Palliative	Patient/Caregiver experience (via administrative data)**; Clinical

Results

Of the 16 studies, five were in an intensive care unit,^{17,23,24,28,30} four in the hospital,^19,20,25,27 three in home health,^21,22,31 one in nursing homes,²⁹ and three in “other” settings (two in cancer centers and one in a dialysis center).^18,26,32 All studies were in a single setting. Seven were longitudinal,^{18,19,22,26,29,30,32}, six used a pre-post design,^{17,20,24,25,27,28} two were random control trials^21,23 and one a retrospective review.³¹ Half (n=8) included a comparison group.^{18,20–23,26,29,31} All studies focused on adults and examined PC experiences. None, despite our efforts to include it, studied efforts to improve hospice experiences.

Six articles included patient-and/or caregiver-reported experience measures and clinical measures.^{19,21,22,24,26,27} Five targeted patient and/or caregiver experiences captured through administrative data (rather than direct report) and clinical measures.^{20,25,29,31,32} Two included all three types of measures: clinical, patient/caregiver-reported experiences, and patient/caregiver experiences captured through administrative data.^17,23 Two targeted only patient- and/or caregiver-reported experience,^28,30 and one study included only experiences captured through administrative data.¹⁸

The most common measures of patient and caregiver-reported experience were global measures of overall care,^{21–24,26,28} followed by interactions with clinicians and staff,^{21–23,28,30} and provider communication.^24,27,28 Few articles targeted respect and trust,¹⁹ timeliness of care,¹⁹ or another aspect of patient experience such as symptom burden,^23,24,31 spiritual concerns,¹⁹ or caregiver knowledge of providers and care plans.¹⁷

Patient and/or caregiver experiences captured through administrative data were most commonly documentation of goals-of-care conversations,^25,31,32 followed by chart-based elements of PC experiences,^20,23,29 information reported through Physician Orders for Life-Sustaining Treatment (POLST) forms,¹⁸ and documentation of bereavement support.²⁹

The most common clinical measure alongside experience measures was symptom management.^{19–21,27,29,32} Other clinical measures included hospital/ICU length of stay,^17,23,24 hospice utilization,^26,29,32 discharge location,³¹ number of process modifications,²⁰ and change in code status.²⁵

Patient- and/or Caregiver Care Experiences

Reported Experiences.

Ten studies focused on efforts to improve patient or caregiver-reported experiences. Three of these studies showed improvement in patient and/or caregiver reported experiences,^17,21,22 two had mixed results,^28,30 four studies showed no improvement in patient or caregiver reported experiences,^23,24,26,27 and one study included only group-level cross-sectional post-improvement measurements of patient-and/or caregiver-reported experiences (without baseline measurement) so no determination of overtime improvement was possible.¹⁹

Several studies found statistically significant improvement (p<0.05 level) in patient and/or caregiver reported experiences. Three of the six studies examining overall ratings of care found a statistically significant increase.^22,28,30 One study measuring caregiver-reported knowledge of physician and patient care plan showed statistically significant improvement.¹⁷ Another study tracking several measures of patient-and/or caregiver-reported experience found improvement in provider communication.²⁸

Experiences Documented Through Administrative Data.

Seven studies focused on efforts to improve patient and/or caregiver experiences documented through administrative data. Five of these studies – one study improving bereavement support (documented in administrative data)²⁹ and four studies on documented goals-of-care conversations^17,18,25,31 – showed statistically significant improvement.

Strategies that Improved Care Experiences

Reported Experiences.

Several strategies had statistically significant improvements in patient- and/or caregiver-reported experiences. Strategies improving global measures of PC experience included:

• interdisciplinary care team, with expertise in symptom management and biopsychosocial intervention responsible for coordinating and managing care across all settings, improved overall patient and caregiver-reported experiences of PC.²²

• interdisciplinary clinician education in PC, trained ICU PC champions, identifying ICU-specific barriers to PC, providing feedback from overall ICU-specific family/caregiver experience data, and employing system supports such as standardized physician order forms, that improved global measures of caregiver-reported experiences.³⁰

Additionally, a strategy that improved caregiver knowledge was a QI effort inviting family members/caregivers to participate on daily rounds, which improved caregiver knowledge of the patient’s physician and care plan.¹⁷ Another study identified a strategy for improving overall PC experience, provider communication and competence, accessibility of information, involvement in decision making, and caregiver-reported satisfaction with the death and dying process. This involved facilitating interaction between medical intensive care unit clinicians, PC teams, and family support staff through regular patient care meetings; staff education about PC; family meetings with all-involved team members; chart documentation about goals-of-care; and ongoing patient follow-up by an advanced practice nurse, with direct and intensive involvement with the patient’s family.²⁸

Experiences Documented Through Administrative Data.

