374 Journal of Pain and Symptom Management

Vol. 20 No. 5 November 2000

Review Article

Complementary and Alternative Medicine in the

Management of Pain, Dyspnea, and Nausea and
Vomiting Near the End of Life: A Systematic
Review
Cynthia X. Pan, MD, R. Sean Morrison, MD, Jose Ness, MD, Adriane Fugh-Berman,
MD, and Rosanne M. Leipzig, MD, PhD
Department of Geriatrics and Adult Development (C.X.P., R.S.M., J.N., R.M.L.), The Mount
Sinai School of Medicine, New York, New York, and Department of Health Care Sciences (A.F.B.), George Washington University School of Medicine, Washington, DC, Geriatric Research
Education and Clinical Center (R.M.L.), Bronx, New York, USA

Abstract
To review the evidence for efficacy of complementary and alternative medicine (CAM) modalities
in treating pain, dyspnea, and nausea and vomiting in patients near the end of life, original
articles were evaluated following a search through MEDLINE, CancerLIT, AIDSLINE, PsycLIT,
CINAHL, and Social Work Abstracts databases. Search terms included alternative medicine,
palliative care, pain, dyspnea, and nausea. Two independent reviewers extracted data, including
study design, subjects, sample size, age, response rate, CAM modality, and outcomes. The efficacy
of a CAM modality was evaluated in 21 studies of symptomatic adult patients with incurable
conditions. Of these, only 12 were directly accessed via literature searching. Eleven were
randomized controlled trials, two were non-randomized controlled trials, and eight were case
series. Acupuncture, transcutaneous electrical nerve stimulation, supportive group therapy, selfhypnosis, and massage therapy may provide pain relief in cancer pain or in dying patients.
Relaxation/imagery can improve oral mucositis pain. Patients with severe chronic obstructive
pulmonary disease may benefit from the use of acupuncture, acupressure, and muscle relaxation
with breathing retraining to relieve dyspnea. Because of publication bias, trials on CAM
modalities may not be found on routine literature searches. Despite the paucity of controlled trials,
there are data to support the use of some CAM modalities in terminally ill patients. This review
generated evidence-based recommendations and identified areas for future research. J Pain
Symptom Manage 2000;20:374387.

U.S. Cancer Pain Relief Committee, 2000.

Key Words
Alternative medicine, pain, dyspnea, end of life, systematic review, palliative care

Introduction
Address reprint requests to: Cynthia X. Pan, MD,
Depart-ment of Geriatrics and Adult Development, Box
1070, Mount Sinai School of Medicine, One Gustave L.
Levy Place, New York, NY 10029-6574, USA.
Accepted for publication: December 13, 1999.
U.S. Cancer Pain Relief Committee, 2000

Complementary and alternative medicine

(CAM) and palliative care are two major social
movements that have gathered momentum and
have captured the attention of both the public and
the medical profession over the
0885-3924/00/$see front matter

Published by Elsevier, New York, New York

PII S0885-3924(00)00190-1

Vol. 20 No. 5 November 2000

Use of Complementary Medicine in Palliative Care

375

We have chosen to focus on this constellation of

symptoms because (a) they are the most
past decade. Both fields have evolved with little
formal connection despite common features, such
as attracting chronically ill patients for whom
there are no curative conventional ther-apies and
emphasizing the relief of symptoms and
improving function. Both also owe a major portion
of their current visibility to health care consumers
who are increasingly vocal about perceived
discrepancies between patients and physicians
treatment goals.
CAM can be defined as a group of medical
practices that are not in conformity with the
standards of the medical community. These
practices are not usually taught at U.S. medical
schools, are not available in hospitals, and are not
commonly reimbursable by third-party pay-ers.
They are diverse and include a variety of unrelated
systems of care that may or may not conform to
biomedical explanatory models (Fig. 1). Most
people who use CAM do so for chronic, incurable,
nonlife-threatening con-ditions,1 although its use
is also common in pa-tients with cancer, human
immunodeficiency virus (HIV), and acquired
25
immune deficiency syndrome (AIDS).
Palliative care has been defined as the ac-tive
total care of patients whose disease is not
responsive to curative treatment. Control of pain,
of other symptoms, and of psychological, social,
and spiritual problems, is paramount. The goal of
palliative care is achievement of the best quality of
6
life for patients and their families.
Dying patients experience a heavy symptom
burden. In one survey, 84% of cancer patients near
death reported severe pain, 49% had dif-ficulty
7
breathing, and 33% had nausea. Treat-ment of
these symptoms may be associated with
troublesome side effects (e.g., opioid-induced
constipation, nausea, delirium, or se-dation). Some
CAM therapies may have fewer adverse effects
than traditional treatments or may offer therapies
that are more consistent with patients cultural or
health care beliefs. If effective, CAM therapies
may serve as useful al-ternatives or adjuncts in the
care of terminally ill patients.
This systematic review focuses on the role of
CAM therapies in the palliation of three of the
most common and distressing symptoms in ter7
minally ill patients : pain, dyspnea, and nausea.

common and morbid; (b) we were unable to find

studies through a previous MEDLINE search that
adequately addressed other symp-toms, such as
anorexia, fatigue, insomnia, or sleep disorders;
and (c) we recognize that spiri-tual, emotional,
social, and economic aspects require abundant
attention in the care of pa-tients near the end of
life, but they are beyond the scope of this review.

