Palliative Care Intervention Trials for Adults Living with Progressive Central Nervous System Diseases and Their Caregivers: A Systematic Review

Abstract

Context:

Interest in implementing palliative care for adults living with progressive central nervous system diseases (PCNSD) and their caregivers is increasing. Objectives: To inform evidence-based practice and future research by critically evaluating randomized clinical trials (RCTs) investigating palliative care interventions (PCIs) for adults living with PCNSD and their caregivers using self-reported outcomes and the patient- and caregiver-reported outcome measures employed.

Methods.

A systematic search using PRISMA methods of EMBASE, PubMed, Scopus, Web of Science databases using index and keyword methods for articles published from inception through 28 February 2021 was performed. RCTs investigating PCI as their primary aim using patient- and/or caregiver-reported outcomes to assess PCI effectiveness in adults living with PCNSD and their caregivers were included for qualitative synthesis.

Results:

Five RCTs met criteria and used 21 unique outcome measures. Pooled patient diagnoses included multiple sclerosis, motor neuron disease and movement disorders, primarily Parkinson’s Disease. All 5 RCTs assessed PCI effectiveness on patient symptom burden and caregiver burden, and 3 RCTs used patient QOL as a primary outcome. Overall risk of bias was low. Pooled positive findings were limited to very modest changes in patient QOL, specific physical symptoms and caregiver burden. Most outcome measures lacked clinimetric responsiveness to detect change whether caused by disease or an intervention to the patient or caregiver.

Conclusion:

Sparse, low-certainty evidence for PCI impact on patient QOL, symptom burden and caregiver burden indicate future research should consider refining study populations, interventions, outcomes assessed and outcome measures to detect any change due to PCI.

Editorial Note:

David Hui, MD

This nicely conducted systematic review examining the role of palliative care in patients with primary central nervous system disease identified only 5 randomized clinical trials of variable methodologic quality, highlighting a major knowledge gap in this field and the need for more high quality research.

Introduction

There is growing interest in investigating the impact of palliative care interventions (PCI) ideally delivered by an interdisciplinary palliative care team (PCT) in adults diagnosed with progressive central nervous system diseases (PCNSD) and their caregivers, especially among those practicing in both neurology and palliative medicine who comprise the emerging field of neuro-palliative care.^1,2 Healthcare providers caring for people living with chronic, incurable and progressive central nervous system diseases and their caregivers are aware of their high burdens across multiple symptom domains.^3,4,5,6,7 As the concept of quality of life (QOL) has become more popular in clinical practice, the importance of addressing QOL issues and symptoms across all domains has become increasingly recognized by healthcare providers including those who specialize in neurology.²

There is a need for high-quality evidence for the provision of palliative care (PC) for the large and increasing number of adults living with PCNSD and their caregivers.^8,9 PCNSD such as cerebrovascular disease, neurodegenerative diseases, neuroinflammatory conditions, and malignancies involving the central nervous system and its coverings commonly occur and all are associated with neurologic and cognitive symptoms giving rise to functional limitations that often worsen over time. As examples of the high prevalence, more than 7 million stroke survivors are living in the United States alone, two-thirds of whom disabled, and over 50 million people worldwide are living with dementia.^10,11 Multiple sclerosis (MS) has an estimated worldwide prevalence of 50 to 300 per 100,000 and a global incidence of roughly 2.3 million people.¹² In the recently published Global Burden of Disease Study assessing the burden of neurologic disorders across the United States between 1990 – 2017, among the 5 most burdensome measured in terms of disability-adjusted life-years was stroke, dementias including Alzheimer’s type, and Parkinson’s Disease, and the 5 leading causes of deaths due to neurological diseases were dementias, stroke, Parkinson’s disease, motor neuron disease and MS.¹³

Reviews of PC in adults living with PCNSD, located by the search for this review, were published between 2005 and 2019: 1 Cochrane review aimed to assess multidisciplinary care involving neurology, rehabilitation medicine and palliative care providers in adults living with motor neuron disease¹⁴; 1 Cochrane review focused on assessing palliative care or multidisciplinary care closely aligned with the same aims as a PCT in adults living with MS¹⁵, and 1 Cochrane review¹⁶ and 1 systematic review¹⁷ each sought to assess PCI in people living with dementia. Sampson et al’s 2005 systematic review and Murphy et al’s 2016 Cochrane review shared the common objective of assessing the investigation of PC in adults living with dementia in any setting.^16,17 Though conducted 11 years apart, these 2 reviews collectively located only 3 unique studies each analyzing process of care.^18–20 Both Sampson et al and Murphy et al reported study design heterogeneity and the paucity of data were barriers to providing conclusions beyond that more quality clinical research in dementia is needed, citing hope future trials investigating principles of PC for advanced dementia patients would provide much needed information.^16,17 Ng et al authored a 2009 Cochrane review seeking to evaluate prospective randomized studies evaluating multidisciplinary care involving palliative medicine, neurology and rehabilitation medicine providers in the care of adults diagnosed with motor neuron disease but were unable to locate any publications meeting their criteria.¹⁴ They, therefore, concluded more quality quantitative research in the use of multidisciplinary care involving PC in this population is needed.¹⁴ The objective of Latorraca et al’s 2019 Cochrane review was to assess the effects, benefits and harms of PCI compared to usual care for people living with MS.¹⁵ Their search for any RCTs, cluster randomized trials or first phase of cross-over trials yielded only 3 RCTs^21–23; these are also evaluated in this systematic review. Their pre-specified primary outcomes were health-related QOL and adverse events (defined as “number of participants with at least one complaint of pain, sleep disturbance, or any even considered as adverse by the participant”) at 6 months.¹⁵ The following secondary outcomes at 6 months were chosen: fatigue; cognitive function; anxiety, depression; hospital admissions; relapse-free survival and sustained progression-free survival.¹⁵ To summarize, Latorraca et al found no evidence of differences between intervention and control groups > 6 months post-intervention and concluded they are uncertain whether PCIs are beneficial for people with MS in the setting of low- or very low-certainty evidence regarding the difference between PCI versus usual care for long-term health-related quality of life, adverse events, and hospital admission in patients with MS.¹⁵ Thus, all 4 review author teams reported the same conclusion: more quantitative research investigating PCI is needed to provide evidence for clinical practice. Two phase III RCTs investigating PCI in adults with PCNSDs using patient-reported outcomes and caregiver-reported outcomes to assess PCI effectiveness were published in 2020, containing 964 total participants (560 patients and 404 caregivers),^24,25 warranting initiation of an updated systematic review.

Integral to palliative medicine and clinical research alike is assessing the patient perspective. Clinical outcomes, defined as an outcome describing or reflecting how an individual functions, feels or survives, can be assessed by reports from a patient, non-clinician observer, a clinician or via a performance-based measure.²⁶ Referred to as clinical outcome assessments, they are categorized based on who the reporter is.²⁶ Patient-reported outcomes are unique as they are direct from the patient without any intermediary or interpretation about concepts and information known only by the individual.²⁶ Instruments are used to collect and measure this data. The quality of this measurement, namely its validity to measure what it is intended, its reliability in measuring consistently, and its responsiveness to detect change, are critical determinants of that first-hand outcome data. In the context of clinical trials investigating PCI effectiveness to impact how patients and caregivers feel and function, patient- and caregiver-reporting of their respective outcomes across domains and the choice of instruments to measure those outcomes are critical.

There has been an unmet need for high-quality palliative care intervention trials to inform evidence-based practice of palliative medicine for adults living chronic, progressive neurologic conditions and their caregivers. This comprehensive review aims to facilitate this endeavor by systematically reviewing RCTs investigating PCI for people living with PCNSDs and their caregivers published to date, and by critically evaluating the outcomes assessed as well as the patient-reported and caregiver-reported outcome measures used to assess PCI effectiveness for their validity and their responsiveness to detect change, in order to determine their utility for neuro-palliative care and provide guidance for future research.