Four studies found a statistically significant increase in documented goals-of-care conversations following improvement activities.^17,18,25,31 Strategies that improved documented goals-of-care conversations included:

• having patients with high risk for mortality participate in a dedicated clinical encounter with their primary provider focusing on advance care planning.¹⁸

• email alert to inpatient attending physicians explaining mortality risk of patients and recommending a PC consultation or goals-of-care conversation.²⁵

• Identifying a family member/caregiver as a point of contact for the critical care team and invited to participate on daily rounds.¹⁷

• Emphasis on medication management, care coordination, patient and caregiver education, advance care planning, spiritual care, and maintenance of patient health records.³¹

Additionally, implementing a half-day of education for nursing home physicians on PC practices improved documented bereavement support.²⁹

Strategies that Aimed to Improve Care Experiences, but Did Not

Strategies aimed to improve patient- and/or caregiver-reported end-of-life care experiences, but did not succeed included:

• clinician education, local champions, academic detailing with feedback to clinicians.^23,24

• patients undergoing chemotherapy received periodic phone calls from a trained nurse to discuss general questions, systematically assess symptoms, refer patients for additional care, and provide education regarding advance care planning.²⁶

• Using social worker with PC expertise to discuss patient prognosis, care preferences, and be present when discussing the discharge plan with patient, caregiver and physician.²⁷

Authors of these studies indicated several possible reasons for the failure to identify improvements in end-of-life care experiences. Two studies indicated that improving caregiver ratings may require more direct contact with caregivers and family; Two studies cited measurement and sample size reasons for not detecting improvements (i.e., lack of variation in measures over time, skewed score distribution, small sample size).

Clinical Outcomes Studied Alongside Care Experiences

Thirteen articles included clinical outcomes alongside experience measures.^{17,19–27,29,31,32} Six studies measured symptom management.^{19–21,27,29,32} Because symptom management aims to prevent or treat the symptoms of a disease, side effects caused by treatment, and psychological, social, and spiritual issues related to a disease or its treatment, it can be considered either a clinical outcome or an aspect of patient experience. It is included in comfort care, PC, and supportive care. Because of this ambiguity as either a clinical or experience measure, we include and discuss below symptom management articles.

Five studies found a statistically significant improvement in symptom management.^{19,20,27,29,32} Successful strategies that significantly improved symptom management included:

• PC consultation with assessment and recommendations to the attending physician, direct regular involvement by the PC team with the patient and family to facilitate decision making regarding goals-of-care and treatment limitations, and ongoing communication with other members of the PC team to determine efficacy of treatments and accomplishment of care goals.¹⁹

• PC clinical pathway consisting of an interdisciplinary care path, nurses’ daily documentation flowsheet, and a standardized physician’s order sheet with guidelines for medical management of 12 symptoms prevalent at end-of-life.²⁰

• social worker with PC expertise to discuss patient prognosis, care preferences, and be present when discussing the discharge plan with patient, caregiver and physician.²⁷

• half-day educational program for nursing home physicians on PC practices.²⁹

• audit and feedback with a twice-yearly data reporting and analysis cycle and offering participating practices extensive, site-specific, and benchmarking reports regarding areas of care that are important to quality and value.³² Sites reporting in multiple periods had significantly better performance on multiple quality measures than sites reporting in one period. This included better performance on six of seven clinical measures (i.e., assessments of pain and dyspnea) and four of seven administrative measures on hospice and PC discussions and referrals.

Discussion

This review systematically assessed documented efforts aimed at improving patient and/or caregiver end-of-life care experiences (i.e., PC or hospice care experiences) and found it focused only on PC experiences. We found that patient and/or caregiver end-of-life care experiences are measured slightly more through patient and/or caregiver report (10 articles) than through administrative data (8 articles), and most often alongside clinical measures (13 articles). For reported care experiences, global measures of care experiences were most common (6 studies), followed by interactions with clinicians and staff (5 studies) and provider communication (3 studies). Goals-of-care conversations (5 studies) was the most common care experience collected through administrative data. Symptom management (6 studies) was the most common clinical measure studied alongside end-of-life care experiences.

We found limited evidence regarding efforts aimed at improving global measures of overall care experiences at end-of-life, with half of the studies showing statistically significant improvement and half showing no improvement. These studies only addressed PC experiences, with none addressing improvements in hospice care experiences. We found limited evidence for strategies that improved aspects of patient and/or caregiver experiences at end-of-life, such as caregiver knowledge or bereavement support.

Across the 16 included studies, we did identify several successful strategies for improving PC experiences, however each was included in a single study; we identified across three studies the strategy of identifying and inviting family/caregivers to participate in inpatient daily rounds which successfully improved documented goals-of-care conversations.

