Palliative Medicine

http://pmj.sagepub.com/

The level of need for palliative care: a systematic review of the literature
Peter J Franks, Chris Salisbury, Nick Bosanquet, Emma K Wilkinson, Suzanne Kite, Anne Naysmith and
Irene J Higginson
Palliat Med 2000 14: 93
DOI: 10.1191/026921600669997774

The online version of this article can be found at:

http://pmj.sagepub.com/content/14/2/93

Published by:

http://www.sagepublications.com

Additional services and information for Palliative Medicine can be found at:

Email Alerts: http://pmj.sagepub.com/cgi/alerts

Subscriptions: http://pmj.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations: http://pmj.sagepub.com/content/14/2/93.refs.html

>> Version of Record - Mar 1, 2000

What is This?

Downloaded from pmj.sagepub.com at NORTH CAROLINA STATE UNIV on November 6, 2013

02pm308.qxd 16/3/00 8:52 am Page 93

Palliative Medicine 2000; 14: 93–104

The level of need for palliative care: a systematic

review of the literature
Peter J Franks Co-director, Centre for Research and Implementation of Clinical Practice, Thames Valley
University, London, Chris Salisbury Consultant and Senior Lecturer, Department of Primary Care, University of
Bristol, Nick Bosanquet Professor of Health Policy, Department of Primary Care and General Practice, Imperial
College School of Medicine at St Mary’s Hospital, London, Emma K Wilkinson Project Researcher, Wessex
Institute for Health Research and Development, University of Southampton, Suzanne Kite Senior Registrar,
Edenhall Marie Curie Centre, London, Anne Naysmith Consultant in Palliative Medicine, Parkside Health
Hospital, Pembridge Palliative Care Centre, London and Irene J Higginson, Professor and Head of Department
of Palliative Care and Policy, King’s College School of Medicine and Dentistry, St Christopher’s Hospice, London

Abstract: Palliative care services have developed rapidly over the past 30 years, with
little evaluation as to how needs have been met by these new services. As part of a
systematic review of palliative care, evidence of the needs of patients and carers has
been evaluated from the current literature. Of the total of 673 articles related to the 10
areas within the main review, 64 provided evidence on the need for palliative care
services over the period from 1978 to 1997. A further nine articles were added in
November 1998 after the end of the study to update the review with more recent
research. Need can be assessed in one of two ways: either by adopting an
epidemiological approach or by examining health service usage. In the former, evidence
is provided on disease-specific mortality, and related to the duration of symptoms prior
to the patient’s death. As an example of this, it is suggested that services may need
to provide pain control for 2800 patients per million (p/M) population dying from cancer
each year and 3400 p/M with noncancer terminal illness. Using health service usage
as an indicator of need, 700–1800 p/M with cancer and 350–1400 p/M with noncancer
terminal illness would require a support team or specialist palliative home care nurse,
with 400–700 cancer p/M and 200–700 noncancer p/M requiring inpatient terminal care.
Studies indicate that at present usage, palliative care is being provided by 40–50 hospice
beds/M. Despite this provision, there remains evidence that in certain areas of care such
as pain control, there still remains a high degree of unmet need.

Key words: needs assessment; palliative care; terminal care; utilization review

Resumé: Les services de soins palliatifs se sont rapidement développés au cours des
30 dernières années, Mais la façon dont les besoins ont été satisfaits par ces nouveaux
services n’a pas été sérieusement évaluée. L’évaluation des besoins des patients et des
soignants en soins palliatifs a été faite à partir d’une revue de la littérature courante,
dans le cadre d’une revue systématique des soins palliatifs. Sur les 673 articles en lien
avec dix sujets d’intérêt de la revue principale, 64 ont prouvé le besoin de services en
soins palliatifs durant la période allant de 1978 à 1997. Neuf articles supplémentaires
ont été ajoutés en novembre 1998 pour actualiser les références. Les besoins peuvent

Address for correspondence: Dr Peter J Franks, Centre for

Research and Implementation of Clinical Practice, Thames Valley
University, Wolfson Institute of Health Sciences, 32–38 Uxbridge
Road, London W5 2BS, UK. E-mail: peter.franks@tvu.ac.uk

© Arnold 2000 0267–6591(00)PM308OA

02pm308.qxd 16/3/00 8:52 am Page 94

94 PJ Franks et al.

être évalués de deux façons: soit par l’épidémiologie, soit en étudiant le recours des
services de santé. Dans le 1er cas, l’évaluation est basée sur la mortalité propre à chaque
maladie et corrélée à la durée des symptômes avant la mort des patients. Ainsi, par
exemple, cette méthode suggère que les services de santé puissent avoir à assurer le
traitement de la douleur de 2800 patients par million (p/M) de patients mourant de
cancer chaque année et de 3400 (p/M) de patients mourant de maladie non cancéreuse.
D’autre part, utiliser le recours aux services de santé comme indicateur des besoins
indiquerait que 700 à 1800 p/M présentant un cancer et 350 à 1400 p/M avec une
maladie terminale non cancéreuse nécessiterait une équipe de soutien ou une infirmière
Macmillan, dont 400 à 700 cancéreux p/M et 200 à 700 non cancéreux p/M nécessitant
des soins terminaux en institution. Ces études indiquent dans la situation présente, que
les soins palliatifs sont dispensés grâce à 40 à 50 lits d’hospitalisation par million. Malgré
cette ressource, il demeure évident que dans certains domaines du soin, tel que le
contrôle de la douleur, il y a encore une forte proportion de besoins non satisfaits.