Methods

The methodology for this review was guided by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement checklist specifying identification of all relevant records, selection of eligible randomized clinical trials, risk of bias assessment, data extraction, qualitative synthesis.²⁷

The following inclusion criteria were chosen: 1) study design: randomized controlled trials; 2) population: adults (≥18 years old) diagnosed with one of the following progressive primary central nervous system diseases: Parkinson disease; Huntington disease; primary movement disorder, spinal cord disease or neoplasm; stroke; multiple sclerosis; motor neuron disease/amyotrophic lateral sclerosis; or Alzheimer disease given that dementia is a clinical syndrome referring to impaired cognition leading to inference with activities of daily living and can arise as either a primary or secondary condition; 3) intervention/exposure: receipt of palliative care services, whether by primary or specialty providers, distinct from usual services provided by a primary medical team and excluding services aiming to provide end of life care or investigate the impact or validity of using a communication, tool or guide focused on a single outcome such as completion of an advanced directive or service enrollment; 4) control: usual or standard of care arm which may receive same services at a later time; 5) setting: ambulatory, home, inpatient not associated with any other adjunctive palliative, hospice, or end-of-life care or services; 6) outcomes: patient- and/or caregiver-centered outcomes such as function, symptoms, burden and health-related quality of life; 7) publication status and language: studies published in a peer-reviewed journal in English with no limits on publication date.

A search strategy of EMBASE, PubMed, Scopus, Web of Science using index and keyword methods for articles published from inception through 28 February 2021 in English was formulated and conducted by an expert research librarian in collaboration with the review authors (Appendix Table 1).

Titles of the resulting articles were screened by the review authors (H.L., T.S.A.) after duplicates were removed. The review authors independently reviewed the abstracts of the screened titles to identify full-text articles through assessment of inclusion and exclusion criteria (Table 1). Discrepancies were few, limited to clarification of articles as having outcomes which were patient-centered, and were resolved through discussion. Reference lists of the included studies and systematic reviews were reviewed to identify any additional relevant articles.

Table 1.

STUDY INCLUSION/EXCLUSION CRITERIA

Inclusion Criteria	Exclusion Criteria
1. Peer-reviewed randomized trials	1. Non-English language
2. Human studies	2. Reviews, abstracts, case reports, dissertations, qualitative and mixed method studies
3. English language	3. Non-primary central nervous system diseases
4. All publication years available	4. Disease-related outcomes
5. Adult patients
6. Progressive primary central nervous system diseases
7. Patient-centered outcomes

The following data were systematically extracted by the review authors from each article: population description (number enrolled, diagnosis, patient and caregiver demographics as detailed in the publication), participant inclusion and exclusion criteria, study setting, intervention description (including primary and secondary outcomes, PCT composition, visit frequency), self-report instruments employed to assess specified outcomes between intervention and control arms at protocol specified time interval, author conclusions with statistical results as provided.

Risk of bias for individual studies was assessed and summarized by the review authors (H.L. and T.S.A.) using the Cochrane Collaboration’s Tool for Assessing Risk of Bias in Randomised trials²⁸ and RoB2: a revised tool for assessing risk of bias in randomized trials.²⁹ Discrepancies pertaining to risk of bias domain assessments for each study and the aggregate assessment were few, limited to clarification of handling of missing data, and were resolved through discussion.

Results

Study Selection

As outlined in the PRISMA Flow Diagram (Figure 1), the search identified 483 records for title review of which 454 were excluded yielding a total of 29 records. Of these 29 articles eligible for full-text review, 4 RCTs^{23–25,30,23,24,} met the inclusion/exclusion criteria listed in Table 1 for qualitative synthesis. One of the 4 RCTs published the protocol, different aspects of the results, and analyses in 5 different articles spanning from 2006 to 2011^{31,32,30,21,33} rather than a single comprehensive publication; each article was reviewed for data extraction and risk of bias analysis. Another RCT published results and analyses in 2 different publications.^25,34 Three Cochrane review^14–16 and 1 systematic review¹⁷ were also identified and screened for any additional RCTs. One RCT²² met this review’s inclusion/exclusion criteria, yielding a total of 5 RCTs for qualitative synthesis. Due to the limited search results, qualitative synthesis without meta-analysis was undertaken.

Figure 1.

PRISMA 2009 Flow Diagram.

Study Characteristics

Study characteristics of the 5 RCTs located, published between 2006 and 2020, are summarized and listed by publication date in Table 2. Three trials were phase II and 2 trials were phase III. Of the 3 phase II RCTs, 2 enrolled people living with MS and their caregivers (in London, England and 3 locations in Italy, respectively) and the third enrolled people living with MS, motor neuron disease or a movement disorder and their caregivers in Turin, Italy.^30,21,22,23 Of the 2 phase III trials;^24,25,34 there was a pragmatic design study investigating an outpatient integrated palliative care model of delivering PCI in an outpatient movement disorder clinic setting at 3 different academic centers ( 2 in the United States and 1 in Canada).²⁴ and the other investigated specialty palliative care delivered as a brief, targeted PCI in the outpatient setting for patients and their caregivers already receiving primary medical and specialty neurologic care for MS, motor neuron disease or a movement disorder at one of 7 different sites across the United Kingdom.^25,34

Table 2:

CHARACTERISTICS OF INCLUDED STUDIES

Author Year	Study Setting Palliative Care Team Composition	Study Design Visit Frequency	Primary Outcomes And Measures Used	Secondary Outcomes and Measures Used	Author Conclusions
Higginson et al 200622, 200823, 200919, 201125/ Edmonds et al 201024	▪ Single site (London, England) ▪ home, outpatient clinics, nursing homes, hospitals ▪ PCT: specialty PC team of a PC physician, a PC clinical nurse specialist, a administrator, and a psychosocial worker connected with a hospital PC psychosocial worker. PCT collaborated with MS clinic neurologist.	▪1:1 randomization to intervention arm for immediate referral to specialty PC service versus usual care arm with 12-week delayed referral to same specialty PC service. ▪ Participants had 1–3 PCT contacts (mean 3; in-person and/or by phone).	• Change in 5 symptoms (pain, nausea, vomiting, mouth problems and sleeping difficulty) as measured by POS-S-MS	PROs: Illness burden: - MSIS Palliative care needs: -POS-8: anxiety, patient and caregiver concerns, practical and informational needs CROs: Caregiver burden: - ZBI–12 - MLPQ	Primary Outcomes: • POS-S-MS 5 symptoms improved in intervention arm and deteriorated in usual care arm at 12 weeks, F-test = 4.75, p=0.035). Secondary Outcomes: Patients: • POS-8 trended towards improvement with effect size of 0.6, p-value non-significant. • Both arms had slight improvement in psychosocial and service needs and little change in MSIS physical or psychological scales. Caregivers: • Reduction in caregiver burden at 12 weeks observed per ZBI-12 score in intervention arm compared to usual care arm, with an effect size of 1.3 (F=7.6, p=0.013). • No data reported for MLPQ.
Veronese et al 201532	▪ Single site (Turin, Italy) ▪ Home location ▪ PCT: specialty PC service team of a physician, a nurse, a psychologist and a physiotherapist employed by hospice and PC organization.	▪ 1:1 randomization to intervention arm for immediate referral to specialty PC service versus usual arm with 16-wk delayed referral to same specialty PC service ▪ Participants visited on average at least weekly by a PCT team member; all patients discussed every 2 weeks.	Co-Primary Outcomes at 16 weeks: ▪ Patient QOL:  - SEIQOL-DW ▪ Caregiver burden:  - CBI	PROs: Symptom severity: -Physical symptoms per numerical rating scale 0–10 ▶Pain ▶Breathlessness ▶Sleep disturbance ▶Urinary symptoms ▶Bowel symptoms ▶Oral symptoms Anxiety and depression: - HADS Psychological, social and spiritual issues per numerical rating scale 0–10 -Feeling of abandonment -Disease coping - Social isolation -Service satisfaction - Meaning of the experience -Help from faith	Primary Outcomes: Patient QOL: • Intervention arm SEIQOL-DW scores 20.2% higher compared to usual care arm at 16 weeks (p<0.05); ≥20% considered clinically relevant and statistically significant per ANCOVA analysis. Caregiver Burden: • No clinically or statistically significant difference in caregiver burden as measured by Caregiver Burden Inventory between arms at 16 weeks. Secondary Outcomes • Symptom reduction calculated as mean differences (in percentages) between groups at 16 weeks: ◦ 24.1% reduction in pain ◦ 22.6% in breathlessness ◦ 21.4% in sleep disturbance, ◦ 20.7% in bowel symptoms ◦ 17.4% in urinary symptoms ◦ 12.4% in oral symptoms ▪ Pain, breathlessness and sleep disturbance reductions p<0.05 • Trends towards improvement in patient and caregiver service satisfaction and meaning and worsening of depression and anxiety measured by HADS in intervention arm compared to usual care arm at 16 weeks.
Solari et al 201820	▪ Multi-site (Catania, Milan, Rome) ▪ Home location. ▪ PCT: a physician (neurologist or physiatrist), a nurse (case manager and team leader), a psychologist, and a social worker. Milan and Rome center nurses had advanced degrees and worked fulltime in PC; Catania center nurse attended a 1-week PC training course and had no prior clinical practice in PC.	▪ 2:1 randomization of severe MS-caregiver dyads to either home-based PC intervention arm or usual care arm. ▪ On average, dyads received 4.9 home visits in the first 3 months and 2.8 in the second 3 months. Nurses performed 33% of visits, psychologist 25%, physician 25%, and social worker 17%. Visits performed by ≥2 PCT members: Milan 4%, Rome 14%, Catania 51%.	Primary Outcome at 6 months: • Patient QOL - SEIQoL-DW if cognitively able, if not then: • Patient symptoms reported by caregivers using POS-SMS [sole primary outcome for participants cognitively unable to complete SEQoL-DW (anticipated in 50% of participants)]	PROs: HRQOL: - EQ-5D-3L Mood: - HADS Function: - FIM Financial toxicity: - MSCQ CROs: QOL: - SF-36 HRQOL: - EQ-5D-3L Mood: - HADS Caregiver burden: - ZBI-22 Comprehensive assessment of dyad needs based on direct observation and information from first visit (“managing everyday life” (38%), “organization” (34%) and “psychosocial” (27%)).	Primary Outcomes: Patient QOL: • Mean change in SEIQoL-DW at 6 months 0.8 (95% CI, −5.3 to 6.9) in intervention arm and −4.0 (−21.1 to 13.1) in usual care arm, with mean between-group difference = 4.8 (95% CI, −13.2 to 22.7), effect size 0.10 (p = 0.57). Patient Symptom Burden • Mean change in POS-S-MS at 6 months 2.3 (95% CI, 0.4 to 4.1) in intervention arm versus 0.3 (−2.0 to 2.6) in usual care arm (p=0.047), with mean between-group difference 1.9 (95% CI, −1.1 to 5.0); effect size = 0.32, p non-significant. Secondary Outcomes: Patients: • No significant difference observed between patients in intervention arm vs usual care arm for change at 6 months in POS, HADS and FIM total scores. Caregivers: • Mean change in ZBI score at 6 months = 0.2 (95% CI, −2.8 to 3.2) in intervention arm and 1.7 (−1.1 to 4.5) in usual care arm, with mean between-group difference = −1.5 (95% CI, −6.1 to 3.1; ES, 0.16). • No significant difference observed between caregiver SF-36 or HADS scores between arms • Fulfillment reported for assessed care need domains of “managing everyday life” improved from 38% to 97%; “organization” from 34% to 73% and “psychosocial” from 27% to 72% ♦ No data reported for patient or caregiver EQ-5D-3L or MSCQ.
Kluger et al 201921	• Multi-site (USA x2, Canada x1) • Outpatient clinic • Integrated primary PC administered by a neurologist, social worker, chaplain, and nurse using PC checklists, with guidance and selective involvement by a palliative medicine specialist physician versus standard of care delivered by primary care physician and neurologist.	• 1:1 randomization to intervention group versus standard of care group stratified by site, presence of caregiver, diagnosis of comorbid dementia • Intervention arm received outpatient clinic visits (in-person or telemedicine) every 3 months for 1 year; usual arm visits recorded but not mandated.	Co-Primary Outcomes at 6 months: • Group difference in change in patient QOL per patient and caregiver perspective: - QOL-AD • Group difference in the change in caregiver burden:- ZBI-12	PROs: Symptom burden: -ESAS-PD Health-related QOL: -PDQ Mood: -HADS Grief: -PGD Spiritual Well-Being: -FACIT-SP Clinical global impression of change on 7-point scale CROs: Mood: -HADS Grief: -PGD Spiritual well-being: - FACIT-SP Caregivers’ clinical global impression of change in patient, on 7-point scale	Primary Outcomes: Patient QOL: • PCI arm reported higher QOL per QOL-AD (mean 0.66 [SD 5.5] improvement versus control arm mean 0.84 [SD 4.2] worsening; treatment effect estimate 1.87 95% CI 0.47–3.27; p=0.0090). Caregiver Burden: • No significant difference in caregiver burden per ZBI-12 (mean [SD], 2.3 [5.0] improvement vs 1.2 [5.6] improvement in the standard care group; treatment effect estimate, −1.62; 95%CI, −3.32 to 0.09; P = 0.06). • Intervention arm caregivers reported a statistically significant reduction in caregiver burden as measured by ZBI-12 scores (−2.28 points; 95% CI, −3.38 to −1.18; P < .001) compared with usual care arm (−1.08 points; 95% CI, −2.28 to 0.12; P = 0.08) at 6 months, but the difference was not statistically significant in primary analysis (mean [SD], 2.3 [5.0] improvement in intervention arm vs 1.2 [5.6] improvement in usual care arm; treatment effect estimate, −1.62; 95% CI, −3.32 to 0.09; p=0.06). Secondary Outcomes: • Significant differences favoring PCI included nonmotor symptom burden, motor symptom severity, advance directive completion, caregiver anxiety and burden at 12 months. No outcomes favored standard of care alone. • Secondary analyses suggest benefits greater for persons with higher PC needs.
Gao et al 202026/Hepgul et al 202027	• Multi-site (7 United Kingdom national hospitals) • Home • PCT: Specialty PC delivered by multiprofessional specialty palliative care teams (physician, nurse, social work and psychosocial counselor) linked with local neurology, rehabilitation, primary care and hospice service.	• 1:1 randomization to intervention arm for immediate referral to specialty PC service versus usual care arm with 12-week delayed referral to same specialty PC service, with minimization for center, diagnosis, and capacity versus impaired or lacking capacity. • PCI delivered during 3 visits over 6–8 weeks.	• Change score between baseline and 12weeks in 8 symptoms (pain, spasms, dyspnea, nausea, vomiting, constipation, difficulty sleeping, mouth problems) as measured by IPOS-Neuro-S8	PROs: Changes in symptoms other than IPOS-NeuroS8 and PC needs assessment -IPOS-Neuro Psychological stress -HADS Health-related QOL -EQ-5D-VAS Satisfaction -FAMCARE-P16 Self-efficacy -SEMCD CROs: Caregiver burden -ZBI-12 Caregiver positivity -ZBI-Positivity Caregiver satisfaction -FAMCARE2	Primary Outcome: Patient Symptom Burden: • No statistically significant differences between-groups in 8 key symptoms as measured by IPOS-Neuro-S8 (effect size -0.16; 95% CI, −0.37 to 0.05, p=0.14). • Intervention arm reported improved scores from baseline to 12 weeks on IPOS Neuro- S8 (−0.78, 95% CI, −1.29 to -0.26). Secondary Outcomes: Patient Symptoms: • Intervention arm reported improved physical symptom burden from baseline to 12 weeks as measured by IPOS-Neuro (–1.95, 99.55% CI -3.60 to -0.30). All other PROs/CROs: • No statistically significant differences in any other secondary outcomes (effect size range, −0.20 – 0.12; p= 0.06 – 0.90).