With hospice care experiences, we found no evidence available on strategies hospices use to improve hospice care experiences. Our review, as did previous systematic reviews of end-of-life care, found this lack of evidence on how to improve hospice care experiences. This lack of evidence for improving hospice care experiences is especially notable since hospices are required by the Center for Medicare and Medicaid Services (CMS) to have a quality-assurance program with a QI component and since 2019, hospices have been using quality measurement and QI to comply with the Medicare Conditions of Participation.

The mixed types of experience measures (i.e., caregiver-reported care experiences or care experiences captured through administrative data) found may be a result of the limitations in caregivers assessments and reports about the intensity of symptoms and quality of life of patients with cancer and patients in hospice.^33,34 However, this literature was published prior to the development of several post-death surveys of bereaved family members about PC and hospice care experiences and before the 2015 CMS requirement that Medicare-certified hospices administer the CAHPS® Hospice Survey⁹ and complete for every patient at admission and discharge the Hospice Item Set.³⁵

Because surveys of patient and/or caregiver experience almost always ask patients or caregivers to provide an overall rating of care, it is not surprising that the emerging evidence included a global measure of patient and/or caregiver care experiences. This may also be a result of the 2015 Measure What Matters project³⁶ that recommends measuring 10 indicators applicable to palliative or hospice care, and includes using a global measure of care based on family or caregiver perspective. However, efforts aimed at improving end-of-life care experiences should not only include overall ratings of care but also include specific measures of the different aspects of care experiences, such as interactions with clinicians/staff, provider communication, emotional support, spiritual care, or symptom management. Specific care experience measures are known to be more actionable for QI.³⁷ Including measures on the specific aspects of patient and/or caregiver experiences is critical for identifying gaps and improving care experiences.

Most notable in our review is the continued absence of literature on improving patient and/or caregiver hospice care experiences. QI is important in hospices, as there is known variation in the quality of care experiences among hospice programs. In the CAHPS Hospice survey, hospice care experiences of Blacks and Hispanics were found to be similar to or better than those of Whites for most dimensions, though caregivers of Blacks and Hispanics were less likely to report receiving their desired amount of emotional and religious support.⁴ National CAHPS Hospice survey data found hospice team communication was the strongest predictor of the overall hospice rating in all care settings except the inpatient care unit (where treating family with respect was the most important predictor).⁵

A key-informant study of 52 palliative and hospice care providers, based on a nationwide sample, identified and described potential barriers and facilitators to QI implementation in hospices ³⁸. Authors identified important barriers: (1) external factors, such as lack of QI models specific to end-of-life care, constrain QI implementation; (2) internal factors, such as lack of resources, limit capacity for QI; and (3) research on best practices for hospice is limited. These findings support the need for further research and documentation of efforts to inform and guide QI endeavors in hospices specifically and for end-of-life care more broadly. Though challenging, efforts aimed at improving patient and/or caregiver end-of-life care experiences are needed to identify feasible recommendations that can help facilitate QI across end-of-life care settings including PC and hospice.

These identified gaps in the literature, the importance of quality end-of-life care experiences, and the current CMS requirements for quality measurement and improvement at hospices, highlight the need for research to identify strategies, best practices, and evidence aimed at improving patient and caregiver end-of-life care experiences (based on either caregiver report or obtained via administrative data) for both PC and hospice care.

Our work has limitations. First, our review intentionally focused on articles about improving end-of-life care experiences. As such, our search included both a quality or improvement term and an experience term; this was not a review of interventions, such as education programs, new or stronger policies, or providing different therapies at end-of-life. This resulted in a review of a small, but emerging focused literature. Second, our review necessitated a search string aimed at identifying care experiences, which is a complex and multi-dimensional construct, sometimes described as “outcomes.” For example, in some hospice literature, the distinction between an outcome and a care experience is not clear. In our review of references, we identified studies on hospice care (by the same author) that studied outcomes of symptom distress, caregiver burden or caregiver quality of life, but did not frame these constructs as care experiences of the patient or caregiver.^39–41 We excluded these studies as they were not examining efforts aimed at improving care experiences (rather were studies of interventions associated with quality of life). Both these challenges are inherent in the fact that the concepts themselves are complex and multi-dimensional. As a result, different studies may focus on different components of each construct or describe them in different ways.

In sum, the evidence for improving patient and/or caregiver end-of-life care experiences is limited, focused on PC, but encouraging. Research is needed to identify best practices and strategies for improving patient and/or caregiver end-of-life care experiences. This research should go beyond overall care experiences to include specific aspects of PC and hospice care experiences.