Methods
Data Sources and Study Selection
Six bibliographic databases (MEDLINE, Cumulative Index to Nursing and Allied Health
Literature [CINAHL], CancerLIT, AIDSLINE,
Social Work Abstracts, and PsycLIT) were
searched for original clinical reports or reviews
that evaluated the use of a CAM modality to treat
pain, dyspnea, and nausea and vomiting in adult
patients with incurable conditions who were near
the end of life. Articles were screened and
excluded based on title and ab-stract information
if they involved primarily pa-tients with chronic
conditions that were not fa-tal or not characteristic
of most dying patients (degenerative joint diseases
and arthritides, burns, chronic pain syndromes,
post-operative pain, spinal cord, or other
neurological inju-ries); were laboratory studies,
case
reports,
an-ecdotes,
surveys,
or
commentaries; or focused primarily on biological
mechanisms, risk fac-tors, predictors, prognosis,
or central nervous system stimulation techniques.
When informa-tion was not available or unclear in
the title or abstract, the full text was obtained for
review. Language was not an exclusion criterion.
Table 1 contains specific details of the differ-ent
searches and results. PubMed was used to search
MEDLINE (1966September 1998), us-ing the
exploded medical subject heading (MeSH)
headings alternative medicine (Fig. 1) and
(palliative care or terminal care or hospice care or
death or dying or mortality) and (pain or dysp-nea
or nausea or vomit*). Alternative medi-cine did
not become a MeSH heading until 1986. However,
articles pertaining to alterna-tive medicine have
been retrospectively in-dexed to the heading,
Alternative Medicine. This search yielded 179
articles, 72 of which met the initial inclusion
criteria; 107 articles were excluded for the reasons
above. MED-

376

Pan et al.

Vol. 20 No. 5 November 2000

Fig. 1. MEDLINE Medical Subject Heading (MeSH) Term Working Group, Office of Alternative Medicine, Na-tional
Institutes of Health, Definition of Alternative Medicine.

LINE was also searched, using the exploded MeSH

headings of alternative medicine and (pal-liative
care or terminal care or hospice care or death or
dying or mortality) and (neoplasm or CHF or

COPD or AIDS). These latter search terms rep-

resent common and potentially fatal condi-tions

observed in dying patients. Congestive heart
failure (CHF) and chronic obstructive pulmonary
disease (COPD) were not MeSH headings and
were searched as text terms. Of

Database
MEDLINE, via PubMed
(19669/1998)

CancerLIT database (199398)

Cumulative Index to Nursing and

Allied Health Literature
(CINAHL; 198212/97)
AIDSLINE (1980Dec 1998)

PsycLIT (196798)
PsycLIT (196798)

Social work abstracts

(19779/97)
References and/or experts
Total No. articles

Exploded MeSH headings alternative

medicine (Fig. 1) and (palliative care or
terminal care or hospice care or death or
dying or mortality) and (pain or dyspnea
or nausea or vomit*).
Exploded MeSH headings alternative
medicine and (palliative care or terminal
care or hospice care or death or dying or
mortality) and text terms (neoplasm or
CHF or COPD or AIDS).
Exploded MeSH headings alternative
medicine and (palliative care or terminal
care or hospice care or death or dying or
mortality).
Exploded MeSH headings alternative
medicine and (palliative care or terminal
care or hospice care or death or dying or
mortality).
Exploded MeSH headings alternative
medicine and (palliative care or terminal
care or hospice care or death or dying or
mortality).
Alternative medicine and (palliative care or
terminal care or hospice care or death or
dying or mortality).
(Palliative care or terminal care or hospice
care or death or dying or mortality) and
journal articles with the individual
therapeutic modalities of anthroposophy,
aromatherapy, biofeedback, color
therapy, diet fads, eclecticism, hypnosis,
imagery, mind-body, mental healing,
music therapy, psychotherapy,
therapeutic touch, relaxation, and
massage therapy.
Alternative medicine and palliative care.

179

No. Citations
Initially
Excluded

No. Articles
Reviewed by
Authors

No. Articles
in Final
Analysis

107

72

138

20

20

23

17

225

216

16
619

4
512

12
107

12
21

143 (overlaps with

first MEDLINE
search)

Use of Complementary Medicine in Palliative Care

MEDLINE (19969/1998)

Search Terms

Total No.
Citations
Retrieved

Vol. 20 No. 5 November 2000

Table 1
Summary of Bibliographic Database Search Strategy and Results

37

378

Pan et al.

Vol. 20 No. 5 November 2000

were reviewed. Study design, inclusion/exclusion

criteria, setthe 143 articles retrieved, five met the initial inclusion criteria. There was considerable over-lap
in the results of these two searches.
The terms alternative medicine and (palliative
care or terminal care or hospice care or death or
dy-ing or mortality) were used to search the
follow-ing databases: CancerLIT database (1993
1998), CINAHL (1982December 1997), AIDSLINE database (1980December 1998), Social
Work Abstracts (1977September 1997), and
PsycLIT database (19671998).
Because PsycLIT does not use MeSH headings, we searched (palliative care or terminal care
or hospice care or death or dying or mortality) and
journal articles with the individual therapeutic
modalities of anthroposophy, aromatherapy, biofeedback, color therapy, diet fads, eclecticism,
hypno-sis, imagery, mind-body, mental healing,
music ther-apy, psychotherapy, therapeutic touch,
relaxation, and massage therapy. By using these
terms, 225 articles were retrieved and nine met
criteria. An additional 16 studies were identified
from the reference lists of retrieved papers and
liter-ature reviews and from consultation with experts. No unpublished studies or abstracts were
included.

Data Extraction
The studies that met initial inclusion criteria
were fully assessed and reviewed indepen-dently
by two authors (C.X.P. and J.N.). The authors
agreed 98% of the time regarding the inclusion
and exclusion criteria for articles, as well as
quality of the evidence. Disagreements were
resolved by discussion, by mediation (by R.M.L.
or R.S.M.), and by consensus.
Of the 107 articles that met initial inclusion
criteria, 86 were excluded for the following
reasons: study patients were not near the end of
life or data could not be extrapolated to pal-liative
care patients, study did not involve CAM
therapies, patients were not adults or did not have
the symptoms pertinent to this review, or study did
8
not meet best evidence criteria.
8

We used a best evidence approach to iden-tify

studies for final inclusion in the paper. This meant
that systematic reviews, including meta-analyses,
randomized controlled trials (RCTs), and other
controlled studies were preferentially considered.
If no such trials were found, studies with weaker
designs (such as large prospective case series)

ting, sample size, CAM modality, specifics of the

CAM treatment (dose, frequency, method),
outcomes assessed, and methods of assessment
were extracted from each study and tabulated.