Mots-clés: soins palliatifs; soins terminaux; évaluations des besoins; revue sur l’usage
des soins palliatifs

Introduction had either not been received or had been published

The contents of this paper formed part of the sys-
tematic review of palliative care, details of which are
given elsewhere.1 Briefly, the review was carried out
using a predefined protocol. Inclusion criteria were
that the article had to be of studies involving pal-
liative care for any disease published in Europe,
North America, Australasia or Israel since 1978, in
English, French, German, Italian or Swedish. Arti-
cles were excluded if they were personal opinions or
case studies, history of palliative care provision, eth-
ical issues, the development of research instruments
or about cancer services generally, without special
reference to palliative care. The data sources
included nine computer-based databases, two
paper-based databases and hand searching of five
specialist journals. Major sponsors of research
work and 24 organizations involved in palliative care
work were contacted. In addition, researchers who
are active in the field of palliative care were con-
tacted to provide further information. A standard
data sheet was used to collect the information held
within each paper, which was then synthesized into
a narrative to evaluate the evidence. The remit of
this paper was to review and evaluate the current
literature on the level of need for palliative care. In
all, 673 papers were retrieved for the overall sys-
tematic review, of which 64 papers were relevant to
the question of level of need and included in this
review. A further nine papers were added in
November 1998 after the end of the study, which
in the interim.
Assessing need
The term ‘need’ is frequently used in evaluating the
effectiveness of health care, but is rarely defined. It
is important to define need in order to make
objective assessments of health care requirements.
In epidemiology, need refers to ‘specific indicators
of disease or premature death which require
intervention because the level is above that
generally accepted within a particular society’.2
Within this definition needs must reflect value
judgements on the ability to control the problem,
and the level considered acceptable within society.
As such, need will be dependent on the type of soci-
ety being evaluated and the evidence provided by
both users (patients) and health care professionals.
When considering the current level of need for
palliative care one must define the patient groups
in whom palliative care may be appropriate. In the
past, palliative care services have often been
restricted to patients suffering from terminal can-
cer, with hospice services dealing almost exclusive-
ly with these patients. However, there is a growing
recognition that other diseases require periods of
palliation. There are three main disease categories
that may require palliative care: cancer, nonmalig-
nant progressive diseases and children’s terminal ill-
nesses (Table 1).3
To estimate the need for palliative care two
strategies may be adopted, either using the epi-
02pm308.qxd 16/3/00 8:52 am Page 95
Table 1 Types of illness that may require palliative care3
Illness Categories
Cancer Lung, trachea, bronchus, ear,
nose and throat, female
breast, lymphatic, digestive
tract, genito-urinary,
leukaemia, haemopoetic
Progressive nonmalignant Circulatory:
disease cardiovascular, cerebrovascular
Respiratory
Nervous system:
motor neurone disease,
multiple sclerosis, dementia
AIDS/HIV
Children’s terminal Hereditary degenerative
illnesses disorders, muscular dystrophy,
cystic fibrosis
demiological approach or through the evaluation of
health service use. The epidemiological approach
uses the cause-specific mortality in diseases likely to
benefit from palliative care, and then relates this to
the type and frequency of symptoms experienced by
patients suffering from the terminal stages of these
diseases. This method is useful in identifying all pal-
liative care needs, since it is not dependent on the
source of the care offered. It identifies patients who
may be receiving specialist care, those who are
receiving care through nonspecialist sources such as
the regular community and acute health services,
and includes patients in whom the need for services
may be unmet.
As an alternative to this methodology health ser-
vice use may provide a useful starting point in eval-
uating need. However, this method may be limited
in only evaluating evidence from patients within
specified health care services, and may not evaluate
alternative services. As such this methodology can-
not evaluate unmet need. To complete the picture
one must evaluate unmet needs not only in the users
of specific services, but also attempt to find
patients who are not receiving care who would ben-
efit from the services on offer.
The current paper will evaluate the evidence pro-
vided from both epidemiology studies and evalua-
tions of current services and attempt to determine
the types of services required for effective care of
patients with terminal illness.
The level of need for palliative care 95
Using the epidemiological approach to
needs assessment
Patients who die can be grouped into three cate-
gories:
1) those undergoing a palliative care period prior to
death;
2) those with stable disease, with few symptoms fol-
lowed by sudden death;
3) those who suffer from stable disease, with peri-
ods of progression of symptoms which would ben-
efit from periods of palliative care.3
For patients suffering from cancer, the predicted
progression would most often fall into category (1).
Nonmalignant disease would quite often fall into
categories (2) and (3), particularly circulatory dis-
eases where death may follow a short symptomatic
period. It is important that the patients who fall into
the latter categories are still considered within the
complete picture of palliative care, as they may also
require symptom relief for a period prior to death.
Mortality statistics in England and Wales
Disease-specific mortality can be estimated from
information supplied through death certificates,
tabulated by the Office for National Statistics (for-
merly Office of Populations, Censuses and Surveys).
The most recent figures available cover 1995.4
While it is acknowledged that death certification
includes many inherent inaccuracies concerning the
cause of death, there is some evidence that the
errors in certification are randomly distributed.5
Thus, although the individual certificate may not be
an accurate reflection of the individual’s cause of
death, overall disease-specific mortality rates are
reasonably accurate. Table 2 gives the death rates
for causes that may require a period of palliation.
While cancer deaths are considered most often in
palliative care services, it is clear that these make up
only a proportion of all deaths, which may require
some form of palliative care. Deaths due to circu-
latory diseases are far more common than cancer
deaths, though relatively few of these will include
prolonged symptomatic periods requiring palliation.
To establish the importance of palliative care in
these diseases one must examine the frequency,
duration and intensity of symptoms.
02pm308.qxd 16/3/00 8:52 am Page 96

