Articles
Early palliative care for patients with advanced cancer:
a cluster-randomised controlled trial
Camilla Zimmermann, Nadia Swami, Monika Krzyzanowska, Breffni Hannon, Natasha Leighl, Amit Oza, Malcolm Moore, Anne Rydall,
Gary Rodin, Ian Tannock, Allan Donner, Christopher Lo
Summary
Background Patients with advanced cancer have reduced quality of life, which tends to worsen towards the end of life.
We assessed the effect of early palliative care in patients with advanced cancer on several aspects of quality of life.
Methods The study took place at the Princess Margaret Cancer Centre (Toronto, ON, Canada), between Dec 1, 2006,
and Feb 28, 2011. 24 medical oncology clinics were cluster randomised (in a 1:1 ratio, using a computer-generated
sequence, stratified by clinic size and tumour site [ four lung, eight gastrointestinal, four genitourinary, six breast, two
gynaecological]), to consultation and follow-up (at least monthly) by a palliative care team or to standard cancer care.
Complete masking of interventions was not possible; however, patients provided written informed consent to
participate in their own study group, without being informed of the existence of another group. Eligible patients had
advanced cancer, European Cooperative Oncology Group performance status of 0–2, and a clinical prognosis
of 6–24 months. Quality of life (Functional Assessment of Chronic Illness Therapy—Spiritual Well-Being [FACIT-Sp]
scale and Quality of Life at the End of Life [QUAL-E] scale), symptom severity (Edmonton Symptom Assessment
System [ESAS]), satisfaction with care (FAMCARE-P16), and problems with medical interactions (Cancer
Rehabilitation Evaluation System Medical Interaction Subscale [CARES-MIS]) were measured at baseline and
monthly for 4 months. The primary outcome was change score for FACIT-Sp at 3 months. Secondary endpoints
included change score for FACIT-Sp at 4 months and change scores for other scales at 3 and 4 months. This trial is
registered with ClinicalTrials.gov, number NCT01248624.
Findings 461 patients completed baseline measures (228 intervention, 233 control); 393 completed at least one followup assessment. At 3-months, there was a non-significant difference in change score for FACIT-Sp between intervention
and control groups (3·56 points [95% CI –0·27 to 7·40], p=0·07), a significant difference in QUAL-E (2·25
[0·01 to 4·49], p=0·05) and FAMCARE-P16 (3·79 [1·74 to 5·85], p=0·0003), and no difference in ESAS (–1·70
[–5·26 to 1·87], p=0·33) or CARES-MIS (–0·66 [–2·25 to 0·94], p=0·40). At 4 months, there were significant
differences in change scores for all outcomes except CARES-MIS. All differences favoured the intervention group.
Interpretation Although the difference in quality of life was non-significant at the primary endpoint, this trial shows
promising findings that support early palliative care for patients with advanced cancer.
Funding Canadian Cancer Society, Ontario Ministry of Health and Long Term Care.
Introduction
The complex needs of patients with advanced cancer and
their caregivers arise many months before the patient’s
death.1 Correspondingly, WHO defines palliative care as
“an approach that improves the quality of life of patients
and their families…by means of early identification and
impeccable assessment and treatment of pain and other
problems, physical, psychosocial and spiritual.”2 Specialised palliative care teams enact this approach, through
the holistic care of patients dying from cancer or other
terminal illnesses, and their presence is increasing
worldwide.3 However, most of these teams provide
terminal care at home or to inpatients rather than in
outpatient settings,4–6 and referral to palliative care teams
for most patients occurs in the last 2 months of life or
not at all.7,8
We did a systematic review9 of randomised controlled
trials from 1984 to 2007 assessing the effectiveness of
specialised palliative care. Not all studies assessed a
palliative care team, with interventions including a coordinating service, a nursing intervention, or counselling.
Only four of 13 studies assessing quality of life had
significant findings. However, most lacked statistical
power and were done late in the disease process, resulting
in difficulties with recruitment, attrition, and co-intervention. None specifically assessed an early palliative care
intervention in patients with cancer.
Since publication of this review,9 results have been
reported from two randomised controlled trials assessing early palliative care interventions in patients with
advanced cancer. The first was a study10 of 322 participants
with advanced cancer and a prognosis of about 1 year;
patients were randomised to routine care or to a palliative
care problem-solving intervention through telephone
contact from advanced practice nurses. In this study, and
in the second, a trial11 of 151 patients with advanced nonsmall-cell lung cancer, patients randomised to the early
palliative care group had better quality of life and mood.
www.thelancet.com Published online February 19, 2014 http://dx.doi.org/10.1016/S0140-6736(13)62416-2
Published Online
February 19, 2014
http://dx.doi.org/10.1016/
S0140-6736(13)62416-2
See Online/Comment
http://dx.doi.org/10.1016/
S0140-6736(13)62676-8
See Online for video
Division of Medical Oncology
and Haematology, Department
of Medicine
(C Zimmermann MD,
M Krzyzanowska MD,
B Hannon MBChB, N Leighl MD,
Prof A Oza MD [Lon],
Prof M Moore MD,
Prof I Tannock MD), and
Department of Psychiatry
(C Zimmermann,
Prof G Rodin MD, C Lo PhD),
University of Toronto, Toronto,
ON, Canada; Department of
Psychosocial Oncology and
Palliative Care (C Zimmermann,
N Swami BSc, B Hannon,
A Rydall MSc, Prof G Rodin, C Lo),
Department of Medical
Oncology (M Krzyzanowska,
N Leighl, Prof A Oza, Prof
M Moore, Prof I Tannock), and
Campbell Family Cancer
Research Institute
(C Zimmermann, Prof G Rodin,
Prof I Tannock), Princess
Margaret Cancer Centre,
University Health Network,
Toronto, ON, Canada; and
Western University, London,
ON, Canada (Prof A Donner PhD)
Correspondence to:
Dr Camilla Zimmermann,
Department of Psychosocial
Oncology and Palliative Care,
Princess Margaret Cancer Centre,
Toronto, ON, Canada
camilla.zimmermann@uhn.ca
1
Articles
However, the former study10 did not assess intervention
by a palliative care team and the latter11 included only
patients with non-small-cell lung cancer.
We report a cluster-randomised controlled trial of early
palliative care in patients with advanced cancer. We opted
for cluster rather than individual randomisation (randomising clinics, rather than individual patients) on the basis
of evidence from the health-services literature9,12,13 and
advice from oncologists that it is difficult to recruit patients
to be individually randomised (or not) to an intervention
such as palliative care, in view of strong preconceived
preferences among patients and their oncologists. The
design implications of cluster-randomised trials were
thoroughly taken into account in our trial.14
We postulated that, compared with standard cancer
care, early intervention (clinical prognosis of 6–24 months)
by a palliative care team would be associated with
improved patient quality of life, symptom control, and
satisfaction with care, and less difficulty with clinician–
patient interactions.