Abbreviations: CBI: Caregivers Burden Inventory; CRO: caregiver reported outcome ESAS-PD: Edmonton Symptom Assessment Scale-revised for Parkinson Disease; EQ-5D-3L: European Quality of Life Five Dimensions-3 level version; EQ-5D-5L: European Quality of Life Five Dimensions-Five level version; EQ-5D-VAS: European Quality of Life Five Dimensions-Visual Analog Scale; FACIT-SP: Functional Assessment of Chronic Illness Therapy-Spiritual Well Being; FAMCARE-P16: FAMCARE patient satisfaction version-16 items; FAMCARE2: FAMCARE caregiver satisfaction version-17 items; FIM: Functional Independence Measure; HADS: Hospital Anxiety and Depression Scale; IPOS-Neuro: Integrated Palliative Outcome Scale for long term neurological conditions; IPOS-Neuro-S8: Integrated Palliative Outcome Scale-Neuro-Symptoms 8; MLPQ: Modified Lawton Positivity Questionnaire; MSCQ: Multiple Sclerosis Foundation Costs Questionnaire; MSIS: Multiple Sclerosis Impact Scale; PC: Palliative Care; PCI: Palliative Care Intervention; PCT: Palliative Care Team; POS-8: Palliative Outcome Scale-8 items modified from Palliative Outcome Scale-10 items; POS-S-MS: Palliative Care Outcome Scale-Symptoms-Multiple Sclerosis version; PDQ: Parkinson’s Disease Questionnaire; PGD: Prolonged Grief Disorder; PRO: Patient reported outcome; QOL: quality of life; QOL-AD: Quality of Life in Alzheimer’s Disease; SEIQOL-DW: Schedule for the Evaluation of Individual Quality of Life Direct Weight; SF-36: Short Form-36 items; SEMCD: Self-Efficacy to Manage Chronic Disease; ZBI: Zarit Burden Interview (−12 item and −22 item versions); ZBI-Positivity: Zarit Burden Interview-Positivity.

The palliative care domains as defined by the National Consensus Project for Quality Palliative Care publication Clinical Practice Guidelines for Quality Palliative Care, 4^th edition³⁵ incorporated by each of the 5 RCTs are summarized in Table 3. The number of domains incorporated ranged from 4 to 6 without correlation to phase, study size, study population or palliative care delivery model. Four RCTs (all 3 phase II trials and 1 phase III trial) investigated specialty palliative care model of PCI delivery and conducted their PCT visits at the patients’ homes^{21–23,25,30–34} with 2 studies also accommodating visits in other medical settings as warranted^{21,25,30–32,34} although only 1 specifying how often this occurred.³² However, only 3 of these 4 specialty palliative care model studies employed palliative medicine trained physicians and nurses in their interdisciplinary PCTs^{21,22,25,30,32–34,36} as 1 study’s PCT physician members were neurologists or physiatrists.²³

Table 3.

DOMAINS OF PALLIATIVE CARE ADDRESSED BY INCLUDED STUDIES

	Higginson 2006, 2008, 2009, 2011/Edmonds 2010	Veronese et al 2015	Solari et al 2018	Kluger et al 2020	Gao et al 2020/Hepgul et al 2020
Domain 1: Structure and Processes of Care	+	+	+	+	+
Domain 2: Physical Aspects of Care	+	+	+	+	+
Domain 3: Psychological and Psychiatric of Care	−	+	+	+	−
Domain 4: Social Aspects of Care	+	+	+	+	+
Domain 5: Spiritual, Religious, and Existential Aspects of Care	−	+	−	+	−
Domain 6: Cultural Aspects of Care	−	−	−	−	−
Domain 7: Care of the Patient Nearing the End of Life	−	−	−	−	−
Domain 8: Ethical and Legal Aspects of Care	+	−	−	+	+

“+” = Based on the study publication text(s), palliative care/interdisciplinary team composition, primary/secondary outcomes assessed, this study was judged as incorporating the respective domain as outlined by in Clinical Practice Guidelines for Quality Palliative Care, 4^th edition, National Coalition for Hospice and Palliative Care.³³

“−“ = Based on the study’s publication text(s), PCT composition, primary/secondary outcomes assessed, this study was judged as not having incorporated the respective domain as outlined by in Clinical Practice Guidelines for Quality Palliative Care, 4^th edition, National Coalition for Hospice and Palliative Care.³³

All 5 RCTs implemented interdisciplinary PCTs as summarized in Table 2. The duration of the PCI ranged from 3 to 6 months across the 5 RCTs.^{21–25,30–34} The frequency of encounters involving at least 2 PCT members during the PCI timeframe ranged from weekly during a 16-week intervention timeframe²² to twice in a 6-month intervention frametime.²⁴ The control arms in 4 of the 5 studies were to receive the same specialty palliative care services as the intervention arm after the primary outcome assessment.^{21–23,25,30–34}

Study Population Characteristics

Sample sizes ranged from 52 to 579 among the 5 RCTs. Studies were conducted in a range of different healthcare systems given their respective locations in the United Kingdom, Italy, Canada and the United States. Of the 1,319 pooled participants whose characteristics are provided in Table 4, 56% were patients, 53% were male and nearly 55% had a movement disorder. The mean pooled patient age was 61.9 years. Pooled patient race and caregiver demographics were unable to be compiled due to underreporting; each study’s demographic information is summarized in Table 5. Three trials reported patient race, the majority of which were white.^{21,24,25,30,34} Each study had clear disease-specific inclusion criteria selecting patients with high disease burden as listed in Table 5. In the case of studies which enrolled those diagnosed with MS, the patients had been living with their illness for many years.^{21,23,25,30,34} Patient cognitive impairment was assessed either formally by testing^{21,24,25,30–34} or informally by clinical judgement^22,23 in each study, and only one study excluded patients unable to self-report outcome assessments.²²

Table 4.

POOLED PARTICIPANT CHARACTERISTICS

Pooled Participant Characteristics	Count	%
Total Participants	1,319
Patients	740	56
Caregivers	579	44
Diagnosis
movement disorder	405	54.7
multiple sclerosis	296	40
motor neuron disease	39	5.3
Patient Sex
Male	393	53
Female	345	47

Table 5:

STUDY PARTICIPANT CHARACTERISTICS

Author Year	Diagnosis	Inclusion/Exclusion Criteria	Demographics
Higgins et al 200622 2008,23 200921, 201125 /Edmonds et al 201024	MS	▪ High EDSS (≥8 advised; referrals based on need, not disability) ▪ Excluded if referring clinician deemed patient to have urgent needs, screened by palliative medicine consultant.	N Patients = 52 - Primary progressive MS = 44% -Secondary progressive MS = 50% - other MS subtype or related disease = 6% -mean disease duration = 18 years (SD 9.6) -mean EDSS score = 7.8 N Caregivers = 43 Intervention arm: n patients = 26  - Female = 65%  - mean age = 53years  - White = 89% n caregivers = 20  - Partner/Spouse = 14  - Child = 4  - Sibling = 1  - Parent = 1 Usual care arm: n patients = 26  - Female = 73%  - mean age = 53years  - White = 92% n caregivers = 23  - Partner/Spouse = 15  - Child = 3  - Sibling = 2  - Parent = 3
Veronese et al 201532	MND, MS, PD, MSA, PSP	General signs of deterioration in neurological diseases: ▶ Progressive deterioration in physical or cognitive function despite optimal therapy ▶ Complex symptoms ▶ Swallowing problems leading to respiratory complications ▶ Speech problems: dysarthria and progressive dysphasia ▪ Excluded if cognitively unable to complete outcome measures (augmentative/alternative communication aids incentivized) Disease specific criteria: -MND:  ▶ Disturbed sleep due to respiratory muscle weakness and dyspnea at rest and/or vital capacity <50% predicted  ▶ Increased cognitive difficulties or barely intelligible speech  ▶ Difficulty swallowing with poor nutritional status/weight loss  ▶ Needs assistance with ADLs  ▶ Medical complications (pneumonia, sepsis) - MS:  1. EDSS score ≥8.0  2. Presence of at least one of the following conditions:   ▶ Significant complex symptoms and medical complications   ▶ Dysphagia admissions with sepsis and poor nutritional status   ▶ Communication difficulties   ▶ Breathlessness - Movement disorder:  1. Hoehn and Yahr stage ≥4  2. No indication for neurosurgical procedures  3. The presence of ≥2 of the following criteria:  ▶ Medications is less effective or increasingly complex regime ▶ Reduced independence, needs help with daily living   ▶ Condition less controlled, less predictable with “off” periods   ▶ Dyskinesias, mobility problems and falls   ▶ Psychiatric signs (depression, anxiety, hallucinations, psychosis)	N Patients = 50 N Caregivers = 45 Of 50 patients:  -n male = 30(60%)  -n female = 20(40%)  -mean age = 61years (n = 8 < 44years)  -n ALS/MND = 16(32%)  -n MS = 18(36%)  -n MD = 16(32%) Of 45 caregivers:  - Wife = 24(53%)  - Husband = 12(27%)  - Daughter = 4(9%)  - Mother = 1(2%)  - Sister = 1(2%)  - Son = 1(2%)
Solari et al 201820	MS (primary or secondary progressive)	• age ≥18 years • non-institutionalized • EDSS score ≥8.0 or higher • complex symptoms • ≥2 unmet care needs	N = 76 dyads Intervention arm n patients = 50  - Female = 62%  - mean age = 60.5years  - mean age @dx = 37.5years +/−13.8years n caregivers = 50 -Female = 62% -mean age = 60.1years  - Spouse/partner = 25(50%)  - Parent = 8(16%)  - Child = 8(16%)  - Other relative = 7(14%)  - Paid caregiver = 2(4%) Usual care arm n patients = 26  - Female = 46%  - mean age = 56.8years  - mean age @dx = 45.7years +/−10.9years n caregivers = 26  - Female = 61%  - mean age = 60.8years  - Spouse/partner = 15(58%)  - Parent = 4(15%)  - Child = 0  - Other relative = 6(23%)  - Paid caregiver = 1(4%)
Kluger et al 201921	Probable PD, MSA, CBD, PSP, LBD	• Moderate to high palliative care needs based on PC-NAT modified for Parkinson disease • Fluent in English ▪ Excluded if already receiving PC or if site investigator judged presence of clinically urgent PC needs or other illnesses possibly requiring PC.	N Patients = 210 N Caregivers = 175 Intervention Group: n patients = 106  - Female = 61%  - mean age = 69.5years  - White = 94% n caregivers = 87  - Female = 71%  - mean age = 65.7years  - White = 94%  - Spouse = 81%  - Child = 12%  - Other = 8% Usual Care Group: n patients = 104  - Male = 67%  - mean age = 70.7years  - White = 89% n caregivers = 88  - Female = 75%  - mean age = 66.4years  - White = 88%  - Spouses = 83%  - Child = 8%  - Other = 9%
Gao et al 202026/Hepgul et al 202027	MS, MND, IPD, PSP, MSA	1. Age ≥18 years old 2. Advanced or progressive stages of disease:  • MS (EDSS score ≥7.5)  • MND (all stages)  • IPD (Hoehn and Yahr stages 4–5)  • PSP (adapted Hoehn and Yahr stages 3–5)  • MSA (adapted Hoehn and Yahr stages 3–5) 3. An unresolved symptom unresponsive to standard care 4. At least 1 of the following:  • Another unresolved symptom  • Cognitive problems or complex psychological issues  • Communication or information issue, or complex social need ▪ Exclusions: already receiving specialist palliative care or had in last 6 months, lacking mental capacity with lack of a proxy to provide data Caregiver inclusion criteria: • Adults ≥18 years old identified by patient as a person close to them who is able and willing to complete questionnaires for study and give informed consent.	N Patients = 350 N Caregivers = 229 Intervention Group: n patients = 176  - Female = 51.1%  - mean age = 67.3years (SD 10.9)  - White = 94.3%  - Median years (range) dx = 12.3 (0–56)  - n MS = 74 (42.1%)  - n MND = 11 (6.3%)  - n IPD = 71 (40.3%)  - n PSP = 13 (7.4%)  - n MSA = 7 (4.0%) n caregivers = 121  - Female = 66.1%  - mean age = 63.3years  - White = 93.4%  - Spouse = 67.2% Usual Care Group: n patients = 174  - Female = 46.6%  - mean age = 66.4years  - White = 86.2%  - Median years (range) dx = 12.4 (0–46)  - n MS = 74 (42.5%)  - n MND = 12 (6.9%)  - n IPD = 69 (39.7%)  - n PSP = 14 (8.1%)  - n MSA = 5 (2.9%) n caregivers = 108  - Female = 63%  - mean age = 65.3years  - White = 90.7%  - Spouse = 84.3%

ADLs: activities of daily living; CBD: corticobasal degeneration; Dx: diagnosis; EDSS: Expanded Disability Status Score; IPD: idiopathic Parkinson disease; LBD: Lewy body dementia; MND: Motor neuron disease; MS: multiple sclerosis; MSA: multiple system atrophy; PC: palliative care; PC-NAT: Palliative Care Needs Assessment Tool; PD: Parkinson’s Disease; PSP: progressive supranuclear palsy.

Outcome Measures

Fifteen unique patient-reported and 6 unique caregiver-reported outcome measures were used across the 5 RCTs; these are summarized with validation and responsiveness assessments in Table 6. Validation and responsiveness criteria were based on the Consensus-based Standards for the selection of health Measurement Instruments (COSMIN) guideline for systematic reviews of patient-reported outcome measures.³⁷ If more than 1 version had been developed over time, the original instrument’s validation and responsiveness was assessed first then that of the modified version(s). The only measure assessed as “insufficient” for its validation was Edmonton Symptom Assessment Scale-Parkinson’s Disease (ESAS-PD). Quality of Life-Alzheimer’s Disease (QOL-AD),^38,39 Palliative Outcome Scale (POS),⁴⁰ and Integrated Neuro Scale (IPOS-Neuro)^41,42 have only been partially validated. Palliative Outcome Scale-Symptoms MS version (POS-S-MS)⁴³ and IPOS-Neuro, a profile comprised of 3 symptom assessments (IPOS-8, IPOS-Neuro-S8⁴¹ and IPOS-Neuro-S24⁴²), were not validated at the time of its initial use and remain only partially validated. Although there was no instrument common to all 5 of the included studies, a version of the Zarit Burden Interview (ZBI) assessing caregiver burden was used in 4 studies.^{30,21,23,24,25,34} Responsiveness to change assessed as “sufficient” was limited to MS Impact Scale,^44,45 Zarit Burden Interview 22-item version,⁴⁶ and European Quality of Life Five Dimensions 3 level version (EQ-5D-3L) in MS patients.⁴⁴ The EQ-5D 3 level⁴⁷ and 5 level⁴⁸ versions were considered as unique instruments given the chronology of their development and validation. Numerical rating scales for physical and non-physical symptom assessments other than mood were used by 1 phase II trial.²²

Table 6.