Results
The efficacy of a CAM modality was evalu-ated
in 21 studies of symptomatic adult pa-tients with
incurable conditions who were in the later stages
of illness. Table 2 lists the defi-nitions of CAM
modalities discussed in this re-view. Many of the
articles (12 of 21) were not identified by extensive
search methods but by reference in the searched
articles or by ex-perts.
Of the 21 studies, there were 11 RCTs, two nonrandomized controlled trials, and eight case series
(Table 3) Fourteen studies ad-dressed pain, six
evaluated dyspnea, and one was on nausea and
vomiting. The majority of articles evaluated CAM
in cancer patients, most often to treat pain.

Pain
Pain is highly prevalent near the end of life. The
study to understand prognoses and prefer-ences
for outcomes and risks of treatments (SUPPORT),
a multicenter trial involving more than 4,000
seriously ill hospitalized pa-tients with diverse
diagnoses, documented that 50% of patients who
died in the hospital had moderate to severe pain at
least half the time in the several days before
9
death. CAM thera-pies might serve as useful
adjuvants to tradi-tional analgesic therapy and
may be ideal in patients who cannot tolerate or
may be reluc-tant to take pain medications. We
identified six CAM therapies directed at treatment
of pain.
TENS (transcutaneous electrical nerve stimulation). One double-blind RCT randomized 15
hospice cancer patients (aged 3874 years) to
receive either TENS, sham-TENS, or no additional treatment in addition to standard thera-pies
for pain and emesis. Each patient received daily
sessions for 6 days. Symptoms were as-sessed
using the European Organization for the Research
and Treatment of Cancer Qual-ity-of-Life
Questionnaire (EORTC QLQ-C30). Although the
study lacked the power to detect any effect on
pain, fatigue was improved 8-fold in TENS
patients compared to sham-TENS,

Vol. 20 No. 5 November 2000

Use of Complementary Medicine in Palliative Care

379

Table 2
Definitions of Modalities Discussed in Review34a
CAM Modality
TENS

Acupuncture

Massage therapy
Aromatherapy

Psychotherapy

Behavior therapy
Hypnosis
Imagery

Cognitive coping
strategies

35

Relaxation techniques
Music therapy

Definition
Electrical stimulation of nerve and/or muscles to relieve pain; it is used less frequently to produce
anesthesia. The optimal placements of electrodes or trigger points may correspond with
acupuncture analgesia points. TENS is sometimes referred to as acupuncture-like when using a lowfrequency stimulus.
The practice of piercing specific peripheral nerves with needles to relieve the discomfort associated
with painful disorders, to induce surgical anesthesia and for therapeutic purposes. This procedure
was originally introduced and practiced in China. Electroacupuncture is a form of acupuncture
using low-frequency electrically stimulated needles to produce analgesia and anesthesia and to treat
disease. Acupressure uses similar principals but without needles; it can be administered with local
pressure with fingers or pressure bands.
Group of systematic and scientific manipulations of body tissues best performed with the hands for the
purpose of affecting the nervous and muscular system and the general circulation.
The use of fragrances and essences from plants to affect or alter a persons mood or behavior and to
facilitate physical, mental, and emotional well-being. The chemicals comprising essential oils in
plants has a host of therapeutic properties and has been used historically in Africa, Asia, and India.
It is often used in conjunction with massage.
A form of therapy in which two or more patients participate under the guidance of one or more
psychotherapists for the purpose of treating emotional disturbances, social maladjustments, and
psychotic states. A generic term for the treatment of illnesses primarily by verbal or nonverbal
communication. Includes behavior therapy, art therapy, hypnosis, imagery, music therapy, group
therapy.
The application of modern theories of learning and conditioning in the treatment of behavior
disorders. Techniques include cognitive therapy, biofeedback, relaxation techniques, and
meditation.
A state of increased receptivity to suggestion and direction, initially induced by the influence of
another person.
The use of mental images produced by the imagination (a form of psychotherapy). It can be classified
by the modality of its content: visual, verbal, auditory, olfactory, tactile, gustatory, or kinesthetic.
Common themes derive from nature imagery (e.g., forests and mountains), water imagery (e.g.,
brooks and oceans), travel imagery, etc. Imagery is often used to help patients cope with other
diseases. Imagery often forms a part of hypnosis, of relaxation techniques, and of behavior therapy.
Attempt to alter patterns of negative thoughts and dysfunctional attitudes to foster more healthy and
adaptive thoughts, emotions, and actions. Major classes include external focus of attention, neutral
imaginings, pleasant imaginings, dramatized coping, rhythmic cognitive activity, and pain
acknowledging. Emphasizes the basic components: education, skills acquisition, cognitive and
behavioral rehearsal, and generalization and maintenance.
The use of muscular relaxation techniques and activities in treatment to reduce feelings of tension.
Often used in conjunction with imagery techniques.
The use of music as an adjunctive therapy in the treatment of neurologic, mental, or behavioral
disorders.

TENS _ transcutaneous electrical nerve stimulation.

a
All definitions come from reference 34, except for cognitive coping strategies.

and 16-fold compared to standard controls. The

overall quality of life also improved com-pared
10
with either control. In a prospective pre/postintervention study of 60 patients (aged 3575
years) with intractable cancer pain, 2 weeks of
TENS resulted in improve-ment rated as excellent
by 28%, fair by 36%, and no response by 35%;
after 3 months 15% had excellent responses, 18%
11
fair, and 67% were considered failures. A case
series of TENS use in nine patients (aged 2074
years) with advanced cancer found immediate pain
relief and reduced analgesic requirements in 66%,
with partial relief in another 22%. The duration of
12
analgesic effect was not reported. Another case
series of 29 frail cancer patients

with severe pain (mean age 55 years; range 18

82) evaluated the effect of acupuncture and
electrical stimulation administered in 30- to 4513
minute sessions spaced per patient needs. Pain
relief or reduction accompanied by discontinuation of injection analgesics was seen in 62% of
patients, and pain reduction with decreased need
for injections was seen in 27%.
Acupuncture. Acupuncture as a treatment for pain
in seriously ill cancer patients has been evaluated
only in uncontrolled studies. When 92 patients
(aged 1970 years) with abdominal pain due to
locally invasive or metastatic can-cer were treated
daily for 1 to 2 weeks at the acupuncture point of
ST 36, pain control was

380

Pan et al.