96 PJ Franks et al.

Table 2 Mortality statistics in England and Wales. Values are deaths per million population4

Circulatory Respiratory Nervous Congenital

Age (years) Sex Neoplasms diseases diseases diseases abnormalities
All Male 2831 4432 1549 174 26
Female 2543 4630 1782 191 25
<1 Male 24 96 223 184 394
Female 28 76 165 127 351
1–4 Male 34 12 20 32 47
Female 29 13 22 34 37
5–14 Male 40 11 9 18 13
Female 31 8 11 12 16
15–24 Male 59 37 22 45 14
Female 46 21 12 25 12
25–34 Male 109 91 48 38 15
Female 141 59 26 28 11
35–44 Male 334 401 98 54 15
Female 489 162 46 44 14
45–54 Male 1338 1557 238 92 15
Female 1539 511 152 76 16
55–64 Male 4503 5330 929 171 20
Female 3627 2041 630 140 22
65–74 Male 12 073 16 345 4343 447 28
Female 7602 8454 2732 366 30
75–84 Male 22 450 41 150 16 099 1443 64
Female 12 029 27 804 9130 975 46
85+ Male 31 605 82 490 51 335 3103 79
Female 15 918 71 122 34 723 2126 54

Symptom prevalence in patients who have died
Two methods have been proposed to assess the
prevalence of symptoms in patients in the terminal
stages of a disease. Patients may be examined while
receiving specialist health care, or relatives/health
care professionals may be contacted following the
patients’ death to ascertain symptoms suffered prior
to death. Both methods are problematic in provid-
ing accurate assessments for patient populations. In
the former method, only patients attending a par-
ticular health facility will be interviewed. It is
acknowledged that patients who attend such units
may be atypical compared with patient populations
as a whole. To overcome this, other studies have
examined proxy reporting of symptoms in random
samples of patients who died from particular caus-
es. This method may produce a representative sam-
ple, but introduces errors due to proxy reporting of
symptoms by the patients’ relative or health care
provider. In a review of the evidence on quality of
life assessments we have reviewed the evidence on
proxy reporting of symptoms.1 While some studies
show that carers overestimate the severity of symp-
toms compared with patients, others suggest that
these may be reliable estimates of patient symptom
prevalence. Others have shown that concurrent rat-
ing by patients and carers is consistent, whereas over
time following death, carers’ descriptions of the fre-
quency and severity of symptoms may change. The
status of the carer may also be an important factor
in symptom rating, with general practitioners
recording fewer symptoms than hospital nurses,
while relatives often record a higher frequency of
symptoms than health professionals.6 There is a
clear need for prospective studies to make accurate
recordings of symptoms in random samples of the
patient population.
Prevalence of symptoms in patients suffering
from terminal cancer
The prevalence of symptoms in patients with ter-
minal cancer are given in Table 3. Clearly, there is
little consistency between different studies. This is
probably a consequence of differences in patient
populations and the tools used to assess symptom
frequency. The difference in reporting of symptoms
is indicated by the number of cells that have miss-
ing data. In some studies prevalence of vomiting
and/or nausea have been combined.7–10 There may
be additional difficulties in distinguishing between
some symptoms with blurring of the boundaries
between symptoms such as weakness and fatigue.
 02pm308.qxd

Table 3 Prevalence of symptoms in patients with terminal cancer (%)

Ref. Patient Pain Dyspnoea Fatigue Vomiting Nausea Poor sleep Weakness Confusion Appetite Incontinence Constipation Depression Anxiety
7 Random 84 47 – – – 51 – 33 71 47 37 – 38
(n = 168)
8 Mixed 67 28 – – – – 35 – – 38 – – 12
(n = 107)
16/3/00 8:52 am

9 Mixed 36 38 – – – – 25 – – 9 – – –
(n = 47)
10 Mixed 52 21 – – – – 12 3 – 19 6 – –
(n = 32)
12 Mixed 11 53 – – 44 – – 12 79 54 – – –
Page 97

(n = 1592)
13 Mixed 34 28 52 – 13 57 49 28 6 7 6 18 4
(n = 90)
14 Mixed 57 64 – 27 30 57 – 47 87 64 40 29 54
( n = 80)
15 Mixed 71 51 – – – 69 81 – 55 – – – –
(n = 75)
18 Mixed 30 47 – 10 13 28 –68 – 55 – 46 39
(n = 176)
19 Mixed 72 40 – – – 33 56 7 – 13 – – 18
(n = 55)
20 Random 88 54 – – – 60 – 41 78 62 – 32 69
(n = 2074)