OUTCOMES MEASURES USED, IN ALPHABETICAL ORDER

Patient-Reported
Instrument	Measurement Strategy Scoring	Domains Population	Validated	Responsiveness
Edmonton Symptom Assessment Scale-Parkinson Disease (ESAS-PD) 66,67	Profile score Specific measure -14-items rated on 11-point scale -Score range 0–140; 0 indicates no symptoms	Pain, fatigue, nausea, anorexia, depression, anxiety, dyspnea, drowsiness, well-being, stiffness, constipation, confusion, dysphagia. • Symptomatic oncology patients receiving palliative care	ESAS + ESAS-PD −	ESAS + ESAS-PD • Responsive to physical symptom changes resulting from medical treatment
European Quality of Life Five Dimensions – Three Levels (EQ-5D-3L) 44,47	Index score Generic measure -5-items rated at 3 levels and a visual analog scale rated 0–100; 0 indicates worse possible health status	Mobility, self-care, usual activities, pain/discomfort, mood, general health status • General population	+	• Moderately sensitive to change in MS patients with median EDSS 5.5
European Quality of Life Five Dimensions – Five Levels (EQ-5D-5L) 48	Single indicator Generic measure -5-items rated at 5-levels and a visual analog scale rated 0–100; 0 indicating worse possible health status	Mobility, self-care, usual activities, pain/discomfort, depression/anxiety, general health status • General population	+	+
FAMCARE-P16 68	Index score Generic measure-16-items rated on 5-point scale -Score range 16–80; higher scores indicate higher satisfaction with palliative care	Satisfaction with information-giving, care availability, physical and psychological care Oncology patients receiving palliative care	+	?
Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being (FACIT-SP-12) 69,70	Index score General measure-12-items rated on 5-point scale -Score range 0–48; 0 indicating low spiritual well-being	Spiritual well-being, meaning, peace, faith Oncology patients	+	?
Functional Independence Measure (FIM) 71,72	Index score Specific measure -18-items rated on 7-point scale -2 subscales (motor function and cognitive function) each scored 0–100; higher scores indicate higher function	Disability in motor and cognitive functions including self-care, mobility, and communication. Patients receiving neurorehabilitation	+	+ • Responsiveness to change limited by scoring guidelines
Hospital Anxiety and Depression Scale (HADS) 73,74	Profile score Generic measure -Two 7-item subscales -Subscale score range 0–21; 0 indicating little to no likelihood for depression or anxiety respectively.	Symptoms of depression and anxiety Adults 16–65 seeking-medical attention in a general outpatient clinic	+	Oncology and cardiac populations +
Integrated Palliative Outcome Scale-Neuro (IPOS-Neuro) 41,42	Profile score Specific measure -42 items rated on 5-point scale -Subscales: IPOS Neuro 8; IPOS Neuro-S8; IPOS Neuro-S24 -Score range 0–168; higher scores indicate more symptom impact.	IPOS Neuro 8: Psychological and spiritual well-being, information needs, practical issues IPOS Neuro-S8: Pain, shortness of breath, nausea, vomiting, constipation, mouth problems, spasms, difficulty sleeping) IPOS Neuro S24: 24 physical symptoms (IPOS Neuro-S8 symptoms plus 16 additional) • People living with MS, MD	IPOS-Neuro   ? IPOS Neuro  • Partially validated IPOS Neuro  • Partially validated using data from Gao et al/Hepgul et al	IPOS-Neuro ? IPOS-Neuro-S8 ? IPOS-Neuro S24 ?
Multiple Sclerosis Impact Scale (MSIS-29) 44,45	Index Score Specific measure -29-items -Physical impact scale score range 20–100; 100 indicating greater impact. -Psychological impact scale score range 9–45; 45 indicating greater impact.	Physical health (symptoms and function) and psychological health (mood, role limitations, autonomy)	+	+
Palliative Outcome Scale (POS) 40	Profile score Specific measure -10-items rated on 5-point scale -Score range 0–40; 0 indicating no needs	Pain, non-pain symptom, patient and caregiver anxiety, financial concerns, psychological, spiritual and informational needs • Advanced cancer patients receiving hospice or palliative care	• Partially validated	• Limited
Palliative Outcome Scale-Symptom List Multiple Sclerosis (POS-S-MS) 43	Profile score Specific measure -18-items rated on 5-point scale -Score range 0–72, lower scores indicating lower symptom burden.	Pain, spasms, fatigue, dyspnea, nausea, vomiting, poor appetite, dysphagia, mouth problems, drowsiness, difficulty sleeping, constipation, bowel and bladder control, pressure sores, arm and leg function, difficulty communicating.	+ • Not validated prior to use in Higginson et al; 2013 validation study used data from 46 patients in Higginson et al	?
Parkinson’s Disease Questionnaire (PDQ-39) 39,56,62,64,75,76	Profile score Specific measure-39-items rated on 5-point scale, grouped into 8 subscales -Score range 0–100; 0 indicating better QOL	Mobility, activities of daily living, emotional well-being, stigma, social support, cognition, communication, bodily discomfort	+	• Responsiveness to change established
Quality of Life in Alzheimer’s Disease (QOL- AD) 38,39,56,60,61	Index score Specific measure with self-rating and proxy rating -13-items rate on 4-point scale -Score range 13–52; 52 indicating excellent QOL	Marriage, family, living situation, daily activities, physical health, energy, mood, finances, friends, memory, enjoyment, global • Alzheimer’s dementia population	+ • Self-rating and proxy rating discrepancie s. Roemhild 2018 • Limited validation in PD population	− • More responsive to QOL changes due to disease progression than palliative care intervention
Schedule for the Evaluation of Individual Quality of Life Direct Weight (SEIQOL- DW) 39,51–54,77	Index score Generic measure Interviewee nominates 5 domains and quantifies importance totaling 100; index score calculated by multiplying the rating of each domain weight and summing all the products. Higher scores indicate higher QOL.	5 self-selected domains • General population • MD • MND • MS	+	General population + MD? MND? MS?
Self-Efficacy to Manage Chronic Disease Scale (SEMCD) 78,79	Index score Specific measure -6-items rated on 9-point scale -Score: mean of 6 items; higher means indicate greater self-efficacy.	Fatigue, pain/discomfort, emotional distress, activity limitation, wellness, illness intrusiveness • Populations with chronic medical conditions (rarely neurologic)	+	+
Caregiver-Reported
Caregiver Burden Inventory (CBI) 57	Index score Specific measure -24-item scale with 5 subscales; lower scores indicate lower burden	Time-dependence, developmental, physical, social, and emotional burden • Dementia patient caregivers	+	?
EQ-5D-3L (as above)
FAMCARE2 80	Index score Specific measure -17-items rated on 5-point scale. -Score range 17–85; higher scores indicate higher caregiver satisfaction with patient’s care	Management of physical symptoms and comfort, provision of information, family support and patient psychological care. • Oncology patient caregivers receiving palliative care	+	?
HADS (as above)
“Modified Lawton Positivity Questionnaire”	“4 questions total (3 items and 1 yes/no)”	No information about this measure able to be located.	?	?
Short Form-36 (SF-36) 44,81	Profile score Generic measure -36 items, scored 0–100; higher scores indicate better perceived health state	Physical functioning, social functioning, role limitations due to physical and emotional problems, mental health, energy and vitality, pain, general health perception • General population	+	+ • Low responsiveness, less appropriate in nonambulatory and moderately to severely disabled populations
Zarit Burden Interview (ZBI) 46,58,65,82	Index score Specific measure ZBI-22 -22-item version rated on 4-point scale -Score range 0–88 ZBI-12 -12-item version rated on 4-point scale -Score range 0–48 -In both versions, higher scores indicate higher caregiver burden.	Health status, patient dependence, exhaustion and uncertainty, guilt and self-criticism, embarrassment, anger, frustration, psychological burden, emotional reactions, personal and role strain. • brain injury, dementia and advanced cancer patient caregivers • PD caregivers	+	ZBI-22 dementia patient caregiver population + ZBI-12 ?
“ZBI-Positivity”	“8 items for which higher scores indicate more positivity related to caregiving” per Gao et al/Hepgul et al protocol	No information about this instrument able to be located.	?	?

Note: No information about Multiple Sclerosis Foundation Cost Questionnaire (MSCQ) able to be located and no data reported. Prolonged Grief Disorder questionnaire is a symptom-diagnostic test for prolonged grief disorder hence its exclusion here as an outcome measurement instrument.

Abbreviations: EDSS: Expanded Disability Scale Score; MD: movement disorder; MND: motor neuron disease; MS: multiple sclerosis; PD: Parkinson’s Disease; QOL: Quality of Life. “+” = sufficient; “−“ = insufficient per COSMIN guideline for systematic reviews of patient-reported outcome measures³⁷,“?” = no information located.

There are several notable exceptions among to the outcome measures. The Prolonged Grief Disorder questionnaire was used by a RCT to measure grief but was excluded as an outcome measurement instrument in Table 5 because this questionnaire was developed and validated as a symptom-diagnostic tool for a specific psychiatric condition, namely prolonged grief disorder.^49,50 While the Multiple Sclerosis Foundation Cost Questionnaire was listed as a secondary outcome measure in 1 RCT²³, no information about its authors, development, content or validation could be located nor was any data reported, therefore, it was also excluded from Table 5. For the Modified Lawton Positivity²¹ and ZBI-Positivity^25,34 questionnaires included in Table 5, no other information could be located other the description of its contents in the trial’s publication. Therefore, the creation and use of these instruments were likely unique to the specific trials despite their lack of validation or responsiveness. Data was reported for ZBI-Positivity^25,34 but not the Modified Lawton Positivity.²¹