Methods
Study design and participants
The study took place at Princess Margaret Cancer Centre,
a comprehensive cancer centre and part of the University
Health Network in Toronto, ON, Canada, between Dec 1,
2006, and Feb 28, 2011. Recruitment involved daily
screening of participating oncology clinics by research
personnel to establish eligibility. Eligible patients were
aged 18 years or older, had stage IV cancer (for breast or
prostate cancer, refractory to hormonal therapy was an
additional criterion; patients with stage III cancer and
poor clinical prognosis were included at the discretion of
the oncologist); an estimated survival of 6–24 months
(assessed by their main oncologist);15 and Eastern
Cooperative Oncology Group (ECOG) performance
Early palliative care intervention
Standard care
Outpatient clinics
Staff
Palliative care physician and nurse
Oncologist and oncology nurses
Visits
Routine once monthly; more often if necessary
Ad hoc; mainly based on chemotherapy or
radiation schedule
Symptom assessment in clinic
Routine, structured assessment during every visit by
palliative care nurse and physician
No structured assessment
Psychosocial assessment in clinic
Routine assessment and discussion of goals of care, of
patient and family support needs, and of patient and family
coping and psychological distress; discussion of advance care
planning according to patient and family readiness
No routine assessment
Telephone follow-up
Routine by palliative care nurse after each visit; more often as As needed by oncology nurse; rare access to oncologist
needed by palliative care nurse and physician
On-call service
24-h on-call service explained during first visit; provided by
specialised palliative care physicians
Access to 24-h on-call service (oncology resident
or clinical associate)
Inpatient care
Direct access to palliative care unit for symptom management
No access to palliative care unit; admission to
oncology ward or medical ward (via emergency
department for urgent care)
Inpatient staff
Primary care by trained palliative care nurses and physicians
Primary care by oncology nurses and clinical associates
Palliative care training for nurses
Formal 10-day training at opening of palliative care unit,
and continued education by palliative care unit advance
practice nurse
No formal palliative care training
Palliative care inpatient follow-up
Follow-up by palliative care team when admitted to
No follow-up by palliative care team
non-palliative-care-unit service at University Health Network
Hospital service
Home care
Community care access centre services* Explained and offered during first visit; reassessed at each visit
Ad hoc; generally no home care referral until referral
to palliative care team
Communication with family physician
and community care access centre
Routine
Rarely; ad hoc
Home palliative care physician†
Explained during first visit; offered when ECOG performance
status ≥3 or when patient requests
None
Multidisciplinary, addressing physical, psychological, social
and spiritual needs
Ad hoc, mainly addressing physical needs
Approach to care
All care providers
ECOG=Eastern Cooperative Oncology Group. *Community care access centre services include personnel such as nursing, personal support, physical therapy, occupational
therapy, and equipment such as hospital bed, walker, wheelchair. †Home palliative care physicians provide either back-up support to family physicians doing house calls or
direct care if (as is the case for most patients) the family physician does not provide house calls.
Table 1: Comparison of early palliative care intervention and standard care
2
www.thelancet.com Published online February 19, 2014 http://dx.doi.org/10.1016/S0140-6736(13)62416-2
Articles
status of 0, 1, or 2 (assessed by their main oncologist).16
Consenting patients completed baseline measures to be
eligible. Exclusion criteria were insufficient English
literacy to complete questionnaires and inability to pass
the cognitive screening test (Short Orientation-MemoryConcentration Test score <20 or >10 errors).17
Patients provided written, informed consent to participate and the study was approved by the Research Ethics
Board of the University Health Network.
Randomisation and masking
In this cluster-randomised controlled trial, oncology
clinics were the units of randomisation and patients the
units of inference.14 A clinic was defined as a unique,
consistent time and location for outpatient oncology care
by a medical oncologist. Written consent was obtained
from all 16 medical oncologists in the five largest site
groups (lung, gastrointestinal, genitourinary, breast, and
gynaecology) to randomise their 24 clinics either to
immediate consultation and follow-up by a palliative care
team, or to standard care. Randomisation was done by
the statistical team at Western University (London, ON,
Canada) using a computer-generated sequence, was in
a 1:1 ratio, and was stratified by clinic size and tumour
site: lung (four clinics), gastrointestinal (eight clinics),
genitourinary (four clinics), breast (six clinics), and
gynaecological (two clinics). Eight oncologists had clinics
in two tumour sites, and were therefore randomised to
two clusters; five oncologists had their two clinics
randomised within the same trial group and three to
different groups.
Although complete masking of interventions was not
possible, patients provided written informed consent to
participate in their own study group, without being
informed of the existence of another group. This form
of masking is common in cluster randomised trials,18
and avoids potential bias from patients in the control
group requesting the intervention or otherwise altering
their behaviour. Oncologists and investigators were
aware of assignment.
depending on the status of the patient, included:
arrangement of home nursing care services; transfer of
care to a home palliative care physician (when the patient’s
ECOG performance status was 3 or worse, or when
requested); and admission to the Princess Margaret Cancer
Centre palliative care unit for urgent symptom control or
terminal care. This model was assessed in our pilot study,20
which showed improvement in symptom control and
satisfaction of patients. The study duration was 4 months;
on completion of the study, patients in the intervention
group were offered continued follow-up in the oncology
palliative care clinic.
The control group received no formal intervention, but
palliative care referral was not denied, if requested.
Participants in the control group referred to the palliative
care service received the same care as patients in the
intervention group, but did not have the same
standardised monthly follow-up. On completion of
the 4-month measures, which represented the end of the
trial, patients in the control group continued with
standard care and were referred (or not) to the palliative
care team when this referral would normally occur.