Site-specific studies
16 Small-cell 45 87 20 38 75 88 – 75 55 – 65 60
lung cancer
(n = 232)
16 Nonsmall-cell 42 87 85 10 25 58 84 – 55 40 – 60 42
lung cancer
(n = 423)
31 Breast – – – – – – – – – – – 18 18
(n = 211)
The level of need for palliative care 97
02pm308.qxd 16/3/00 8:52 am Page 98
98 PJ Franks et al.
Within this type of study there is also little infor-
mation on the severity of symptoms and the inca-
pacity that they may cause the patient. Different
studies have described a range of symptoms,
though there is little evidence on the relative impor-
tance of such symptoms to patients.
Pain is the most frequently investigated symptom
in patients suffering from terminal cancer. Howev-
er, there is considerable variation in its reported fre-
quency, varying from 11% to 84% in different
patient groups.7–20 This variation may be caused by
methodological differences, not least the method of
symptom ascertainment, selection of patients, and
the use of proxy reporting. It is surprising that
reported pain was highest in the only truly random
sample of patients, since it might be expected that
pain would be worse in the most severe cases being
treated by specialist units. Similar levels of pain
have been reported in a sample of 45 hospice
patients and 126 cancer patients treated else-
where.21 In all, 93% of hospice patients experienced
pain, compared with 80% of other cancer patients.
In a review of the literature22 it was estimated that
moderate to severe pain occurs in 45–100% of
patients with advanced cancer, in whom it is inade-
quately managed in 20–40%.
Respiratory problems are also frequently cited in
studies of terminal cancer patients, with prevalence
varying from 21% to 64% with mixed cancer
sites.7–10,12–15,18–20 As expected, this is substantially
higher (87%) in the patients suffering from lung
cancer.16 Frequency of nausea varies in different
studies from 13% to 44%,12–14,16,18,23 and vomiting
from 10% to 27%.14,16,18,23 In studies where vomit-
ing and nausea have been combined, frequency
ranges from 9% to 51%.7–10
Using the figures derived from a random sample
of patients dying from cancer,7 it has been estimat-
ed that 2400 patients per million (p/M) of the 2800
cancer deaths per million would have experienced
pain prior to their death, with 1300 p/M suffering
breathing difficulties and 1400 p/M experiencing
symptoms of vomiting or nausea.3
Prevalence of symptoms in patients suffering
from nonmalignant terminal disease
A number of studies have investigated the preva-
lence of symptoms in cancer and noncancer deaths
combined,6,24–26 with only one random sample
examining symptoms in patients dying from non-
cancer causes (Table 4).7 Other studies have exam-
ined patient groups with other specific conditions
including human immunodeficiency virus (HIV),27
motor neurone disease (MND),28 coronary heart
disease29 and dementia.30 Prevalence of symptoms
in these groups appear to mirror those found in the
random sample, except for symptoms relating to the
particular disease process. Examples of this are the
high prevalence of weakness in patients suffering
from MND27 and confusion in those suffering from
dementia.28
Using the data on 471 noncancer deaths in the
random sample, a lower proportion of these
patients experienced pain (67%) than the cancer
patients (87%), though there were still high pro-
portions of patients who experienced respiratory
problems (49%) and nausea/vomiting (27%). It has
been estimated that of 6900 p/M with progressing
nonmalignant disease, 3400 p/M would experience
pain, 3400 p/M would suffer from respiratory prob-
lems and 1900 p/M would experience vomiting or
nausea.3
Psychological impact of terminal disease
While much work has examined the prevalence of
disease-related symptoms, several studies have also
examined the psychological impact of the disease.
Hopwood et al.31 identified that 27% of women with
advanced breast cancer suffered from either anxi-
ety, depression or a combination of both. In a ran-
dom sample of nonsudden deaths6 anxiety was
recorded in 24% of patients and depression in 18%.
This has been supported from other studies of
patients with cancer7,8,13,14,16 and mixed cancer and
noncancer terminal illness.25,28 Using these esti-
mates it has been proposed that approximately
700 cancer p/M and 1600 noncancer p/M per year
suffer from acute anxiety.3 In addition, it is recog-
nized that families coping with a terminally ill mem-
ber may also suffer from psychological distress.32
Assuming that one-third of families experienced
severe anxiety, it has been estimated that 930
families/M with a terminal cancer patient and 2200
families/M with a noncancer terminally ill relative
would suffer.3 Clearly, palliative care services must
address the psychological as well as physical symp-
toms associated with the disease process, not only
in the patient, but also in the family units support-
ing the patient.
 02pm308.qxd

Table 4 Prevalence of symptoms in patients with cancer and noncancer terminal disease (%)

Ref. Patient Pain Dyspnoea Fatigue Vomiting Nausea Poor sleep Weakness Confusion Appetite Incontinence Constipation Depression Anxiety
16/3/00 8:52 am

6 Mixed 52 42 – 20 16 19 52 23 – – 35 24 18
(n = 262)
24 Mixed 60 20 7 – – – 35 5 – 25 – – –
(n = 846)
25 Mixed 66 62 – – – – 81 38 66 42 51 30 25
Page 99

(n = 53)
26 Mixed 69 69 88 27 54 88 – 35 – 54 31 – –
(n = 26)

Noncancer studies
7 Random 67 49 – – – 36 38 – 38 32 33 – 36
(n = 471)