Patient and Caregiver-Reported Outcomes

All 5 RCTs used patient- and caregiver-reported outcomes as their primary and secondary outcomes (Table 2).^21–24,30 Outcomes assessed as primary outcomes, whether as single or co-primary, across the 5 RCTs included patient symptoms in 3 studies,^{21,23,25,30–34} patient QOL in 3 studies^22–24 and caregiver burden in 2 studies.^22,24 Patient symptom burden and caregiver burden were assessed using specific measures. One study chose improvement in patient QOL as the primary outcome for patients without cognitive impairment and improvement in physical symptom burden for patients with cognitive impairment impeding their ability to complete the chosen primary outcome assessment’s QOL instrument.²³ Patient QOL was assessed using the same generic global QOL measure in 2 studies, though only 1 reported statistically significant improvement, and using a specific health-related QOL measure in a third study that reported statistically significant improvement.^22–24 Of note, this generic global QOL instrument, SEIQOL-DW, was assessed as sufficiently validated but of unknown responsiveness in the patient populations being studied,^51–55 and the specific health-related QOL instrument, QOL-AD, had been validated with concerns regarding its proxy rating and whose responsiveness had not been clarified prior to its implementation within this specific trial.^38,56

Among the 3 studies that used patient physical symptoms as a primary outcome (1 used caregiver proxy reporting for patients cognitively unable self-report),^{21,23,25,30–34} only 1 study reported statistically significance for improvement restricted to 5 symptoms.^21,30–33 The outcome measures used had not been validated prior to their implementation in 2 of the 3 studies and have unknown responsiveness.^41–43 Neither of the 2 studies that used caregiver burden as a co-primary outcome reported statistically significant improvement, employing 2 different outcome measures, both of which were assessed as sufficiently validated but of unknown responsiveness.^46,57,58

Secondary outcomes assessed in patients include physical and non-physical domain symptoms, mood (anxiety and depression), impact of disease reflecting health-related QOL, grief, satisfaction with palliative care, self-efficacy, and spiritual well-being.

Positive results reported as statistically significant for any of these secondary patient outcomes were isolated to a two RCTs.^22,24 Secondary outcomes assessed in caregivers included caregiver burden, global and health-related QOL, mood (anxiety and depression), satisfaction with palliative care, and spiritual well-being. Positive results reported as statistically significant for any secondary caregiver outcome within the pre-specified assessment time frame were isolated to a single RCT reporting reduction in caregiver burden.^21,30–33

Risk of Bias Assessment

The overall assessment for risk of bias of the studies resulted as low risk. Appendix Table 2 summarizes each risk of bias domain appraisal for each study. Figure 2 presents the overall risk of bias assessments for each study. As all 5 RCTs investigated PCI via a PCT approach, whether in a specialty, integrated or primary palliative care model, each was assessed as having the same inherent risks of bias with respects to blinding of participants and personnel (low risk) and blinding of outcome assessment (some concerns) in the absence of any evidence indicating additional risks were present per Cochrane recommendations.^28,59 Possibly to due to evolving standards in clinical trial reporting across time and journals, reporting bias was a particular concern in 2 studies; 1 study lacked data reporting of 2 outcome assessments, raising concerns whether these data were collected or selectively not reported due to missing data or negative results,²³ and another study published highly inconsistent outcomes data of unclear distribution across arms.^25,34 The attrition rates ranged from 4% to 13% across studies. Reporting of missing data management and the impact of missing data, dependent on cause of missingness, on the study’s respective statistical analysis was inconsistent across the publications for the 5 included studies. Allocation concealment was of unclear risk in 2 studies due to lack of specification regarding referral screening^22,23 and some concerns in 1 study due to site investigator participation in the screening process.²⁴ In a phase II study there was a large discrepancy in the number of PCT visits conducted at 1 site compared to the other 2 sites.²³

Figure 2.

ROB chart grayscale.

Discussion

There is growing interest in optimizing the care of people living with chronic, progressive neurologic conditions and their caregivers using palliative medicine principles. High-quality evidence demonstrating the positive impact of these principles are needed to guide care. The goal of this review is to identify such evidence stemming from the existing literature and to inform future research investigating PCI in this specific population. Our search for RCTs investigating the effectiveness of PCI in adults living with PCNSD and their caregivers using self-reported outcomes yielded only 5 RCTs, of which 3 were pilot studies, therefore a qualitative synthesis without meta-analysis was conducted. The PCIs were associated with a modest lowering of physical symptom burden of specific symptoms, primarily pain and sleeping difficulty,^{21,22,30–33} a modest improvement in patient QOL of uncertain clinical relevance^22,24 and a slight reduction in caregiver burden.^{21,24,30–33} The confidence in the certainty and reliability of these findings are low due to the low number of total participants, and inconsistency in findings across the 5 RCTs in setting of data missingness and heterogeneous approaches to measuring outcomes. In only 1 trial did participants not have access to palliative medicine trained physicians and in this 3-site trial a single site’s nurse had no prior clinical palliative medicine experience.²³ Of note, this particular trial reported no positive results of statistical significance for any of its specified outcomes in its advanced stage MS study population.²³

The primary outcomes assessed across the 5 RCTs were reduction in physical symptoms, improvement in patient QOL, and reduction in caregiver burden. Patient QOL was reported as improved by 2 trials using different instruments and analytic methods to calculate their results and no improvement in the three additional trials. Veronese et al used SEIQOL-DW, a generic global QOL measure validated in each of the phase II study’s disease populations of MS, movement disorders and motor neuron disease but without clear evidence of responsiveness to change of any etiology. With a total study population of only 50 randomized 1:1 into 2 arms, Veronese et al reported a statistically significant 20.19 (20.2.%) between group difference in mean change scores at 16 weeks and designated this finding as clinically significant based on their criteria of differences between groups as 20% or higher as being clinically relevant.²² Veronese et al did not specify whether an intent-to-treat or per-protocol analysis was undertaken in the setting of attrition of 4 participants in the intervention arm and 2 in the control arm and no description of missing data management. In contrast, Kluger et al used a patient health-related QOL measure, QOL-AD, in their phase III pragmatic design study of 210 patient participants living with a movement disorder, primarily Parkinson’s disease.²⁴ This trial also randomized 1:1 into 2 arms; at the primary outcome assessment timepoint of 6 months, the control arm contained 94 participants and the intervention arm had 96. At 6 months, the authors reported mean improvement of 0.66 (SD 5.5) compared to 0.84 (SD 4.2) worsening with treatment effect estimate 1.87 [(95% CI, 0.47–3.27); p=0.009]. An interpretation of this treatment effect estimate figure and whether there is an associated minimal important difference to determine clinical relevance or significance was not offered in the trial publication. However, the authors provide a change in the QOL-AD score of at least 3 points as being a clinically significant change for an individual patient’s score. Kluger et al also conducted a clinimetric comparison of QOL scales including PDQ-39 and QOL-AD using the trial’s data that was published separately. This clinimetric study estimated an absolute minimally clinically important difference (MCID) value for QOl-AD as 3.9 and for PDQ-39 as 12.7.⁵⁶ The baseline mean QOL-AD score for the intervention arm was reported as 33.9 (SD 5.7) and for the control arm as 34.3 (SD 5.6).²⁴ If the treatment effect estimate is the mean between group difference, then the MCID of 3.9 can be compared to the between group difference of 1.87, indicating the threshold for a clinically important change in patient health-related QOL as measure by QOL-AD was not met. This comparison aligns with the study’s publication graphical depiction of the QOL-AD scores ranging between 32 and 36. Furthermore, the clinimetric study summarized in its findings that QOL-AD responsiveness to change was determined to have occurred in the control arm which the study authors thought was more likely experiencing disease progression relative to the PCI arm, therefore, they did not strongly endorse use of QOL-AD as a primary outcome measure in PCI studies in Parkinson’s disease populations.⁵⁶ Additionally, Kluger et al cited part of their rationale for using the QOL-AD was based on its proxy-rating and utilized proxy rating in their study. The reliability of proxy-rating for QOL-AD has been found to be uncertain given significant discrepancies between the self-reporting and proxy-rating of this instrument have been reported⁶⁰ and QOL-AD proxy-rating by community-dwelling dementia caregivers has been found to be highly impacted by caregiving experience and caregiver mood.⁶¹

The inclusion by Kluger et al of another patient health-related QOL measure, the PDQ-39, as a secondary outcome allows for another vantage point of whether their PCI was effective in improving patient health-related QOL. The PDQ-39 is scored as a profile of 8 subscales with higher values as better, is validated and has been found to be at least partially responsive to change including at the level of individual subscales.^56,62–64 The treatment effect estimate for PDQ-39 at 6 months was reported as −2.63 (95%CI −5.72 to 0.46, p=0.10)²⁴ would appear to not be clinically significant because this value is less than the reported MCID of 12.9. In the clinimetric study by Holden et al, PDQ-39 was reported to show more responsiveness to change than QOL-AD in the intervention group in this trial, which Holden et al suggested may mean PDQ-39 is more sensitive to QOL changes in people diagnosed with Parkinson’s disease or a related disorder receiving PCI.⁵⁶ However, the standardized response means of both QOL-AD and PDQ-39 yielded values at or below the 0.2 cut off for “small” effect per Cohn’s criteria, the statistical test for assessing responsiveness to change.⁵⁶ Further analysis of this trial’s patient- and caregiver-reported outcome results may provide insights into the dynamic relationship between health-related QOL of both caregivers and patients, and overall future research into the changes in PDQ-39 and QOL-AD due to PCI may be warranted. Examining QOL outcomes in illnesses with heterogeneous rates of progression by determining clinically significant improvement or worsening may be a useful approach for future studies.