Study measures consisted of several scales. The
Functional Assessment of Chronic Illness Therapy—
24 clusters (clinics) randomised
12 clinics allocated to control
1367 patients screened for eligibility
1134 excluded
944 ineligible to participate
99 declined participation
57 not interested
23 time required
9 too ill
10 other reason
91 did not complete
baseline assessment
12 clinics allocated to intervention
1926 patients screened for eligibility
1698 excluded
1357 ineligible to participate
251 declined participation
87 not interested
75 time required
54 no symptoms
14 too ill
7 palliative wording
14 other reason
90 did not complete
baseline assessment
Procedures
Table 1 shows how the intervention differs from standard
care. The Princess Margaret Cancer Centre palliative care
service consists of an outpatient oncology palliative care
clinic, a 12-bed palliative care unit, and an inpatient
consultant team.19 The core intervention was consultation
and follow-up in the oncology palliative care clinic by a
palliative care physician and nurse, consisting of: (1)
comprehensive,
multidisciplinary
assessment
of
symptoms, psychological distress, social support, and
home services, within 1 month of recruitment (60–90 min
duration); (2) routine telephone contact from a palliative
care nurse 1 week after the first consultation, and thereafter
as needed; (3) monthly outpatient palliative care follow-up
(20–50 min); and (4) a 24-h on-call service for telephone
management of urgent issues. Ancillary interventions,
233 assigned to control
17 died
24 withdrew
192 completed at least one follow-up
assessment
9 died
28 withdrew
155 completed study
228 assigned to intervention
15 died
12 withdrew
201 completed at least one follow-up
assessment
29 died
41 withdrew
131 completed study
Figure 1: Trial profile
www.thelancet.com Published online February 19, 2014 http://dx.doi.org/10.1016/S0140-6736(13)62416-2
3
Articles
Intervention
(n=228)
Age (years)
Control
(n=233)
61·2 (12·0)
60·2 (11·3)
Women
136 (59·6%)
125 (53·6%)
Married or common law partner
156 (68·4%)
167 (71·7%)
43 (18·9%)
42 (18·0%)
Living alone
Employment status
Retired
104 (45·6%)
101 (43·3%)
Employed
45 (19·7%)
59 (25·3%)
Unemployed
29 (12·7%)
24 (10·3%)
Disability
50 (21·9%)
49 (21·0%)
Below high school
18 (8·0%)
24 (10·3%)
High school
56 (24·8%)
57 (24·6%)
152 (67·3%)
151 (65·1%)
Lung
55 (24·1%)
46 (19·7%)
Gastrointestinal
74 (32·5%)
65 (27·9%)
Genitourinary
27 (11·8%)
51 (21·9%)
Breast
41 (18·0%)
31 (13·3%)
Gynaecological
31 (13·6%)
40 (17·2%)
Education*
College, university, or other
Tumour site
Active chemotherapy
No
29 (12·7%)
25 (10·7%)
Awaiting chemotherapy
25 (11·0%)
26 (11·2%)
174 (76·3%)
182 (78·1%)
Radiation treatment
Yes
16 (7·0%)
13 (5·6%)
CCI total score >0†
75 (32·9%)
71 (30·5%)
ECOG performance status‡
0
61 (26·8%)
76 (32·6%)
1
149 (65·4%)
143 (61·4%)
2
18 (7·9%)
14 (6·0%)
FACIT-Sp§ (n=443)
101 (20·3)
105 (18·8)
QUAL-E¶ (n=436)
73 (11·1)
74 (11·5)
ESAS|| (n=461)
28 (15·5)
23 (15·7)
FAMCARE-P16** (n=449)
64 (9·7)
68 (9·7)
Score
CARES-MIS†† (n=448)
4
4·7 (5·6)
Spiritual Well-Being (FACIT-Sp) scale measures quality
of life, including physical, social and family, emotional,
functional, and spiritual domains (score range 0–156,
higher scores are better).21,22 The Quality of Life at the End
of Life (QUAL-E) scale measures quality of life in
domains of life completion, effect of symptoms,
relationship with health provider, and preparation for
end of life (range 21–105, higher scores are better).23 The
Edmonton Symptom Assessment System (ESAS)
consists of nine numerical scales with anchors of 0 (best)
and 10 (worst) for pain, fatigue, drowsiness, nausea,
anxiety, depression, appetite, dyspnoea, and wellbeing;
individual symptom scores are summed for the ESAS
Distress
Score,
ranging
from
0–90.24
The
FAMCARE-P16 scale measures satisfaction with
information-giving, availability of care, psychological
care, and physical care in patients with advanced cancer
(range 16–80, higher scores are better).25 The Cancer
Rehabilitation Evaluation System Medical Interaction
Subscale (CARES-MIS) assesses specific problems of
patients in their interactions with nurses and doctors,
including those related to information seeking,
communication, and control of the medical team
(range 0–44, higher scores are worse).26
Follow-up measures were distributed in person or by
mail. Patients who did not respond within 2 weeks
received a reminder telephone call and an offer for
research staff to help with completion of forms. Patients
who could not be reached or declined help were deemed
lost to follow-up. Medical records for each patient were
reviewed at recruitment and at each follow-up visit.
All measures were completed by the patient monthly
for 4 months after enrolment. The 4-month study interval
was chosen on the basis of two considerations: that there
should be sufficient time for improvement in quality of
life (or deterioration in patients in the control group,
which tends to occur in all patients with advancing
3·9 (5·4)
Intervention
group
(n=228)
Control
group
(n=233)
Data are mean (SD) or n (%). CCI=Charlson Comorbidity Index; ECOG=Eastern
Cooperative Oncology Group. FACIT-Sp=Functional Assessment of Chronic Illness
Therapy—Spiritual Well-Being scale. QUAL-E=Quality of Life at the End of Life
scale. ESAS=Edmonton Symptom Assessment System. FAMCARE-P16=FAMCARE
patient satisfaction with care measure. CARES-MIS=Cancer Rehabilitation
Evaluation System Medical Interaction Subscale. *Data missing for two patients
in intervention group and one patient in control group. †The CCI is a measure of
comorbidity for patients with cancer; it generates a weighted score on the basis of
the presence of various medical disorders, each disorder is assigned a score of 1, 2,
3, or 6 based on the CCI scoring index, scores are summed to provide a total score
for each patient.28 ‡An ECOG score of 0 indicates fully active at predisease
performance; 1 ambulatory but restricted in physically strenuous activity; 2 not
fully ambulatory but lying or sitting <50% of the day.16 §The FACIT-Sp scale ranges
from 0 to 156, with higher numbers representing better quality of life. ¶The
QUAL-E scale ranges from 21 to 105, with higher numbers representing better
quality of life. ||The ESAS scale ranges from 0 to 90, with higher numbers
representing worse symptom control. **The FAMCARE-P16 scale ranges from
16 to 80, with higher numbers representing better patient satisfaction with care.
††The CARES-MIS subscale ranges from 0 to 44, with higher numbers
representing greater problems with medical interactions.
Data are n (%). *Numbers for admissions, consultations, and referrals are not
exclusive (ie, one patient might have a consultation and a palliative care unit
admission).