Disease-specific studies
27 HIV – – 57 – – 65 – – 63 – – – –
(n = 40)
28 Motor 77 50 – 23 – 64 100 – – 86 18 32 27
neurone
disease
(n = 22)
29 Coronary 63 51 – – 45 – – 43 37 – 30 59 –
heart disease
(n = 600)
30 Dementia 64 – – – – – 83 57 59 72 – 61 –
(n = 170)
The level of need for palliative care 99
02pm308.qxd 16/3/00 8:52 am Page 100
100 PJ Franks et al.
Evidence of need derived from
specialist sources
In the UK, palliative care services are provided
from three main sources: hospital, hospice and com-
munity services. These services contribute different
types of care to patients with terminal illness and
cater for different groups. While the hospital ser-
vices concentrate on symptom control and
respite,33 home care services are chiefly perceived
to provide support, symptom management and
counselling.34 On average, 90% of the last year of a
terminally ill patient is spent at home, with home
care being highly valued by patients and carers.35
There are no figures available on the precise
requirements of the three services, though there is
some evidence that patients may not be receiving
the care they wished. In a needs assessment exercise
in Scotland, 62% of patients died in hospital com-
pared with 28% at home and 3% in hospices.36 Half
of these patients preferred to die at home, one-
quarter in a hospice and only 10% wished to die in
hospital. Only 34% people died in their preferred
place of death. People who wished to die at home
were most often unable to do so due to poor symp-
tom control. Referrals from GPs appear to be fre-
quently related to poor control of symptoms and
poor pain management.37 Patients who are admit-
ted to hospices generally want admission to the
facility, whilst referral to hospitals is only request-
ed by one-quarter of patients.38,39 Of cancer
patients dying at home, 97% felt it was the right
place, whereas only half of hospital patients felt the
same way.14 Patients are more likely to die in hos-
pital or hospices if elderly,40 single and poor.41 In
general, hospital services have been criticized for
poor symptom control, overtreatment, and an
uncaring attitude.6 In an examination of a random
sample of 2074 cancer deaths, five factors were
identified that were associated with admission to
hospices: being in pain in the last year, constipation,
dependency on others to perform activities of daily
living, having a diagnosis of breast cancer and being
aged under 85 years.42 One of the conclusions from
this study was that patients were admitted to hos-
pices more by chance than by need.
While current services appear to be geared
towards the last few weeks of life, the World Health
Organization (WHO) considers that palliative care
should be initiated and developed from diagnosis
onwards.43 Despite this, there is little evidence of
current services adopting this principle. However, in
an attempt to overcome this, a palliative care team
was established in a hospital service to encourage
earlier referral for palliative care.44 Success was
gauged by earlier referral within the hospital to this
service, with patients being referred to hospice ser-
vices earlier. Despite the extended duration from
referral to death, 80% of patients were still
referred within 3 months of their death. The time
between GP referral to hospice and death was par-
ticularly short.
In England it has been suggested that 15–25% of
cancer deaths receive inpatient hospice care, and
25–65% receive input from a support team or
specialist palliative home care (e.g. Macmillan)
nurse.7,14,21,32,42,44–45 Using these figures, it has
been estimated that 700–1800 cancer patients p/M
would require support, while 400–700 p/M would
require inpatient hospice care per year.3 For
patients with noncancer progressing illness
350–1400 p/M would require a support team and
200–700 p/M would require inpatient palliative care.
In the UK, duration of terminal inpatient hospice
care varies, though the average is generally
between 2 weeks and 1 month prior to death.40,46–51
This closely relates to the recorded deterioration in
ability to perform activities of daily living 1 month
prior to death.43 In the USA, the time between
application for hospice treatment and death is
around 8 weeks.53 Using the average inpatient dura-
tion of stay and bed occupancy in the UK, it has
been estimated that 40–50 hospice beds are
required p/M.47,48 However, these figures are likely
to be an underestimate of need. Using a wide-rang-
ing consultation process, Nottingham Health
Authority identified a number of areas of need, par-
ticularly respite care, pain and other symptom con-
trol, practical and financial support and
counselling.54 In the USA unmet needs most fre-
quently revolved around inability to pay medical
bills (52%), transport costs (47%) heavy house-
keeping (42%), and other activities of daily liv-
ing.55–56 Need for bereavement services and other
emotional burdens on family and friends have also
been identified as unmet needs of some ser-
vices.35,57,58 In the UK many GPs have difficulties in
helping relatives and patients to cope with emo-
tional stress.37
02pm308.qxd 16/3/00 8:52 am Page 101
Specific patient groups
The needs of certain patient groups may be differ-
ent from the majority of patients receiving palliative
care. There is a general perception of poor uptake
of palliative care services from patients from ethnic
minority groups,59 which may be due to these ser-
vices failing to appreciate different religious or cul-
tural practices.60 Similarly, hospices that were
principally established to meet the need of patients
with cancer have until recently failed to provide ser-
vices for patients with other progressing dis-
eases.36,61 Diseases that have been highlighted are
end-stage renal failure,62 heart failure,63 stroke61