Vol. 20 No. 5 November 2000

Table 3
Designs and Sample Sizes for Studies of Complementary and Alternative Medicine Modalities for the Symptoms
of Patients at the End of Life
Symptom
Pain

CAM Modality
TENS

RCT

Nonrandomized
Controlled Study Case Series

10

N _ 15

11

N _ 60 _
12

N_9

Acupuncture

N _ 29

13

N _ 92

14

15

N _ 183
16

N _ 239

[HIV]
Massage

N _ 28

17
18

N_9

Massage _
aromatherapy
Psychologic therapies
(behavioral, CBT,
relaxation, imagery)

Music therapy
Hypnosis

19

N _ 103

20

N _ 94

22

N_9
N _ 58
(_ support
group)

Dyspnea

Acupuncture

N _ 58

21

N _ 24

23

24

(COPD)
N _ 20

25

(Cancer)
Acupressure

N _ 31

26

(COPD)

Nausea/
vomiting

27

PMR/breathing
retraining

N _ 20

Psychoanalysis

N _ 65
(COPD)

Breathing/coping/
counseling

N _ 20

Acupressure

(COPD)
28

29

(cancer)
30

N_9 _

Results
Underpowered for pain outcomes
Excellent pain relief in 28% of
patients, fair in 36%
Immediate pain relief and
reduced analgesics intake in
66%, partial relief in 22%
62% had pain relief/reduction
and discontinuation of injection
analgesics
Almost all patients with mild/
moderate pain and 72% of those
with severe pain were controlled
for one month.
48% reported _3 days of pain
relief or improved mobility.
No significant difference between
real and sham acupuncture for
neuropathic pain.
Men had immediate, short term
pain relief (P _ 0.01)
Significant pain reduction on days
1 and 2 (P _ 0.025)
Pain relief in 33% of patients
Pain reduced in 38% of patients.
For oral mucositis pain,
relaxation/imagery and
cognitive coping skills both
improved pain (P _ 0.01)
compared to usual treatment
___ therapist support.
No significant differences
Pain sensation and suffering were
reduced in breast cancer patients
receiving support group therapy
(P _ 0.05); self-hypnosis
enhanced pain relief (P _ 0.05).
Acupuncture decreased
breathlessness (P _ 0.02) and
improved 6-minute walk.
70% of patients had marked
improvement in breathlessness
and anxiety (P _ 0.001).
Real acupressure decreased
dyspnea compared to sham
(P _ 0.01).
PMR decreased dyspnea
compared to controls (P _ 0.04).
Nurse therapy reduced dyspnea
compared to psychoanalytic
therapy (P _ 0.05).
Interventions reduced
breathlessness at worst by 35%
(P _ 0.05) and distress caused by
breathlessness by 53% (P _ 0.01).
No significant differences

RCT _ randomized controlled trial; TENS _ transcutaneous electrical nerve stimulation; PMR _ progressive muscle relaxation; CBT _ cognitive behavioral
therapy; COPD _ chronic obstructive pulmonary disease; N _ Number of subjects; _ _ pre/postintervention, single group study; _ _ crossover or N-of-1
design.

Vol. 20 No. 5 November 2000

Use of Complementary Medicine in Palliative Care

381

achieved for at least 1 month in essentially all patients with mild or moderate pain and 72% with
14
severe pain. Pain was assessed by using the
World Health Organization grading system of
mild, moderate, or severe. Another case series
examined 183 patients from a cancer pain clinic
who were unresponsive to conventional therapy,
146 of whom had pain directly related to cancer or
as a result of treatment. Patients were treated with
manual acupuncture in 5- to 15-minute ses-sions
one to four times weekly. Forty-eight per-cent (70
patients) reported more than 3 days of pain relief
and/or an increase in mobility. The duration of
response was greater for patients with malignant
muscle spasm pain, post-irradia-tion fibrosis, and
15
ischemic limb pain.
In HIV-positive patients (mean age 40 years), a
large multicenter, single-blind RCT of 239 patients
compared real versus sham acu-puncture for the
treatment of painful periph-eral neuropathy.
Acupuncture was adminis-tered twice weekly for
6 weeks, followed by once weekly for 8 weeks.
No statistically signifi-cant difference was found
between the two groups at 6 or 14 weeks of
16
follow-up.

Behavioral and Relaxation Therapies. These approaches are often used to treat pain, but few
studies have been done in dying patients. A case
series of 58 patients with advanced cancer referred
to relaxation therapy in hospice found that pain
20
was reduced in 38% of pa-tients. However, the
assessment of pain was vague and the regimen of
relaxation therapy was individualized for each
patient. In an RCT of 94 bone marrow transplant
patients (mean age 36 years) with painful oral
mucositis, four therapies were compared: usual
treatment (UT), UT _ therapist support, UT _
relax-ation/imagery training, and UT _ a package
of cognitive behavior coping skills including
21
relaxation/imagery.
These interventions consisted of 30-minute sessions twice weekly for 5
weeks. Patients who received relaxation/imag-ery
or the package skills reported less pain (on a VAS
0100 mm) than patients in the other two groups
(P _ 0.01); however, no greater pain relief was
obtained by adding cognitive-behavioral skills to
relaxation/imagery. There was no significant
difference in analgesic in-take among the groups.

Massage. Two studies evaluated massage for pain

in palliative care patients. In an un-blinded RCT,
28 patients (mean age 61.5 years) with cancer
were assigned to either Swedish massage therapy
or a visitor for 10 minutes. Pain was assessed by a
visual analog scale (VAS) from 0 to 10 cm. Men
experi-enced immediate pain relief (VAS from 4.2
to 2.9, P _ 0.01), but this effect subsided by an
hour after the massage. There was no signifi-cant
benefit in women, although their base-line level of
17
pain was mild (1.7/10). In a case series, 9 male
cancer patients (mean age 56.6 years) who
received two consecutive 30-minute evening
massages reported significant reductions in pain
(according to VAS 0100 mm) as compared to
18
baseline. There was a mean reduction in pain of
29.5 mm on day 1 (P _ 0.007) and 22.6 mm on
day 2 (P _ 0.025). There was also a reduction in
anxiety and enhanced feelings of relaxation. In a
case series involving 103 patients with cancer, a
combination of massage and aromatherapy
promoted pain relief in 33% of patients who
concluded the study (47%). There was no
quantitative assessment of the extent of pain
19
relief.