Table 2: Baseline characteristics of trial participants
Table 3: Palliative care intensity for intervention and control groups
Palliative care clinic visits
None
0
213 (91·4%)
1
23 (10·1%)
9 (3·9%)
2
30 (13·2%)
5 (2·1%)
3
28 (12·3%)
3 (1·3%)
4
68 (29·8%)
0
≥5
79 (34·6%)
3 (1·3%)
17 (7·5%)
0
Inpatient palliative care consultations*
18 (7·9%)
2 (0·9%)
Palliative home nursing referrals*
39 (17·1%)
7 (3·0%)
Home palliative care physician referrals*
18 (7·9%)
7 (3·0%)
Palliative care unit admissions*
www.thelancet.com Published online February 19, 2014 http://dx.doi.org/10.1016/S0140-6736(13)62416-2
Articles
illness), and that there should not be excessive dropout
and therefore reduced power. On further consideration
and before trial initiation, 3 months was designated as
the primary endpoint to further ensure sufficient power
for analyses; 4 months (trial end) was designated a
secondary endpoint, because it would show whether or
not an effect was sustained. Change score for FACIT-Sp
from baseline to 3 months was the primary outcome.
Change score for FACIT-Sp at 4 months, and change
scores for all other scales at 3 months and 4 months,
were secondary outcomes. As an exploratory substudy,
we also collected data from caregivers; these data will be
reported separately, as will data regarding economic cost
and qualitative findings.
Statistical analysis
The intracluster correlation coefficient accounts for the
greater similarity of responses of patients within clusters
compared with between clusters.14 Our initial sample size
Intervention
Control
n
n
Mean observed
change from
baseline (SD)
estimation showed that 380 patients (190 per group),
would provide 80% power at the two-sided 5% level of
significance to detect a between-group difference in
FACIT-Sp of 0·45 SD (medium effect size) by the primary
endpoint of 3 months. Sample size was recalculated
in 2008, on the basis of the observed SD (from aggregated
baseline data of 245 patients), intracluster correlation
coefficient, attrition, and adherence. The revised sample
size was 450 patients (225 per group).
Patients were analysed according to their original
randomisation. To account for clustering,14 we used mixed
effects models, implemented with PROC MIXED or
GLIMMIX using SAS software (version 9.3). We
controlled for baseline variables of tumour site, age,
chemotherapy treatment, and ECOG score, on the basis
of their association with quality of life in a previous
(unpublished) analysis of our baseline data. To limit
multiple testing, available cases analyses were focused on
changes observed at the primary (3-month) and secondary
Available cases analysis*
Mean observed
change from
baseline (SD)
Adjusted difference
between change scores
(95% CI)
p value
Effect size†
ICC
FACIT-Sp
1 month
154
1·86 (11·99)
168
–1·34 (10·12)
··
··
··
2 months
138
0·58 (13·09)
151
–2·71 (12·92)
··
··
··
··
3 months
140
1·60 (14·46)
141
–2·00 (13·56)
3·56 (–0·27 to 7·40)
0·07
0·26
0·035
4 months
122
2·46 (15·47)
149
–3·95 (14·21)
6·44 (2·13 to 10·76)
0·006
0·44
0·024
1 month
154
1·09 (6·79)
162
–1·19 (7·22)
··
··
··
2 months
137
1·38 (7·49)
151
–0·61 (8·13)
··
··
··
3 months
139
2·33 (8·27)
139
0·06 (8·29)
2·25 (0·01 to 4·49)
0·05
0·28
0·036
4 months
121
3·04 (8·33)
148
–0·51 (7·62)
3·51 (1·33 to 5·68)
0·003
0·45
0·015
1 month
180
–0·72 (13·01)
172
1·13 (10·79)
··
··
··
2 months
158
0·89 (14·83)
160
1·45 (14·08)
··
··
··
3 months
151
0·14 (16·93)
149
2·12 (13·88)
–1·70 (–5·26 to 1·87)
0·33
–0·13
0·067
4 months
131
–1·34 (15·98)
155
3·23 (13·93)
–4·41 (–8·76 to –0·06)
0·05
–0·31
0·034
1 month
160
1·77 (8·14)
169
–2·64 (7·96)
··
··
··
2 months
140
1·95 (9·12)
157
–2·26 (7·36)
··
··
··
··
3 months
142
2·33 (9·10)
145
–1·75 (8·21)
3·79 (1·74 to 5·85)
0·0003
0·47
<–0·0001
4 months
121
3·70 (8·58)
153
–2·42 (8·33)
6·00 (3·94 to 8·05)
<0·0001
0·73
–0·018
··
··
QUAL-E
··
··
ESAS
··
··
FAMCARE-P16
··
CARES-MIS
1 month
157
–0·45 (4·33)
170
0·88 (3·32)
··
··
··
2 months
144
–0·28 (4·57)
156
0·86 (3·73)
··
··
··
3 months
139
–0·16 (5·50)
147
0·85 (4·06)
–0·66 (–2·25 to 0·94)
0·40
–0·21
0·018
4 months
123
–0·35 (4·38)
154
0·61 (3·60)
–0·84 (–1·91 to 0·22)
0·11
–0·24
0·023
··
FACIT-Sp=Functional Assessment of Chronic Illness Therapy—Spiritual Well-Being scale. QUAL-E=Quality of Life at the End of Life scale. ESAS=Edmonton Symptom Assessment
System. FAMCARE-P16=FAMCARE patient satisfaction with care measure. CARES-MIS=Cancer Rehabilitation Evaluation System Medical Interaction Subscale. ICC=intracluster
correlation coefficient. *Differences in change scores between groups and associated tests of effect were estimated by regression, adjusting for clustering and baseline covariates;
negative ICCs were assumed to arise by chance and therefore treated as zero; regression analyses were done only at the 3-month and 4-month follow-up intervals to limit multiple
testing. †Effect sizes are Cohen’s d: an effect size of 0·20 is small, 0·50 is moderate, and 0·80 is large.29
Table 4: Change scores relative to baseline and the regression-estimated differences in change scores between groups
www.thelancet.com Published online February 19, 2014 http://dx.doi.org/10.1016/S0140-6736(13)62416-2
5
Articles
(4-month) endpoints. Sensitivity analyses were done
using three different methods for handling missing data:
last value carried forward, complete case analysis, and
multiple imputation. For multiple imputation we used
the Markov Chain Monte Carlo method to generate 30 imputations, in line with evidence that this number
would minimise power loss with 35% missing data at
the 3-month primary endpoint.27
This trial is registered with ClinicalTrials.gov, number
NCT01248624.
