386 Journal of Pain and Symptom Management Vol. 46 No.

3 September 2013

Original Article

Traditional Healers’ Views of the Required

Processes for a ‘‘Good Death’’ Among Xhosa
Patients Pre- and Post-Death
Nicola Graham, MBChB, DipHIVManSA, MPhilPallMed,
Liz Gwyther, MBChB, FCFP, MSc,
Thozama Tiso, Dip Nursing Science, Adv Cert HIV Management,
and Richard Harding, BSc (Joint Hons), MSc, PhD, DipSW
Palliative Care Programme (N.G., L.G., R.H.), Division of Family Medicine, School of Public Health
and Family Medicine, Faculty of Health Sciences, University of Cape Town, Rondebosch; Eastern Cape
Department of Health (T.T.), Bhisho, South Africa; and Department of Palliative Care Policy &
Rehabilitation (R.H.), Cicely Saunders Institute, King’s College London, United Kingdom

Abstract
Context. South Africa faces enormous HIV-related mortality and increasing
cancer incidence. Traditional healers are the preferred source of advice and care
in Africa, and this is true for the large Xhosa ethnic group.
Objectives. To provide more appropriate multidimensional, culturally suitable
care at the end of life; this study aimed to identify the care needs and cultural
practices of Xhosa patients and families at the end of life, from the perspective of
traditional healers.
Methods. The study design was qualitative and cross-sectional. The research took
place in a 300 km radius around East London, Eastern Cape, South Africa.
Interviewees were Xhosa individuals who were recognized by their communities as
traditional healers. Data from two focus groups and eight individual interviews were
analyzed, using an inductive thematic approach.
Results. Data were elicited around the facilitation of a good death in terms of
care needs before death and important rituals after death. Care needs before
death focused on relief of psychosocial suffering; the importance of the spoken
word at the deathbed; and the importance of a relationship and spiritual
connection at the end of life. There were broad similarities across the rituals
described after death, but these rituals were recognized to differ according to
family customs or the dying person’s wishes.
Conclusion. Awareness of potential needs at the end of life can assist clinicians
to understand the choices of their patients and develop effective end-of-life care
plans that improve the outcomes for patients and families. J Pain Symptom
Manage 2013;46:386e394. Ó 2013 U.S. Cancer Pain Relief Committee. Published by
Elsevier Inc. All rights reserved.

Address correspondence to: Richard Harding, BSc Road, London, United Kingdom SE5 9PJ. E-mail:
(Joint Hons), MSc, PhD, DipSW, Department of Pal- richard.harding@kcl.ac.uk
liative Care, Policy & Rehabilitation, Cicely Saun- Accepted for publication: August 29, 2012.
ders Institute, King’s College London, Bessemer