In contradistinction to these 2 positive patient QOL results are the multiple negative results from the other patient QOL assessments used as secondary outcomes across the 5 RCTs assessed using multiple different measures, as well as the negative results from a recently published systematic review of PCI in adults with chronic noncancer illness which found there was no significant difference in QOL in its analysis of pooled data from 13,664 participants across 28 trials involving fewer than 50% cancer patients. In summary, the confidence in the reliability and the clinical significance of the positive patient QOL improvement results in 260 participants in this review are low. Discussions about the assessment of patient QOL in the context of different advanced serious illnesses may be warranted for researchers of palliative medicine interventions in planning future research and interventions.

Caregiver burden reduction was used as a co-primary outcome by both studies reporting patient QOL improvement. Interestingly, neither study reported statistically significant reduction of caregiver burden between arms within their pre-specified assessment timeframes of 16 weeks and 6 months, respectively,^22,24 although, Kluger at al reported the difference between groups became statistically significant at 12 months. Caregiver burden in Parkinson’s disease populations has been predicted by patient mood and disease severity including disability, caregiver psychological well-being and both patient and caregiver health-related QOL.⁶⁵ Though these 2 trials with movement disorder populations did not demonstrate caregiving burden reduction, Higginson et al/Edmonds et al assessed caregiver burden as a secondary outcome and reported a between group reduction at 12 weeks with an effect size of 1.3 in their advanced stage MS population trial.^21,30–33 Two different instruments, the Caregiver Burden Inventory⁵⁷ (CBI) and the Zarit Burden Interview-12 item version (ZBI-12),^46,58 were used to assess caregiver burden in these 3 RCTs.^{21,22,24,30–33} No rationale was offered by Veronese et al regarding their choice of the CBI; this instrument is validated in caregivers of adults living with dementia and its responsiveness to change is uncertain.⁵⁷ Kluger et al cited their rationale for choosing ZBI-12 was its wide use in other studies evaluating caregiver burden in Parkinson’s disease populations.²⁴ Indeed ZBI-12 is well-validated including in the Parkinson’s disease population, although its responsiveness has been not well-established.⁶⁵ In summary, impact on caregiver burden was not consistently demonstrated across studies, and within those studies reporting patient QOL improvement, the signal of caregiver burden reduction was weak. The lack of caregiver burden reduction may be due to several factors including, but not limited to, the chronicity of high disease burden especially those with MS or intensity of disease and symptom burden with acute to subacute changes in movement disorder and motor neuron disease in addition to insufficient duration and/or frequency of PCI, and variation in missing data management.

Lastly, only a limited number of specific physical symptoms were reported as statistically significantly improved between groups. The only trial reporting positive between group results for physical symptoms assessed as a primary outcome, Higginson et al/Edmonds et al reported a statistically significant difference in pain, nausea, vomiting, mouth problems and sleeping difficulty at 12 weeks, F = 4.75 (p=0.035).^21,30–33 POS-S-MS was the outcome measure used, which at that time had not been validated and its responsiveness has yet to be established.⁴³ Veronese et al used change in patient physical symptoms rated on a numerical scale as a secondary outcome, and of the 6 physical symptoms surveyed reported a 20.2% reduction (p<0.05) between arms in pain, breathlessness, sleep disturbance and bowel symptoms assessed at 16 weeks.²² Thus overall, Veronese et al were the only authors to report improvement in both patient QOL and in specific physical symptoms in their primary results, and among 740 total patients pain and sleeping difficulty were the only individual physical symptoms common to both Veronese et al and Higginson et al/Edmonds et al reported as improved.

In summary, positive findings among the results of these 5 RCTs investigating the effectiveness of PCI in adults living with PCNSDs and their caregivers are limited to small signals of change in specific symptoms, patient QOL and caregiver burden. A plausible reason for the disparity in these results may be due to the heterogeneity of PCI and the use of instruments not responsive to change attributable to PCI.^56,64,66 Another possible factor is the heterogeneity of diagnoses. Indeed, Gao et al/Hepgul et al cited their negative results may arise from the heterogeneity of their phase III study’s population of adults living with MS, a movement disorder or motor neuron disease.^25,34

Future work should seek to explore symptom management and its potential impact on patient and caregiver QOL as well as caregiver burden. Future researchers will need to be deliberate in whether to target their PCI studies in heterogeneous PCNSD populations or focused disease groups, and in their choice of outcome measures used to assess the effectiveness of PCI in different domains accordingly. In addition, establishing an instrument’s ability to detect change due to receiving PCI independent of validation testing during assessment of disease-focused medications or devices may also be important. Additional challenges in patient-centered outcomes within palliative care research among people living with cognitive impairment exist due to waning ability to self-report as cognitive ability decline over time, especially in those who have the clinical syndrome of dementia in later stages leading researchers to implement only caregiver-reported and provider-reported outcomes.^83,84 The utility of proxy-reporting when severe cognitive impairment is present or is likely to arise within the context of PCI effectiveness research as implemented by Kluger et al in their integrated outpatient palliative care model is an approach identified as needing more research pertaining to its clinimetric and ethics.⁸⁵ Notably, no studies involving people living with stroke, primary spinal cord diseases or neoplasms were identified and are populations for future research.

Limitations

This review has several limitations. We did not evaluate all dementia subtypes, instead we targeted specific primary neurologic diseases resulting in cognitive impairment, for example, Alzheimer’s disease, primary movement disorders, motor neuron disease, multiple sclerosis, and stroke. Our second limitation is we did not search the grey literature and only focused on published, peer-reviewed study publications to identify quantitative research to contribute to evidence-based practice for palliative medicine in adults with PCNSD and their caregivers. Thirdly, observational studies that might have useful information were also not included. Finally, because the search was limited to English-language texts, there is a risk of language bias.

Conclusions

This review highlights there has been sparse quantitative research investigating PCI in adults living with PCNSD published to date, and that those published demonstrated PCI only very modestly impacted burden of specific physical symptoms, patient QOL and caregiver burden using patient- and caregiver-reported outcomes. There remains a significant opportunity to address the unmet need for high-quality quantitative research in those living with PCNSD and their caregivers to inform evidence-based clinical practice. Studies to date used a variety of metrics, timepoints and criteria to evaluate the effectiveness of PCI delivered in interdisciplinary teams whether in a specialty, integrated or primary palliative care model for optimizing the care of people living with PCNSD and their caregivers. Standardization of metrics, diversity of subjects enrolled and appropriate training of team members are crucial considerations when evaluating PCI in the context of qualitative research. Future research dedicated to optimizing outcome measurement instruments for studies involving people living with cognitive impairment impeding activities of daily living and their caregivers is needed and should include psychometric assessment and ethical considerations of the proxy-reporting approach. It is essential in both designing and conducting high-quality quantitative research to be well-informed about patient- and caregiver-centered outcomes and the requisite assessments. Furthermore, for trial results to be applicable in more than a single patient population and comparable across trials, the development of patient-reported outcome measures validated and relevant to a range of diagnoses is critical and should be a considered a research priority.

Key Message.

This review comprehensively summarizes and evaluates the randomized controlled clinical trials published to date investigating the effectiveness of palliative care intervention in adults living with progressive primary CNS disease and their caregivers using self-reported outcomes. This review also summarizes and analyses the respective patient- and caregiver-reported outcome measures employed.