Mean change
8
6
4
2
0
–2
–4
–6
–8
–10
Mean change
FACIT-Sp
6
5
4
3
2
1
0
–1
–2
–3
QUAL-E
Role of the funding source
The study sponsors provided operating funds specifically
for the trial after a competitive scientific peer-review
process (Canadian Cancer Society), or indirectly via
programme funding (Ontario Ministry of Health and
Long-Term care), but had no role in study design, data
collection, data analysis, data interpretation, or writing
of the report. The Data Safety Monitoring Board at
Princess Margaret Cancer Centre oversaw trial integrity
and safety and had full access to the study data, as did
study investigators (CZ, NS, AD, CL). CZ had final
responsibility for the decision to submit for publication.
ESAS
8
Mean change
6
4
Results
2
0
–2
–4
–6
FAMCARE-P16
6
Mean change
4
2
0
–2
–4
–6
CARES-MIS
2·5
Mean change
2·0
1·5
1·0
0·5
0
–0·5
–1·0
–1·5
Control
3 months
Intervention
Control
Study group
4 months
Intervention
Figure 2: Adjusted mean changes in quality of life, symptom distress, satisfaction with care, and problems
with medical interactions at 3 months and 4 months in the control and intervention groups
Change scores were estimated by regression, adjusting for clustering and baseline covariates. Bars are 95% CIs. For
FACIT-Sp, QUAL-E, and FAMCARE-P16, positive changes are improvements; for ESAS and CARES-MIS, negative
changes are improvements. For FACIT-Sp=Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being
scale. QUAL-E=Quality of Life at the End of Life scale. ESAS=Edmonton Symptom Assessment System. FAMCAREP16=FAMCARE patient satisfaction with care measure. CARES-MIS=Cancer Rehabilitation Evaluation System
Medical Interaction Subscale.
6
992 patients were eligible and 461 completed baseline
measures: 223 patients were in the control group
(mean number of patients per cluster 19·4 [SD 12·1])
and 228 were in the intervention group (mean number of
patients per cluster 19·0 [12·9]). Figure 1 shows reasons
for declining participation. Although presence of symptoms was not an entry criterion, 54 patients eligible for
the intervention group declined to participate on the
basis of lack of symptoms. 301 patients completed
measures at 3 months (152 intervention; 149 control)
and 286 patients completed the 4-month trial
(131 intervention; 155 control).
There were no substantive differences between groups
at baseline for demographic or medical data, except for a
larger number of patients in the control group with
genitourinary cancers than in the intervention group
(table 2). Outcome measure scores tended to be worse in
the intervention group at baseline (table 2). However,
during the trial, there was no difference between groups
in the number of patients receiving chemotherapy (86%
[195 of 228] in the intervention group vs 89% [208 of 233]
in the control group, p=0·36) or radiation (21% [47 of 228]
vs 15% [34 of 233], p=0·14).
All participants in the intervention group had at least
one visit to the oncology palliative care clinic (median
four; range one to nine). The intensity of core and
ancillary interventions is shown in table 3, as is contact of
the control group with the intervention.
The primary outcome of change scores for FACIT-Sp
at 3 months did not differ significantly between groups
(mean change score in intervention group +1·60
[SD 14·46] vs control group –2·00 [13·56], p=0·07). For
www.thelancet.com Published online February 19, 2014 http://dx.doi.org/10.1016/S0140-6736(13)62416-2
Articles
Last value forward
Complete cases
Multiple imputation
Adjusted difference
between change scores
(95% CI)
p value
0·02
3·20 (–0·07 to 6·47)
0·06
0·005
4·07 (0·64 to 7·51)
0·02
3·76 (0·81 to 6·71)
0·02
2·23 (0·26 to 4·19)
0·03
4·65 (1·85 to 7·44)
0·003
2·76 (0·84 to 4·67)
0·005
Adjusted difference
between change scores
(95% CI)
p value
Adjusted difference
between change scores
(95% CI)
3 months
3·89 (–0·02 to 8·01)
0·06
7·19 (1·57 to 12·8)
4 months
4·34 (0·70 to 7·98)
0·02
7·70 (2·64 to 12·8)
3 months
2·35 (0·17 to 4·54)
0·04
4 months
2·75 (0·56 to 4·95)
0·02
p value
FACIT-Sp
QUAL-E
ESAS
3 months
–0·95 (–4·54 to 2·64)
0·59
–3·92 (–7·15 to –0·69)
0·02
–2·39 (–6·18 to 1·41)
0·22
4 months
–2·07 (–6·75 to 2·62)
0·37
–4·05 (–8·40 to 0·30)
0·07
–3·85 (–8·13 to 0·42)
0·08
3 months
4·95 (3·09 to 6·81)
<0·0001
4·70 (2·38 to 7·03)
<0·0001
4·70 (2·87 to 6·54)
<0·0001
4 months
5·59 (3·65 to 7·52)
<0·0001
5·63 (3·15 to 8·12)
<0·0001
6·41 (4·44 to 8·37)
<0·0001
3 months
–1·34 (–2·60 to –0·08)
0·04
–0·88 (–2·71 to 0·95)
0·32
–0·80 (–2·01 to 0·40)
0·19
4 months
–1·38 (–2·19 to –0·58)
0·002
–1·25 (–2·37 to –0·13)
0·03
–1·01 (–1·80 to –0·22)
0·01
FAMCARE-P16
CARES-MIS
Differences in change scores between groups and associated tests of effect were estimated by regression, adjusting for clustering and baseline covariates. Negative
intracluster correlation coefficient were assumed to arise by chance and therefore treated as zero. FACIT-Sp=Functional Assessment of Chronic Illness Therapy—Spiritual WellBeing scale. QUAL-E=Quality of Life at the End of Life scale. ESAS=Edmonton Symptom Assessment System. FAMCARE-P16=FAMCARE patient satisfaction with care
measure. CARES-MIS=Cancer Rehabilitation Evaluation System Medical Interaction Subscale.
Table 5: Sensitivity analyses using three methods for handling missing data
secondary quality of life and symptom distress outcomes
at 3 months, the difference between groups in change
scores for QUAL-E was significant (+2·33 [8·27] vs +0·06
[8·29], p=0·05), but the difference between groups in
change score for ESAS was not significant (+0·14 [16·93]
vs +2·12 [13·88], p=0·33). At the secondary 4-month
endpoint, the differences in change scores were significant
for FACIT-Sp (+2·46 [15·47] vs –3·95 [14·21], p=0·006),
QUAL-E (+3·04 [8·33] vs –0·51 [7·62], p=0·003), and
ESAS (–1·34 [15·98] vs +3·23 [13·93], p=0·05; table 4,
figure 2).