Music Therapy. In a randomized, placebo-controlled, unblinded, multiple crossover trial, 9

terminally ill cancer patients were exposed to
three interventions: UT, UT _ background sound,
and UT _ music therapy. The expo-sures happened
for 15 minutes, twice daily, for 2 days. There was
no statistically significant dif-ference among the
three groups (P _ 0.07) when pain relief was
assessed, using a graphic rating scale (010 cm).
This study was under-powered to detect small
differences in the ef-fect of music therapy on
22
palliation of pain.
Psychological Therapies. In an unblinded RCT of
58 women (mean age 54.5 years) with ad-vanced
breast cancer, patients were assigned to standard
treatment or to standard treatment plus group
therapy that consisted of weekly meetings with a
support group. The latter group was further
randomized into no-hypno-sis and self-hypnosis
arms to manage pain. Pain was assessed by a
rating scale from 0 to 10. The patients receiving
group therapy experienced a statistically
significant reduction in pain sen-sation and pain
suffering (both P _ 0.01) over 10 months of

follow-up, but there was no dif-ference in

frequency and duration of pain epi-

382

Pan et al.

sodes. Self-hypnosis provided a further reduc-tion

23
in pain sensation (P _ 0.05). However, this
patient population differed from the usual
palliative care population in that they had a longer
survival period.
No well-designed studies were identified on the
effect of biofeedback, hypnosis alone, or
aromatherapy in alleviating pain in patients at the
end of life.
In summary, relaxation with imagery can improve pain in patients with oral mucositis, a
condition that can be found in dying patients. Case
series and a small RCT suggest that TENS may
provide short-term pain relief in dying pa-tients or
in patients with intractable cancer pain.
Acupuncture may be helpful in relieving advanced
or refractory cancer pain, although data stem only
from uncontrolled studies. Acu-puncture has thus
far not been shown to re-lieve HIV-related
peripheral neuropathic pain. Supportive group
therapy can improve pain in breast cancer patients,
and this effect can be enhanced by hypnosis.
Massage therapy may provide immediate, but not
prolonged, pain relief in cancer patients; men may
re-spond better than women. There are few welldesigned studies in music therapy, and no recommendations can be made based on data thus far.

Dyspnea
Dyspnea, the subjective sensation of breathlessness, is experienced by close to 50% of pa7
tients prior to dying. Although the most effec-tive
treatment of dyspnea is to treat the underlying
etiology (such as COPD, heart fail-ure, or pleural
effusions), many times this is not possible in the
terminally ill. We identified five trials addressing
the treatment of severe dyspnea with CAM
therapies. These therapies include the following.
Acupuncture. A single-blind RCT of 24 COPD
patients (mean age 64 years) with disabling
shortness of breath (SOB) compared 13 ses-sions
of acupuncture over 3 weeks to placebo
acupuncture for 3 weeks. Patients treated with
acupuncture had less subjective breathlessness as
measured by both a 5-point SOB scale (P _ 0.01)
and a modified Borg VAS (P _ 0.02). Pa-tients
treated with acupuncture could also walk further
in 6 minutes than those treated with sham
24
acupuncture (P _ 0.05). A prospective

Vol. 20 No. 5 November 2000

study of 20 patients (median age 60 years) with

cancer-related breathlessness at rest treated with
acupuncture at LI4 (a point on the large intestine
meridian) for 10 minutes found that 70% of
patients
reported
marked
symptomatic
improvement in subjective reports of breathlessness, relaxation, and anxiety that peaked at 90
minutes (P _ 0.001) and lasted up to 6 hours (P _
0.005); respiratory rate also signifi-cantly
decreased. Eight of the responders had been
resistant to other treatments, including opioids,
25
steroids, nebulizers, and oxygen.
Acupressure. In a single-blind, randomized
crossover study of 31 patients with severe COPD
(mean age 67 years; mean forced expi-ratory
volume at 1 second that is 40% of pre-dicted)
enrolled in a 12-week pulmonary reha-bilitation
program, patients were taught acupressure to be
practiced daily at home for 6 weeks, alternating
with sham acupressure for 6 weeks. Dyspnea as
measured by VAS was signif-icantly less when
patients were receiving 6 weeks of real rather than
26
sham acupressure (P _ 0.009).
Behavioral and Psychological Therapies. An RCT
of 20 COPD outpatients (mean age 61 years; mean
FEV1 46% of predicted) were randomly assigned
to either progressive muscle relax-ation (PMR) or
being told to relax for 45 min-utes during four
weekly sessions. Half the pa-tients had severe
COPD, and another 25% had moderate disease.
PMR consisted of tension-release exercises in 16
muscle groups for four weekly sessions plus daily
home practice, using taped instructions. Patients in
the PMR group had less dyspnea by VAS
compared with con-trols during each session (P _
0.04), but no significant improvement occurred
between the first and last sessions. There were
significant improvements in anxiety, respiratory
rate, and heart rate during each session compared
27
to non-intervention controls. In a double-blind
RCT of 65 patients with severe COPD (mean age
66 years; mean FEV1 0.9 L), patients were
assigned to one of four interventions: nurse
therapy (reassurance, without psychotherapeu-tic
training), supportive therapy with psycho-analysts
(but not using transference tech-niques), analytic
therapy with psychoanalysts (using transference),
and control group (weekly labs). There was a
significant reduc-