and pulmonary diseases.61 In areas where referral
of noncancer progressing terminal diseases are
encouraged, these may make up one-fifth of all
referrals.24
Neurological disease
As Table 2 indicates, deaths attributed to neuro-
logical disorders are relatively rare; however,
patients with diseases such as multiple sclerosis and
motor neurone disease have different characteris-
tics from patients suffering from cancer, and are
likely to require different services from other can-
cer and noncancer terminal illness. The key issue for
these patients is the long duration of symptoms. Pal-
liative care for these patients revolves around the
need for respite care, with only 20% of hospice
referrals for motor neurone disease required for
terminal care.28
Patients with HIV and AIDS
There are still relatively few deaths caused by
acquired immunodeficiency syndrome (AIDS) and
related problems, with only 19 deaths/M in men and
two deaths/M in women in 1995.4 In the UK as a
whole in 1994 there were 1189 diagnoses of AIDS
made and 1065 deaths from AIDS-related diseases.3
Although this number is relatively small, the needs
of this group of patients may be quite different from
those of patients with other terminal illnesses. In
general, these patients are considerably younger
than most patients receiving palliative care, while
many may be homeless, making home care prob-
lematic. AIDS is characterized by acute episodes of
infection requiring hospitalisation, followed by
periods of remission. Mean time from referral to
death has been estimated at 31 weeks.64 While pain
The level of need for palliative care 101
has not been considered a major problem, recent
work has indicated that pain experienced may be
different between the sexes, most frequently neu-
ropathic in men (36%), but muscular in women
(17%).65 Patients suffering from AIDS experience
a multitude of symptoms, particularly related to the
pulmonary and gastrointestinal systems and skin
problems.65,66 The stigma of AIDS, homophobic
attitudes, high levels of anxiety caused by lack of
effective treatment and poor symptom control, have
all been proposed as reasons for AIDS-specific ser-
vices,66 though these are considered uneconomical,
and referrals are encouraged to existing hospice ser-
vices.45 In one study 57% of patients suffering from
AIDS died in hospital, 22% at home and 21% in
hospices.64 Others may experience long-term dete-
rioration in health, which will gradually increase the
burden on the health services. In the terminal stages
of the disease the prevalence of symptoms is fre-
quently higher than for cancer patients. Symptoms
are moderate to severe in 35–70% of cases.27,64,67
Childhood diseases
Relative to the adult population, few children
require palliative care, but it must be acknowledged
that children may require different care from those
of adult patients. Deaths due to congenital abnor-
malities occur most frequently in infants under 1
year of age (Table 2). Cancer deaths are more rare
in this age group, but increase with increasing age.
Hospices for children were pioneered in the UK,
and developed in Europe and Canada68 with chil-
dren’s hospitals also providing specialist services.69
For children with terminal illness the aim has been
for the child to die at home.70–73 Most families of
children suffering from cancer take their children
home after treatment is stopped, though between
one-quarter and one-third die in hospital.69,74
Clearly, where this is the case, there is a need for
the provision of services able to deal with dying
children in their own homes.
Conclusions
The evidence surrounding the need for palliative
care is generally of poor quality and often provides
conflicting evidence. Thus, assessing the needs for
palliative care can only provide broad indications of
the levels of need required. While the epidemio-
02pm308.qxd 16/3/00 8:52 am Page 102
102 PJ Franks et al.
logical approach gives a much broader perspective
on the levels of palliative care required for popula-
tions, it has a number of limitations. Different stud-
ies have shown that symptom prevalence can vary
greatly, even between patient groups with similar
disease profiles. Moreover, prevalence of symptoms
may be of limited value in assessing the require-
ments for palliative care. Not only may the presence
of a symptom be important, but also the severity of
that symptom and the impact it is likely to have on
the patient’s level of disability and discomfort. Evi-
dence based simply on symptom prevalence will also
fail to identify the type of care required and how
effectively that care may be delivered, both in terms
of cost effectiveness and user and professional sat-
isfaction with care.75
The evaluation of current service usage is also
limited in its approach, since it can only evaluate
patients who are receiving care within that service.
For a complete evaluation of services this informa-
tion must be supplemented by patients who fall out-
side the service in order to determine whether they
are receiving the appropriate care for their disease.
There is evidence that patients suffering from non-
cancer terminal illness may require at least as much
palliative care as patients dying from cancer, yet
proportionally fewer resources are used to alleviate
suffering in these patients. As Addington-Hall et
al.42 noted, admissions to hospices appear to be
more by chance than by the needs of patients. Clear-
ly, services must be rationalized to deliver appro-
priate care to all patients. Evaluations of services
must not only look at specialist palliative care ser-
vices, but also the delivery of palliative care within
the general hospital and community services where
the majority of patients are receiving care.
The palliative care needs of society are likely to