There were significant differences in the secondary
outcome of change scores for satisfaction with their care
at the 3-month endpoint (mean change score for
FAMCARE-P16 +2·33 [SD 9·10] in the intervention group
vs –1·75 [8·21] in the control group, p=0·0003) and at
the 4-month endpoint (+3·70 [8·58] vs –2·42 [8·33],
p<0·0001; table 4, figure 2). There was no significant
difference in change scores for problems with medical
interactions at the 3-month endpoint (mean change score
for CARES-MIS, –0·16 [SD 5·50] in the intervention
group vs 0·85 [4·06] in the control group, p=0·40), or at
the 4-month endpoint (–0·35 [4·38] vs 0·61 [3·60], p=0·11).
Table 5 shows results of the sensitivity analyses. At
the 3-month endpoint, the non-significant difference in
FACIT-Sp between control and intervention groups was
consistent with last value carried forward and multiple
imputation (p=0·06 for both); the complete cases analysis showed a significant difference (p=0·02). Differences
for the QUAL-E (p≤0·04 for all approaches) and
FAMCARE-P16 scales (p<0·0001 for all approaches) were
consistent with the main analyses. At the 4-month endpoint, the significant differences for FACIT-Sp, QUAL-E,
and FAMCARE-P16 were robust across analyses, whereas
the significant difference for the ESAS was not; CARESMIS change scores showed significant differences
between groups across sensitivity analyses at 4 months,
although the main available cases analysis was not
significant.
Discussion
In this trial of 461 patients with advanced cancer, early
referral to a palliative care team did not significantly
improve quality of life (measured by the FACIT-Sp scale)
at 3 months compared with usual cancer care. However,
quality of life according to the QUAL-E scale was significantly improved in the intervention group compared
with the control group at 3 months, as was satisfaction
with their care (FAMCARE-P16). Changes in symptom
severity (ESAS) and problems with medical interactions
(CARES-MIS) were not significantly different between
groups at 3 months. By 4 months, differences between
groups were significant for all scales (including FACITSp) except CARES-MIS, favouring the early palliative
care intervention.
WHO has recommended that all countries implement
comprehensive palliative care programmes to improve
quality of life for patients with cancer or other lifethreatening illness,30 and a growing majority of countries
worldwide have such programmes.3 However, palliative
www.thelancet.com Published online February 19, 2014 http://dx.doi.org/10.1016/S0140-6736(13)62416-2
7
Articles
care services are fully integrated within the health-care
system of only a small proportion of these countries,3 and
referrals are usually late in the disease process5,6 despite
guidelines recommending early involvement.31 This
cluster randomised trial provides promising findings
that support the early involvement of specialised
palliative care in the treatment of patients with a wide
range of advanced cancers (panel).
Quality of life was chosen as the primary outcome for
this trial because it is a central focus of palliative care.2,32 We
used as our primary outcome a cancer-specific measure
encompassing physical, functional, social, and psychological domains, and the spiritual domain, which is of
particular relevance in patients with advanced cancer.33
Additionally, we used a secondary outcome measure that
was developed to assess the effectiveness of interventions
targeting improved care at the end of life.23 Quality of life
in patients with cancer tends to decrease gradually in the
last year of life, with a steep decline in the last 2–3 months.34,35
Panel: Research in context
Systematic review
We did a systematic review of randomised controlled trials to assess the effectiveness of
specialised palliative care,9 using the keyword groupings: “palliative”, “terminal”, or “hospice”
and “quality of life”, “quality of care”, “satisfaction”, “wellbeing”, “economic” or “cost”. We
searched Medline, Ovid Healthstar, CINAHL, Embase, and the Cochrane Register of
Controlled Trials, from inception until Jan 31, 2008. Of 22 studies identified, not all assessed
a clearly defined palliative care team, with other interventions including coordinating
services, nursing interventions, or counselling services. Only four of 13 studies assessing
quality of life had significant findings. However, most had insufficient statistical power, and
took place late in the disease process, resulting in difficulties with recruitment, loss to
follow-up, and co-intervention in patients in the control groups. Only five studies were done
in outpatient clinics, with most done in patients’ homes; none specifically assessed an early
palliative care intervention in patients with cancer.
We updated our search on Aug 30, 2013, using the same keyword groupings. Since our
2008 review,9 two other randomised controlled trials were reported assessing early
palliative care interventions in patients with advanced cancer. Bakitas and colleagues10
randomised 322 participants with advanced cancer and a prognosis of about 1 year to
routine care or to a palliative care problem-solving intervention provided by advanced
practice nurses through telephone contact. Temel and colleagues11 randomised
151 patients with advanced non-small-cell lung cancer to an early palliative care team
intervention or to oncology care alone. In both studies, early palliative care improved
quality of life and mood.
Interpretation
The findings of our cluster-randomised trial of 461 outpatients with a wide range of
advanced cancer diagnoses suggest that early palliative care might be beneficial for
patients with a wide range of advanced solid tumour malignancies. Such a collaborative
approach between oncology and palliative care differs from the traditional view of
palliative care services being used only at the very end of life, and is increasingly endorsed
in oncology guidelines.31 Our findings are consistent with those of previous studies that
have shown a benefit of early palliative care for improving quality of life,10,11 and also show
that this approach improves satisfaction of patients with their care. Taken together, these
studies show the benefits of integrated palliative care services, and provide support for
early involvement of specialised palliative care in tandem with standard oncology care.
8
Despite the worse status of the intervention group at
baseline, which might have placed it at a steeper point
in this decline, the present study showed a non-significant
improvement in quality of life at the primary 3-month
endpoint, and a significant benefit at 4 months. This
finding extends evidence from a previous trial of early
involvement by a palliative care team, which showed
improvement in quality of life for patients with non-smallcell lung cancer.11 The effect sizes for quality of life (FACITSp and QUAL-E) in our study were 0·26–0·28 at 3 months
and 0·44–0·45 at 4 months (table 4), compared with those
of 0·41–0·52 identified after early palliative care for
patients with metastatic non-small-cell lung cancer.11
Satisfaction of patients with their care improved
significantly at both endpoints in the early palliative care
intervention group, whereas it deteriorated in the control
group. The measure used for the present trial was
validated specifically for outpatients with advanced
cancer25 and included recommended domains of
accessibility and coordination of care, symptom
management, communication and education, support
and personalisation of care, and support of decisionmaking.36 This patient-centred outcome is indicative of
quality of care, which is particularly important for patients
with advanced illness, who spend a large proportion of
their time interacting with medical teams.36,37
In addition to measuring overall satisfaction of patients
using the FAMCARE-P16 measure, we explored specific
problems with clinician–patient interactions using
CARES-MIS. Although the two measures contain similar
items, in CARES-MIS they are negatively worded, with
emphasis on the patient rather than on the clinician, (eg,
“Doctor’s attention to your description of symptoms” in
FAMCARE-P16 vs “I have difficulty telling my doctor
about new symptoms” in CARES-MIS). This negative
phrasing and emphasis on the patient might have
contributed to the highly positively skewed baseline
scores for the CARES-MIS, which showed little variability
or room to improve. The small number of patients
acknowledging such difficulties also meant that power
was affected by any data loss.