Vol. 20 No. 5 November 2000

Use of Complementary Medicine in Palliative Care

383

placebo wrist band, and no band. Each treatment

condition lasted 4 or 8 hours. No
tion in dyspnea by the Fletcher scale (ordinal
rating of incapacity because of dyspnea) in the
nurse group compared with other groups (P _
28
0.05) but not by the VAS-dyspnea.
Oddly,
walking distance was significantly improved in the
control group.
In 20 patients with advanced small cell and nonsmall cell lung cancer (with a median sur-vival of
200 days), an RCT was conducted to evaluate the
effect of non-pharmacological in-tervention for
29
breathlessness in lung cancer. Breathlessness
was measured by VAS. All pa-tients had
completed chemotherapy or radio-therapy and
were suffering from breathless-ness. The median
age was 55 years for the intervention group and 69
years for the control group. Intervention consisted
of weekly 1-hour sessions (over 36 weeks) with a
nurse practi-tioner, who offered breathing
retraining exer-cises, counseling, relaxation, and
teaching of coping and adaptation strategies.
Control pa-tients were assessed but not taught
such skills. At 3 months, the intervention group
showed a median improvement in breathlessness
at worst of 35% (P _ 0.02), improvement in distress caused by breathlessness of 53% (P _ 0.02),
and a 17% improvement in functional capacity (P
_ 0.03). The median scores of the control group
were static or had worsened.
In summary, a number of RCTs support the use
of acupuncture and acupressure to relieve dyspnea
in patients with moderate to severe COPD. Other
RCTs support using muscle re-laxation with
breathing retraining to reduce breathlessness in
COPD and lung cancer pa-tients. Breathing
retraining, in combination with coping strategies,
may reduce breathless-ness and improve
functional capacity. Modali-ties that improve
function (i.e., walking) are important because they
may also enhance qual-ity of life. Also,
acupuncture
may
relieve
can-cer-related
breathlessness.

Nausea and Vomiting

We identified only one relevant study that
evaluated nausea and vomiting in six hospice
patients (mean age 68 years; range 3684) that
30
tested acupressure by using an N-of-1 design.
Each patient underwent three treatment conditions in alternating sequence: acupressure wrist
band (at the commonly accepted P6 acu-point),

significant differences were found between the

treatment conditions, although the sample size was
small and two patients had no nausea dur-ing the
treatment period.
Most studies of CAM therapies for nausea and
vomiting were conducted in chemother-apyassociated and anticipatory nausea and vomiting,
which occurs in 24% to 65% of can-cer patients
undergoing serial chemotherapy. Such CAM
therapies
include
behavioral
thera-pies,
acupuncture, and ginger. It is unclear whether
treatments for chemotherapy-associ-ated nausea
and vomiting can be extrapolated to those not
undergoing chemotherapy. This discussion is
outside the scope of this review.
In summary, no large-scale trials have been
done in terminally ill patients with nausea and
vomiting that is not associated with chemotherapy.

Discussion
In this article, we have undertaken a system-atic
review of CAM therapies to improve the comfort
of patients near the end of life or in the late stages
of illness. We have found that trials of CAM
modalities may not be readily identified on routine
literature searches. De-spite the paucity of
controlled trials, there are data that support the use
of some CAM modal-ities in terminally ill
patients. Relaxation tech-niques, acupuncture, and
TENS may improve intractable pain in dying
patients. Acupunc-ture and breathing retraining
(with or without additional coping strategies) may
ease dyspnea in moderately to severely impaired
COPD pa-tients, who have a limited prognosis,
albeit a less predictable one compared to cancer
pa-tients. These results should not be extrapo-lated
to those dying with a malignant pleural effusion or
with end-stage heart failure; in-stead, studies
should be done in these popula-tions directly to
assess their effectiveness. Table 4 outlines
evidence-based recommendations derived from
this review. Of note, no studies that used herbal
and dietary supplements were found during this
literature review, although herbal medicines were
included in the search terms (under the heading of
Alternative
Medi-cine).
Conducting
methodologically rigorous research in palliative
care and CAM is chal-lenging. Both of these fields
traditionally have not attracted research funding,
making it diffi-

384

Pan et al.

Vol. 20 No. 5 November 2000

highly prev-alent symptoms in patients near the

end of life,
cult to undertake large-scale studies. In pallia-tive
care, symptomatically ill patients may be
unwilling to participate in a placebo-controlled
RCT or may have impaired ability to compre-hend
informed consent or to complete out-come
assessments because of severe illness and
31
fluctuating mental status. Complex ethical issues also surround the recruitment of termi-nally
ill patients in research studies. Outcomes of
concern to patients (e.g., quality of life or
existential meaning) may be difficult to evalu-ate
because of the lack of formal assessment tools and
difficulty quantifying this type of in-formation. In
addition, there may be vast incon-sistency in the
duration of endpoints, thus mak-ing it difficult to
compare and contrast results.
In terms of CAM, additional barriers exist.
First, many CAM modalities involve hands-on
treatment, such as the use of acupuncture, TENS,
massage, or chiropractic maneuvers. Al-though
many of the studies we reviewed used sham
procedures, there is no consensus on what
constitutes appropriate sham controls and it is
difficult to maintain blinding during the studies.
Hands-on CAM therapies may not be amenable to
study using the established RCT designs that work
so well with pharmaco-logic therapies. Second,
studies involving CAM therapies may be less
likely to be published in refereed, indexed
journals. It is important to note that many of the
articles (12 of 21) were not identified by extensive
search methods but by reference in the searched
articles or by ex-perts. Third, CAM study results
testing herbal or dietary supplements may not be
reproduc-ible in clinical practice as the
composition of commercially available CAM
products cannot be guaranteed in the United
States, because most are classified as dietary
supplements and, therefore, not subject to Food
and Drug Ad-ministration oversight regarding the
32,33
purity or quantity of ingredients.
There are a number of strengths to this re-view.
First, it is comprehensive, including six databases
encompassing the medical, cancer, AIDS, nursing
and allied health, and social work literature.
Foreign languages literature was reviewed and,
interestingly, consisted of general overviews rather
than original reports. To our knowledge, this
represents the first sys-tematic review of CAM use
in the palliative care setting. Second, we examined