change in the future, though the exact nature of these
changes is difficult to gauge accurately. On one hand,
the incidence of chronic diseases such as cancer
appears to be reducing, with advances in prevention,
early detection and more effective treatments. How-
ever, this is likely to be counterbalanced by earlier
intervention of palliative care services, as advocated
by WHO recommendations. Further attempts must
be made to assess unmet need and to intervene
where necessary. In particular, symptom control must
be improved in patients suffering from the terminal
stages of all disease, to alleviate both physical and
psychological distress, while providing a place to die
that patients and relatives feel is appropriate for that
individual. This will require detailed analysis of the
requirements of users to be assured that future plan-
ning can achieve the optimum care for patients.
References
1 Salisbury C, Bosanquet N, Killbery E et al. The
impact of different models of palliative care on
patients’ quality of life: a systematic review. Palliat
Med 1999; 13: 3–17.
2 Last JM ed. A dictionary of epidemiology, 3rd
edition. Oxford: Oxford University Press. 1995: 111.
3 Higginson IJ. Health care needs assessment: palliative
and terminal care. Winchester: Wessex Institute of
Public Health Medicine, 1995.
4 Office for National Statistics. 1995 mortality statistics:
cause. England and Wales, series DH2, No. 22.
London: Office for National Statistics (OHMS), 1997:
176–79.
5 Gau DW, Diehl AK. Disagreement among general
practitioners regarding cause of death. Br Med J
1982; 284: 239–41.
6 Wilkes E. Dying now. Lancet 1984; i: 950–52.
7 Cartwright A. Changes in life and care in the year
before death, 1969–87. J Public Health Med 1991; 13:
81–87.
8 Hardy JR, Turner R, Saunders M, Ahern R.
Prediction of survival in a hospital based continuing
care unit. Eur J Cancer 1994; 30: 284–88.
9 Boyd KJ. Short term admissions to a hospice. Palliat
Med 1993; 7: 289–94.
10 Myers KG, Trotman IF. Palliative care needs in a
district general hospital: a survey of patients with
cancer. Eur J Cancer Care 1995; 5: 116–21.
11 Zylicz Z. The Netherlands: status of cancer pain and
palliative care. J Pain Symptom Manage 1996; 12:
136–38.
12 Reuben DB, Mor V, Hiris J. Clinical symptoms and
length of survival in patients with terminal cancer.
Arch Intern Med 1988; 148: 1586–91.
13 Coyle N, Adelhardt J, Foley KM, Portenoy RK.
Character of terminal illness in the advanced cancer
patient: pain and other symptoms during the last four
weeks of life. J Pain Symptom Manage 1990; 5: 83–93.
14 Addington-Hall J, MacDonald LD, Anderson HR,
Freeling P. Dying from cancer: the views of bereaved
family and friends about the experiences of the
terminally ill cancer patients. Palliat Med 1991; 5:
207–14.
15 Dudgeon DJ, Raubertas RF, Doerner K, O’Connor
T, Tobin M, Rosenthal SN. When does palliative
care begin? A needs assessment of cancer with
recurrent disease. J Palliat Care 1995; 11(1): 5–9.
02pm308.qxd 16/3/00 8:52 am Page 103
16 Hopwood P, Stephens RJ. Symptoms at presentation
for treatment in patients with lung cancer:
implications for the evaluation of palliative treatment.
The Medical Research Council (MRC) Lung Cancer
Working Party. Br J Cancer 1995; 71: 633–36.
17 Talmi YP, Bercovici M, Waller A, Horowitz Z,
Adunski A, Kronenberg J. Home and in-patient
hospice care of terminal head and neck cancer. J
Palliat Care 1997; 13(1): 9–14.
18 Conill C, Verger E, Henriquez I et al. Symptom
prevalence in the last week of life. J Pain Symptom
Manage 1997; 14: 328–31.
19 Field D, McGaughey J. An evaluation of palliative
care services for cancer patients in the Southern
Health and Social Services Board of Northern
Ireland. Palliat Med 1998; 12: 83–97.
20 Addington-Hall J McCarthy M. Dying from cancer:
results of a national population based investigation.
Palliat Med 1995; 9: 295–305.
21 Seale C. A comparison of hospice and conventional
care. Soc Sci Med 1991; 32: 147–52.
22 Stjernsward J. WHO cancer pain relief programme.
Cancer Surv 1988; 7: 195–208.
23 Seamark DA, Williams S, Hall M, Lawrence CJ,
Gilbert J. Palliative terminal cancer care in
community hospitals and a hospice: a comparitive
study. Br J Gen Pract 1998; 48: 1312–36.
24 Lickiss N, Wiltshire J, Glare PA, Chyne RW. Central
Sydney Palliative Care Service: potential and
limitations of an integrated palliative care service
based in a metropolitan teaching hospital. Ann Acad
Med Singapore 1994; 23: 264–70.
25 Cowley S. Supporting dying people Nurs Times
1993; 89: 52–55.
26 Hockley JM, Dunlop R, Davies RJ. Survey of
distressing symptoms in dying patients and their
families in hospital and the response to a symptom
control team. Br Med J 1988; 296: 1715–17.
27 Welch JM. Symptoms of HIV disease. Palliat Med
1991; 5: 46–51.
28 Hicks F, Corcoran G. Should hospices offer respite
admissions to patients with motor neurone disease?
Palliat Med 1993; 7: 145–50.
29 McCarthy M, Lay M, Addington-Hall J. Dying from
heart disease. J R Coll Physicians (Lond) 1996; 30:
325–28.
30 McCarthy M, Addington-Hall J, Altmann D. The
experience of dying with dementia: a retrospective
study. Int J Geriatr Psychiatry 1997; 12: 404–409.
31 Hopwood P, Howell A, Maguire P. Psychiatric
morbidity in patients with advanced cancer of the
breast: prevalence measured by two self rating
questionnaires. Br J Cancer 1991; 64: 349–52.
32 Higginson IJ, Wade AM, McCarthy M.
Effectiveness of two palliative support teams. J
Public Health Med 1992; 14: 50–56.
The level of need for palliative care 103
33 Hinton J. Which patients with terminal cancer are
admitted from home care? Palliat Med 1994; 8:
197–210.
34 Nash A. Reasons for referral to a palliative nursing
team. J Adv Nurs 1993; 18: 707–13.
35 Neale B, Clark D. Informal palliative care: a review