Much the same as in other studies,9,10 the present trial
showed only a marginal difference in symptom intensity,
and only at the secondary 4-month endpoint. This finding
contrasts with our pilot study, in which symptom scores
improved substantially after the same intervention.20 A
possible reason is that for the present trial, the mean
baseline ESAS distress score was only 28 (of a
possible 90); thus, similar to CARES-MIS, there was little
room for improvement. Inclusion of a minimum level of
symptom distress as an entry criterion should be
considered for future studies, as was done in a successful
trial of a nurse-led intervention for cancer-related
fatigue.38 The difference in effect sizes for quality of life
and symptom severity (table 4) affirms that quality of life
is a broader construct, which can improve despite a
relative lack of change in symptoms.
www.thelancet.com Published online February 19, 2014 http://dx.doi.org/10.1016/S0140-6736(13)62416-2
Articles
In complex interventions it is important to define the
core intervention, required for all patients randomised to
the intervention group, and additional interventions,
which are an extension of the core, but are conditional
upon the situation of the individual patient.39,40 In our
study, the core intervention consisted of outpatient
consultation and follow-up by a palliative care physician
and nurse. Although there are differences in how
palliative care is practised worldwide, physicians and
nurses are usually at the core of hospital palliative care
services, which are mainly directed at inpatients.5,6 Most
palliative care clinics, including ours, began as a half-day
clinic per week, which need not be a large drain on
existing inpatient palliative care resources.4
We did not gather data for the level of proficiency in
palliative care of participating oncologists. However, in a
survey of Canadian oncologists, 50% (285 of 568) had
completed a rotation in palliative care during their
training; and 43% (244 of 568) reported they felt
comfortable providing palliative care.6 Those who had
completed a palliative care rotation during their training
were more likely to refer early to palliative care, suggesting
that this experience might have increased their awareness
of possible benefits of this practice.
Limitations of this trial include that it was done at one
centre and interventions were not masked. Availability
of a well-functioning palliative care team provided a
clinically relevant intervention, but exposure of controls
to the intervention was unavoidable because of oncologists’ familiarity with this team, and because three
oncologists were randomised to both trial groups; this
exposure might have reduced power. Although we
cannot exclude that the benefits of the intervention
were attributable to increased attention in general,
rather than specifically from a palliative care team,
attention to concerns of patients is itself an important
aspect of palliative care.32 There was also selection bias,
which is common in cluster-randomised studies
because of randomisation of clusters before consent of
individuals.18 A larger number of patients declined
participation in the intervention group, including
because of lack of symptoms. Aware of this potential
limitation, we opted for cluster randomisation to
maximise recruitment, and were able to attain our
planned sample size.
Early involvement of specialised palliative care is
becoming a quality standard for patients with cancer31
but it is currently underutilised.5–8 This study suggests
that early palliative care might improve quality of life and
increase satisfaction with their care for patients with a
large range of advanced solid tumour malignancies.
These results provide support for earlier involvement of
palliative care teams for patients with advanced cancer.
Further studies are needed to establish which patients
are most likely to benefit; a cost analysis is underway, and
will establish the economic implications of implementing
this model.
Contributors
CZ contributed to conception and design, analysis and interpretation of
data, drafting of the report, critical revision of the report, and obtaining
funding. NS contributed to data collection, drafting of the report, analysis
and interpretation of data, critical revision of the report, and administrative
support. MK and IT contributed to study conception and design, data
collection, analysis and interpretation of data and critical revision of the
report. BH contributed to analysis and interpretation of data, drafting of
the report, and critical revision of the report. NL, AO, and MM contributed
to study conception and design, data collection, and critical revision of the
report. AR contributed to study conception and design, critical revision of
the report, and administrative support. GR contributed to study conception
and design, analysis and interpretation of data and critical revision of the
report. AD contributed to study conception and design, analysis and
interpretation of data, statistical analysis and critical revision of the report.
CL contributed to analysis and interpretation of data, statistical analysis,
drafting of the report, and critical revision of the report.
Declaration of interests
We declare that we have no competing interests.
Acknowledgments
Results of this study were presented, in part, at the American Society of
Clinical Oncology (ASCO) Annual Meeting; Chicago, IL, USA; June 4,
2012; and at Best of ASCO, 2012. This study was funded by grants 017257
and 020509 from the Canadian Cancer Society (CZ) and by the Ontario
Ministry of Health and Long Term Care. CZ is also supported by the
Rose Family Chair in Supportive Care, Faculty of Medicine, University of
Toronto. The views expressed in the study do not necessarily represent
those of the sponsors. We thank the patients who participated in this
study, and their caregivers. We thank the medical oncologists who
referred patients to this study, and the clinical and administrative staff of
the palliative care team at Princess Margaret Cancer Centre. We thank
Debika Burman, Nanor Kevork, and Ashley Pope (Department of
Psychosocial Oncology and Palliative Care, Princess Margaret Cancer
Centre) for their assistance with preparation of study materials,
recruitment of patients, and data entry and preparation.
References
1
Harrison JD, Young JM, Price MA, Butow PN, Solomon MJ.
What are the unmet supportive care needs of people with cancer?
A systematic review. Support Care Cancer 2009; 17: 1117–28.
2
WHO. WHO definition of palliative care. www.who.int/cancer/
palliative/definition/en (accessed Nov 13, 2013).
3
Lynch T, Clark D, Connor SR. Mapping levels of palliative care
development: a global update 2011. London: Worldwide Palliative
Care Alliance, 2011.
4
Meier DE, Beresford L. Outpatient clinics are a new frontier for
palliative care. J Palliat Med 2008; 11: 823–28.
5
Hui D, Elsayem A, De la Cruz M, et al. Availability and integration
of palliative care at US cancer centers. JAMA 2010; 303: 1054–61.
6
Wentlandt K, Krzyzanowska MK, Swami N, Rodin GM, Le LW,
Zimmermann C. Referral practices of oncologists to specialized
palliative care. J Clin Oncol 2012; 30: 4380–86.
7
Lamont EB, Christakis NA. Physician factors in the timing of
cancer patient referral to hospice palliative care. Cancer 2002;
94: 2733–37.
8
Osta BE, Palmer JL, Paraskevopoulos T, et al. Interval between first
palliative care consult and death in patients diagnosed with
advanced cancer at a comprehensive cancer center. J Palliat Med
2008; 11: 51–57.