which are clinically relevant, common, and

morbid. Third, we employed a rigorous hierar-chy
of evidence by using a best evidence ap-proach.
That is, we sought data initially from metaanalyses and from randomized con-trolled trials. If
such data were not available, we reviewed the
literature for other controlled tri-als and, finally,
consulted case series. Small case reports or
anecdotal reports were not con-sidered.
Several limitations to this study should be
noted. Our search was constrained by indexing
limitations, such as changes in terminology over
time and difficulty finding literature about dying
patients. We were limited by the content of the
databases and may have missed publications that
are not indexed in them. We have found that trials
of CAM modalities may not be easily found on
routine literature searches. Also, we were limited
by the quality and quantity of existing data on this
subject. A meta-analysis could not be performed
because of the breadth and heterogeneity of
therapeu-tic modalities and symptomatology and
be-cause most studies were not randomized trials.
Despite these limitations, our review sug-gests
that there is evidence to support the use of some
CAM therapies in terminally ill pa-tients and,
indirectly, in treating symptoms that are highly
prevalent in such patients (Ta-ble 3). In addition to
the use of traditional an-algesics (including
opioids) for pain manage-ment, it may be valuable
to complement this with massage, relaxation and
imagery, hypno-sis, or TENS therapy. In addition
to treating underlying causes of dyspnea in
patients with moderately to severely compromised
breath-ing, it can be useful to add acupuncture,
acu-pressure, relaxation, and breathing retraining
and coping strategies. For many patients, such
approaches may be culturally sensitive, as well as
enhance the patients sense of control over their
illness.
Additionally, we identified other CAM modalities that appear promising, based on small
controlled studies or case series (e.g., massage for
pain relief, acupuncture for cancer-related pain
and dyspnea) and that would benefit from study in
larger controlled trials. Finally, we identified gaps
in the literature that need to be addressed in future
research, including treatment of dyspnea in dying
cancer patients and the use of simple measures
such as P6 acu-

Vol. 20 No. 5 November 2000

Use of Complementary Medicine in Palliative Care

385

Table 4
Summary of Evidence and Recommendations for CAM Use in Dying or Severely Ill Patients
Symptom
Pain

Dyspnea

Nausea/
vomiting

Recommendations

Reference No.

1. TENS may provide short-term pain relief in

patients with intractable or advanced cancer pain.
2. Acupuncture may provide short-term relief in
terminally ill patients with cancer pain.
3. Acupuncture does not appear to provide pain relief
in patients with HIV neuropathy.
4. Massage with or without aromatherapy, might
provide short-term relief for patients with intractable
cancer pain.
5. Relaxation with imagery and cognitive coping skills
can improve pain in patients with oral mucositis.
6. Relaxation might relieve pain in terminally ill
patients.
7. Support group therapy can improve pain in
advanced cancer patients.
Hypnosis can enhance such pain relief.
1. Acupuncture may provide relief in patients with
malignancy-related breathlessness.
2. Acupuncture/acupressure and breathing retraining
with progressive muscle relaxation improve dyspnea
in severe COPD patients.
3. Psychoanalytic therapy does not appear to relieve
dyspnea in COPD patients.
1. Acupressure does not appear to relieve cancerrelated nausea in a small study.

Evidence Strength

1113

3.1 3.2

1415

3.1 3.2

16

17, 18, 19

3.3

21

20

3.3

23

25

3.1

24, 26, 27

28

24

3.1

TENS _ transcutaneous electrical nerve stimulation; HIV _ human immunodeficiency virus; COPD _ chronic obstructive pulmonary disease. USPSTF
(United States Preventive Services Task Force) Guidelines:
1 _ Evidence obtained from a systematic review of relevant randomized controlled trials. 2 _
Evidence obtained from at least one properly designed randomized controlled trial.
3.1 _ Evidence obtained from well designed controlled trials without randomization.
3.2 _ Evidence obtained from well designed cohort or case control analytic studies, preferably from more than one center or research group.
3.3 _ Evidence obtained from multiple time series with or without the intervention.
b
4 _ Opinions of respected authorities, based on clinical experience, descriptive studies, or reports of expert committees. Only
studied in one trial.

pressure to alleviate opioid-induced nausea and

emesis. There are also scarce data con-cerning
symptom relief in HIV patients near the end of
life.
Many CAM therapies are simple to use and
appear to have minimal side effects. Our re-view
found little patient dropout because of ad-verse
side effects in these studies. Other studies support
36
this finding. For example, Yamashita et al.
reported that in an acupuncture center performing
55,291 treatments over a 5-year pe-riod, only 64
adverse events occurred. Most events were
transient and without serious se-quela, such as
failure to remove needles, tran-sient dizziness,
discomfort, perspiration, and minor ecchymosis.
There is little data in the lit-erature on the adverse
effects of TENS; the few that exist report minor
side effects, such as skin irritation from lubricant
37,38
jellies, or none at all.
This probably reflects a
combination of

rare adverse effects as well as lack of a formal

reporting system.
Many CAM therapies have a favorable benefit/risk ratio and may be useful adjunctive therapies to conventional (allopathic) medicine.
However, reimbursement poses a major bar-rier to
the use of CAM modalities, even those with
demonstrated efficacy, because some CAM
therapies can be costly and are not cov-ered by
third-party payers.
It is important to keep in mind that both the
field of CAM and palliative medicine are evolving and dynamic. Some of the CAM therapies
mentioned here, such as TENS and behavioral
therapy, may be straddling the sometimes arbitrary and blurry boundary between conven-tional
and complementary medicine. Similarly, a new
concept of palliative care is developing,
recognizing that there need not be a dichot-omy
between the models of palliative and cura-

386

Pan et al.

Vol. 20 No. 5 November 2000

8. Slavin RE. Best evidence synthesis: an intelligent

tive care. Future studies in CAM and palliative

care require sound designs, larger sample sizes,
reliable blinding, and specific and clinically relevant outcomes measures, including the effect on
the concomitant use of conventional thera-pies.
With continued open-minded research, we should
be able to integrate the best thera-pies from both
domains to advance comfort and to ease suffering
in dying patients.

Acknowledgments
This study was supported in part by the Commonwealth Fund. Dr. Pan is a fellow in Pfizer/
American Geriatrics Society and in the John A.
Hartford Foundation Center of Excellence. Dr.
Morrison is the recipient of a National In-stitute
on
Aging
Mentored
Clinical
Scientist
Development Award (K08AG00833-01) and is a
faculty scholar for the Open Society Institute
Project on Death in America. Dr. Leipzig is the coprincipal investigator for the Mount Sinai John A.
Hartford Foundation Center of Excel-lence. The
authors would like to thank Dr. Christine K.
Cassel for her valuable advice and review of the
manuscript. We also gratefully ac-knowledge the
patience and help of Veronica Serrano and
Deborah George for administra-tive assistance,
and Kristy Kime for manuscript preparation.