of research on needs and services. J Cancer Care
1992; 1: 193–98.
36 Scott SN. Report of the task force on care of the dying
and the bereaved in Argyll and Clyde Health Board.
Paisley: Argyll and Clyde Health Board, 1995.
37 Seamark DA, Lawrence C, Gilbert J. Characteristics
of referrals to an inpatient hospice and a survey of
general practitioner perceptions of palliative care.
J R Soc Med 1996; 89: 79–84.
38 Parkes CM. Terminal care: home, hospital or
hospice? Lancet 1985; i: 155–57.
39 Bruera E, Kuehn N, Emery B, Macmillan K,
Hanson J. Social and demographic characteristics of
patients admitted to a palliative care unit. J Palliat
Care 1990; 6(4): 16–20.
40 Cartwright A. Dying when you’re old. Age Ageing
1993; 22: 425–30.
41 Komesaroff PA, Moss CK, Fox RM. Patients’
socioeconomic background: influence on selection
of inpatient or domiciliary hospice terminal care
programmes. Med J Aust 1989; 151: 199–201.
42 Addington-Hall J, Altmann D, McCarthy M.
Which terminally ill cancer patients receive
hospice in-patient care? Soc Sci Med 1998; 46:
1011–16.
43 Stjernsward J, Colleau SM, Ventafridda V. The
World Health Organisation Cancer Pain and
Palliative Care Program. Past present and future.
J Pain Symptom Manage 1996; 12: 65–72.
44 Bennett M, Corcoran G. The impact on community
palliative care services of a hospital palliative care
team. Palliat Med 1994; 8: 36–43.
45 Higginson IJ. Palliative care: a review of past
changes and future trends. J Public Health Med
1993; 15: 3–8.
46 Rosenthal MA, Gebski VJ, Kefford RF, Stuart
Harris RC. Prediction of life expectancy in hospice
patients: identification of novel prognostic factors.
Palliat Med 1993; 7: 199–204.
47 Rogers WF, Clarke C, Whitfield AGW. Hospice
planning in North Staffordshire. J R Coll Physicians
(Lond) 1981; 15: 25–27.
48 Frankel S, Kammerling M. Assessing the need for
hospice beds. Health Trends 1990; 22: 83–86.
49 James ML, Gebski VJ, Gunz FW. The need for
palliative care services in a general hospital. Med J
Aust 1985; 142: 448–49.
50 Chan A, Woodruff RK. Palliative care in a general
teaching hospital. 1. Assessment of needs. Med J
Aust 1991; 155: 597–99.
02pm308.qxd 16/3/00 8:52 am Page 104
104 PJ Franks et al.
51 Severs MP, Wilkins PSW. A hospital palliative care
ward for elderly people. Age Ageing 1991; 20:
361–65.
52 Morris JN, Sherwood S. Quality of life of cancer
patients at different stages of the disease trajectory.
J Chron Dis 1987; 40: 545–53.
53 Forster LE, Lynn J. Predicting life span for
applicants to inpatient hospice. Arch Intern Med
1988; 148: 2540–43.
54 Nicholas A, Frankenberg R. ‘Towards a strategy for
palliative care’, Department of Public Health
Medicine, Nottingham 1992, cited in Clark et al.
‘Contracting for palliative care’. Soc Sci Med 1992;
40; 1193–202.
55 Siegel K, Mesagno FP, Karus DG, Christ G.
Reducing the prevalence of unmet needs for
concrete services of patients with cancer. Evaluation
of a computerized telephone outreach system.
Cancer 1992; 69: 1873–83.
56 Houts PS, Yasko JM, Harvey HA et al. Unmet needs
for persons with cancer in Pennsylvania during the
period of terminal care. Cancer 1988; 62: 627–34.
57 Payne S, Relf M. The assessment of need for
bereavement follow up in palliative and hospice
care. Palliat Med 1994; 8: 291–97.
58 Peace G, O’Keefe C, Faulkner A, Clark J.
Childhood cancer: psychosocial needs: are they
being met? J Cancer Care 1992; 1: 3–13.
59 Hill D, Penso D. Opening doors: improving access to
hospice and specialist palliative care services by
members of the black and ethnic minority
communities, Occasional paper 7. London: National
Council for Hospice and Specialist Palliative Care
Services, 1995.
60 House N. Helping to reach an understanding.
Palliative care for people from ethnic minority
groups. Prof Nurse 1993; 8: 329–32.
61 Wilson IM, Bunting JS, Curnow RN, Knock J. The
need for inpatient palliative care facilities for
noncancer patients in the Thames Valley. Palliat
Med 1995; 9: 13–18.
62 Andrews PA. Palliative care for patients with
terminal renal failure [letter]. Lancet 1995; 346:
506–507.
63 Beattie JM. Palliative care in terminal cardiac
failure: small numbers of patients with terminal
cardiac failure may make considerable demands on
services. Br Med J 1995; 310: 1411.
64 Butters E, Higginson I, George R, Smits A,
McCarthy M. Assessing the symptoms, anxiety and
practical needs of HIV/AIDS patients receiving
palliative care. Qual Life Res 1992; 1: 47–51.
65 Kelleher P, Cox S, McKeogh M. HIV infection: the
spectrum of symptoms and disease in male and
female patients attending a London Hospice. Palliat
Med 1997; 11: 152–58.
66 Schofferman J. Hospice care for the patient with
AIDS. Hosp J 1987; 3: 51–74.
67 Cole RM. Medical aspects of care for the person
with advanced aquired immune deficiency syndrome
(AIDS): a palliative perspective. Palliat Med 1991; 5:
96–111.
68 Davies B. Assessment of need for a childrens
hospice program. Death Stud 1996; 20: 247–68.
69 Goldman A, Beardsmore S, Hunt J. Palliative care
for children with cancer-home, hospital or hospice?
Arch Dis Child 1990; 65: 641–43.
70 Bennett P. The dying child. A care team for
terminally ill children. Nurs Times 1984; 80: 26–27.
71 Singleton R. Palliative home care program for
terminally ill children. Leadersh Health Serv 1992; 1:
21–27.
72 Chambers TL. Hospices for children? Br Med J
1987; 294: 1309–10.
73 Martinson IM. Dying children at home. Nurs Times
1980; 76: 129–32.
74 Kohler JA, Radford M. Terminal care for children
dying of cancer: quantity and quality of life. Br Med
J 1985; 291: 115–16.
75 Wilkinson EK, Salisbury C, Bosanquet N, Franks PJ,
Kite S, Lorentzon M, Naysmith A. Patient and carer
preference for, and satisfaction with, specialist
models of palliative care: a systematic literature
review. Palliat Med 1999; 13: 197–216.