9
Zimmermann C, Riechelmann R, Krzyzanowska M, Rodin G,
Tannock I. Effectiveness of specialized palliative care: a systematic
review. JAMA 2008; 299: 1698–709.
10 Bakitas M, Lyons KD, Hegel MT, et al. Effects of a palliative care
intervention on clinical outcomes in patients with advanced cancer:
the Project ENABLE II randomized controlled trial. JAMA 2009;
302: 741–49.
11 Temel JS, Greer JA, Muzikansky A, et al. Early palliative care for
patients with metastatic non-small-cell lung cancer. N Engl J Med
2010; 363: 733–42.
12 Fairhurst K, Dowrick C. Problems with recruitment in a
randomized controlled trial of counselling in general practice:
causes and implications. J Health Serv Res Policy 1996; 1: 77–80.
www.thelancet.com Published online February 19, 2014 http://dx.doi.org/10.1016/S0140-6736(13)62416-2
9
Articles
13
14
15
16
17
18
19
20
21
22
23
24
25
26
10
Gilbody S, Whitty P. Improving the delivery and organisation of
mental health services: beyond the conventional randomised
controlled trial. Br J Psychiatry 2002; 180: 13–18.
Donner A, Klar N. Design and analysis of cluster randomization
trials in health research. London: Arnold, 2000.
Glare P, Virik K, Jones M, et al. A systematic review of physicians’
survival predictions in terminally ill cancer patients. BMJ 2003;
327: 195.
Oken MM, Creech RH, Tormey DC, et al. Toxicity and response
criteria of the Eastern Cooperative Oncology Group.
Am J Clin Oncol 1982; 5: 649–55.
Katzman R, Brown T, Fuld P, Peck A, Schechter R, Schimmel H.
Validation of a short Orientation-Memory-Concentration Test of
cognitive impairment. Am J Psychiatry 1983; 140: 734–39.
Campbell MK, Piaggio G, Elbourne DR, Altman DG. Consort 2010
statement: extension to cluster randomised trials. BMJ 2012;
345: e5661.
Zimmermann C, Seccareccia D, Clarke A, Warr D, Rodin G.
Bringing palliative care to a Canadian cancer center: the palliative
care program at Princess Margaret Hospital. Support Care Cancer
2006; 14: 982–87.
Follwell M, Burman D, Le LW, et al. Phase II study of an outpatient
palliative care intervention in patients with metastatic cancer.
J Clin Oncol 2009; 27: 206–13.
Webster K, Cella D, Yost K. The Functional Assessment of Chronic
Illness Therapy (FACIT) measurement system: properties,
applications, and interpretation. Health Qual Life Outcomes 2003;
1: 79.
Peterman AH, Fitchett G, Brady MJ, Hernandez L, Cella D.
Measuring spiritual well-being in people with cancer: the functional
assessment of chronic illness therapy—Spiritual Well-being Scale
(FACIT-Sp). Ann Behav Med 2002; 24: 49–58.
Steinhauser KE, Clipp EC, Bosworth HB, et al. Measuring quality of
life at the end of life: validation of the QUAL-E. Palliat Support Care
2004; 2: 3–14.
Bruera E, Kuehn N, Miller MJ, Selmser P, Macmillan K. The
Edmonton Symptom Assessment System (ESAS): a simple method
for the assessment of palliative care patients. J Palliat Care 1991;
7: 6–9.
Lo C, Burman D, Rodin G, Zimmermann C. Measuring patient
satisfaction in oncology palliative care: psychometric properties of
the FAMCARE-patient scale. Qual Life Res 2009; 18: 747–52.
Schag CA, Heinrich RL, Aadland RL, Ganz PA. Assessing problems
of cancer patients: psychometric properties of the cancer inventory
of problem situations. Health Psychol 1990; 9: 83–102.
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Graham JW, Olchowski AE, Gilreath TD. How many imputations
are really needed? Some practical clarifications of multiple
imputation theory. Prev Sci 2007; 8: 206–13.
Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of
classifying prognostic comorbidity in longitudinal studies:
development and validation. J Chronic Dis 1987; 40: 373–83.
Cohen J. Statistical power analysis for the behavioral sciences,
2nd edn. New York: Taylor & Francis Group, 1988.
Sepulveda C, Marlin A, Yoshida T, Ullrich A. Palliative care:
the World Health Organization’s global perspective.
J Pain Symptom Manage 2002; 24: 91–96.
Smith TJ, Temin S, Alesi ER, et al. American Society of Clinical
Oncology provisional clinical opinion: the integration of palliative
care into standard oncology care. J Clin Oncol 2012; 30: 880–87.
Morrison RS, Meier DE. Clinical practice. Palliative care.
N Engl J Med 2004; 350: 2582–90.
Whitford HS, Olver IN, Peterson MJ. Spirituality as a core domain
in the assessment of quality of life in oncology. Psychooncology 2008;
17: 1121–28.
Giesinger JM, Wintner LM, Oberguggenberger AS, et al. Quality of
life trajectory in patients with advanced cancer during the last year
of life. J Palliat Med 2011; 14: 904–12.
Hwang SS, Chang VT, Fairclough DL, Cogswell J, Kasimis B.
Longitudinal quality of life in advanced cancer patients: pilot study
results from a VA medical cancer center. J Pain Symptom Manage
2003; 25: 225–35.
Dy SM, Shugarman LR, Lorenz KA, Mularski RA, Lynn J.
A systematic review of satisfaction with care at the end of life.
J Am Geriatr Soc 2008; 56: 124–29.
Morrison RS, Siu AL, Leipzig RM, Cassel CK, Meier DE. The hard
task of improving the quality of care at the end of life.
Arch Intern Med 2000; 160: 743–47.
de Raaf PJ, de Klerk C, Timman R, Busschbach JJ,
Oldenmenger WH, van der Rijt CC. Systematic monitoring and
treatment of physical symptoms to alleviate fatigue in patients with
advanced cancer: a randomized controlled trial. J Clin Oncol 2013;
31: 716–23.
Craig P, Dieppe P, Macintyre S, Michie S, Nazareth I, Petticrew M.
Developing and evaluating complex interventions: the new Medical
Research Council guidance. BMJ 2008; 337: a1655.
Rinck GC, van den Bos GA, Kleijnen J, de Haes HJ, Schade E,
Veenhof CH. Methodologic issues in effectiveness research on
palliative cancer care: a systematic review. J Clin Oncol 1997;
15: 1697–707.
www.thelancet.com Published online February 19, 2014 http://dx.doi.org/10.1016/S0140-6736(13)62416-2