References
1. Eisenberg DM, Kessler RC, Foster C et al. Unconventional medicine in the United States: preva-lence,
costs, and patterns of use. N Engl J Med 1993; 328:246
252.
2. Lerner KJ, Kennedy BJ. The prevalence of questionable methods of cancer treatment in the United
States. CA Cancer J Clin 1992;42:181191.
3. Montbriand MJ. Freedom of choice: an issue
concerning alternate therapies by patients with can-cer.
Oncol Nurs Forum 1993;20:11951201.
4. Ernst E. Complementary AIDS therapies: the good,
the bad and the ugly. Int J STD AIDS 1997;8: 281285.
5. McKnight I, Scott M. HIV and complementary
medicine. Med J Aust 1996;165:143145.
6. World Health Organization. Cancer Pain and Relief
(Technical Report Series 804). Geneva: World Health
Organization; 1990.
7. Seale C, Cartwright A. The Year Before Death.
Brookfield, VT: Ashgate Publishing, 1994.

alternative to meta-analysis. J Clin Epidemiol 1995;

48:918.
9. SUPPORT Principal Investigators. A controlled trial
to improve care for seriously ill hospitalized pa-tients.
The study to understand prognoses and pref-erences for
outcomes and risks of treatments (SUP-PORT). JAMA
1995;274:15911598.
10. Gadsby JG, Franks A, Jarvis P et al. Acupuncturelike transcutaneous electrical nerve stimulation within
palliative care: a pilot study. Complement Therap Med
1997;5:1318.
11. Avellanosa AM, West CR. Experience with transcutaneous electrical nerve stimulation for relief of
intractable pain in cancer patients. J Med 1982;13: 203
213.
12. Ostrowski MJ. Pain control in advanced malig-nant
disease using transcutaneous nerve stimula-tion. Br J
Clin Pract 1979;33:157162.
13. Wen HL. Cancer pain treated with acupuncture and
electrical stimulation. Mod Med Asia 1977;13: 1216.
14. Xu S, Liu Z, Li Y. Treatment of cancerous abdominal pain by acupuncture on Zusanli (ST 36): a
report of 92 cases. J Tradit Chin Med 1995;15:189
191.
15. Filshie J, Redman D. Acupuncture and malig-nant
pain problems. Eur J Surg Oncol 1985;11:389 394.
16. Shlay JC, Chaloner K, Max MB et al. Acupunc-ture
and amitriptyline for pain due to HIV-related
neuropathy: a randomized controlled trial. JAMA
1998;280:15901595.
17. Weinrich SP, Weinrich MC. The effect of mas-sage
on pain in cancer patients. Appl Nurs Res 1990; 3:140
145.
18. Ferrell-Torry AT, Click OJ. The use of therapeu-tic
massage as a nursing intervention to modify anxi-ety
and the perception of cancer pain. Cancer Nurs
1993;16:93101.
19. Wilkinson S. Aromatherapy and massage in palliative care. Int J Palliat Nurs 1995;1:2130.
20. Fleming U. Relaxation therapy for far-advanced
cancer. Practitioner 1985;229:471475.
21. Syrjala KL, Donaldson GW, Davis MW et al. Relaxation and imagery and cognitive-behavioral train-ing
reduce pain during cancer treatment: a con-trolled
clinical trial. Pain 1995;63:189198.
22. Curtis S. The effect of music on pain relief and
relaxation of the terminally ill. J Music Ther 1986;
23:1014.
23. Spiegel D, Bloom J, Yalom I. Group support for
patients with metastatic cancer. A randomized prospective outcome study. Arch Gen Psychiatry 1981;
38:527533.
24. Jobst K, Chen JH, McPherson J et al. Controlled
trial of acupuncture for disabling breathlessness. Lancet
1986;ii:14161418.

Vol. 20 No. 5 November 2000

Use of Complementary Medicine in Palliative Care

25. Filshie J, Penn K, Ashley S et al. Acupuncture for

the relief of cancer related breathlessness. Palliat Med
1996;10:145150.
26. Maa SH, Gauthier D, Turner M. Acupressure as an
adjunct to a pulmonary rehabilitation program. J
Cardiopulm Rehab 1997;17:268276.
27. Renfroe KL. Effect of progressive relaxation on
dyspnea and state anxiety in patients with chronic
obstructive pulmonary disease. Heart Lung 1988;17:
408413.
28. Rosser R, Denford J, Heslop A et al. Breathlessness and psychiatric morbidity in chronic bronchitis and
emphysema: a study of psychotherapeutic management. Psychol Med 1983;13:93110.
29. Corner J, Plant H, AHern R, Bailey C. Nonpharmacological intervention for breathlessness in lung
cancer. Palliat Med 1996;10:299305.
30. Brown S, North D, Marvel MK, Fons R. Acupressure wrist bands to relieve nausea and vomiting in
hospice patients: do they work? Am J Hosp Palliat Care
1992;9:2629.
31. Calman SK, Hanks G. Doyle D, Hanks GWC,
MacDonald, eds. Oxford Textbook of Palliative

387

Medicine, 2nd ed. Oxford: Oxford University Press,

1998:159165.
32. Liberti LE, Der-Marderosian A. Evaluation of
commercial ginseng products. J Pharm Sci 1978;67:
14871489.
33. OReilly JR. Food and Drug Administration, 2nd
ed. New York: Shepards/McGraw-Hill, 1995.
34. National Library of Medicine. PubMedMEDLINE
(1966-1999). Available at://www.ncbi.nlm.nih.gov/
PubMed/. Accessed.
35. Wack JT, Turk DC. Latent structure in strategies for
coping with pain. Health Psychol 1984;3:2743.
36. Yamashita H, Tsukayama H, Tanno Y et al. Adverse events related to acupuncture (letter). JAMA
1998;280:15631564.
37. Johnson MI, Ashton CH, Thompson JW. An indepth study of long-term users of transcutaneous
electrical nerve stimulation (TENS). Implications for
clinical use of TENS. Pain 1991;44:221229.
38. Carroll D, Tramer M, McQuay H, Nye B, Moore A.
Randomization is important in studies with pain
outcomes: systematic review of transcutaneous electrical nerve stimulation in acute postoperative pain. Br J
Anaesth 1996;77:798803.