Ó 2013 U.S. Cancer Pain Relief Committee. 0885-3924/$ - see front matter
Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpainsymman.2012.08.005
Vol. 46 No. 3 September 2013 Traditional Healers at the End of Life
Key Words
End of life, traditional healer, culture, traditional medicine, Africa, death
Introduction
During 2008, there were 1251 daily deaths
from progressive illness in South Africa,1e3
placing a great strain on clinicians who require
the skills to manage highly prevalent multidi-
mensional problems.4e6 The vast majority of
evidence regarding end-of-life care has been
generated in non-African populations,7
although recent advances in the development
of outcome measurement tools have meant
that it is possible to measure and improve the
outcomes that matter to African patients.8e10
Locally generated data are essential because
each person experiences death through their
own context, and patient needs may differ
from those of their health practitioner.11
Therefore, clinicians have a responsibility to
understand the health philosophies and
systems that their patients find helpful12 to
provide patient-centered care that meets the
proposed standards and recommendations
for culturally appropriate spiritual palliative
care provision in Africa.13,14
The World Health Organization estimates
that 80% of African patients make use of
both Western biomedicine and traditional
healing practitioners.15 The use of traditional
healing practices is closely linked to traditional
African beliefs and customs.16 Traditional
healers also are described as the ‘‘guardians
of cultural norms and social order, and bas-
tions of cultural practice.’’17 Consequently,
traditional healers can provide insights into
death processes and meanings to inform rele-
vant clinical practice. There is a national18e22
and international15,23 trend toward collabora-
tion with traditional health systems.
The amaXhosa are an ethnic group constitut-
ing 18% of the national population, comprising
many large family clans and lineages. The amaX-
hosa are concentrated in South Africa’s Eastern
Cape, with 7.9 million speaking isiXhosa as their
home language.24 It is necessary to generate
evidence on the traditional perspectives of
health behaviors and beliefs at the end of life
that can inform palliative care practitioners
who, therefore, may provide more appropriate
patient-centered care. This study identifies the
387
care needs and cultural practices of Xhosa
patients and their families at the end of life to
achieve a good death both pre- and post-death,
from the perspective of traditional healers.
Methods
Study Design and Setting
This was a qualitative cross-sectional study
across the urban, peri-urban, and rural Eastern
Cape of South Africa, and was approved by the
Research Ethics Committee of the University
of Cape Town (Ref:079/2008). The study
comprised individual interviews and focus
groups. Given the exploratory nature of the
research and the ‘‘outsider’’ status of the lead
researcher, a qualitative-interpretive research
orientation was used, aiming to understand
the actions of individuals and the meaning
they attribute to those actions.25 The recogni-
tion that social phenomena are continuously
being constructed and revised through social
interactions positions this research within a con-
structionist paradigm, that is, the findings of
the research cannot be considered definitive
but are put forward as a specific view supported
with data. Therefore, this study aimed to elicit
sensitizing rather than definitive data.
Sampling and Recruitment
Using snowball sampling through personal
and professional contacts, eight key informants
(including a security guard, a housekeeper
from a local hospital, a home-based carer, and
a technician from a local nongovernmental or-
ganization) acted in their capacity as commu-
nity members to introduce the researcher to
the healers across a 300 km distance from the
main provincial town. Traditional healers were
invited to participate if they practiced within
their community from a Xhosa cultural back-
ground. Both inyangas/iixwele (herbalists) and
sangomas/iiqirha (diviners) were included, as
well as those who also may incorporate
Western-style health care. Traditional birth atten-
dants and traditional surgeons were excluded.
388 Graham et al.
Data Collection
Interviews were requested with nominated
traditional healers, and signed consent
obtained. Approximately 25% of those ap-
proached declined, with reasons volunteered
as concern that the study wanted to know the
secrets of their practice; that they did not feel
equipped or able to answer on behalf of tradi-
tional healers; and time pressures. The topic
guide was iteratively reviewed and refined in
light of the emerging data. Two focus group in-
terviews and eight individual interviews were
conducted. One focus group contained rural
participants and the other contained urban
participants. The iterative process of individual
and group interviews continuously developed
the semi-structured interview guide. An initial
series of individual interviews informed the sub-
sequent focus group discussions, and findings
from these expanded discussion groups were
then pursued in greater detail in the final indi-
vidual interviews. The lead author conducted
the interviews with the translator present. All
interviews were audio-recorded and additional
field notes documenting observations and
impressions were collected.
Translation
It was anticipated that the interviews would be
translated fully between isiXhosa and English.
However, it was apparent that the translator
wielded power in the interview setting, and the
interviewees often preferred to make themselves
understood in English. Most interviews were con-
ducted in a mixture of English and isiXhosa and
participants made variable use of the translator
present, sometimes extensively, sometimes only
to check English meanings of isiXhosa words.
Full interview recordings were transcribed, in-
cluding isiXhosa and English translations, and
these transcripts were verified by another inde-
pendent first-language isiXhosa speaker.
Analysis
The lead researcher’s position as a young,
white, female, Western-trained doctor had to
be considered in contacts with interviewees,
and in analysis of the interview data.26 These
reflections were recorded and were used to
inform the analysis of the results.27
The applied nature of the research made an
inductive approach within a functional
Vol. 46 No. 3 September 2013
framework most appropriate.28 After familiar-
ization with the transcripts, alongside the con-
text of the reflexive journal, the data were
initially coded, and subsequently the general
categories were transformed and fragmented
to better fit the data. These more detailed
codes were clustered into nodes and reviewed
by a colleague, and out of these nodes the
themes for analysis emerged inductively.
Results
The two focus groups contained seven and
six members, respectively, and eight individual
interviews were conducted.
Sample Characteristics
The age of the 13 focus group attendees
ranged from 38 to 64 years, eight were female,
one group was conducted with rural respon-
dents and one with peri-urban respondents.
Their years in practice of traditional healing
ranged from six months to 40 years.
The age of the eight individual interviewees
ranged from 30 to 80 years, and five were fe-
male. Two interviewees practiced in an urban
area, four in a peri-urban area, and two in a rural
area. Additional occupations included subsis-
tence farming, businessperson, management
in an HIV nongovernmental organization,
a nursing sister, and other health care setting
roles such as volunteer and laboratory assistant.
Main Findings
Before describing the main coding frame,
the following central finding gives context to
the data.
The Interaction Between Traditional Healing and
Death. Across the sample, there was a de-
scribed reluctance in managing death directly
in their professional capacity as traditional
healers. There were complex reasons given
for this reluctance, including the fear of blame
by the family and the negative impact of death
on their practice, described as ‘‘that dark
cloud,’’ and the difficulty posed by rituals nec-
essary for a traditional healer to ‘‘wash away’’
the contagious nature of death. Although
they were able to describe the cultural practice
and processes at the end of life, and were pres-
ent, managing the ‘‘good death’’ was not
Vol. 46 No. 3 September 2013 Traditional Healers at the End of Life
perceived as their role. The coding framework
is shown in Fig. 1.
The four data codes under the main head-
ing of ‘‘before death’’ are described first. The
‘‘good death’’ was defined mostly in terms of
relief of psychosocial suffering. Although phys-
ical symptoms were mentioned, there was
a more significant focus on the psychosocial
and spiritual aspects of suffering, and the im-
portance of addressing these to die well.
Somebody should also be given a chance to
talk about things. They may say, ‘‘I wish to be
discharged so that I can go and talk with my
ancestors.’’ There are various places where
the ancestors are, and a person may want
to go there. He should be allowed to do
that.dParticipant 8 (55-year-old woman, urban
setting, formally trained with 13 years’ experience)
If he has been believing in something, you
should call that, you should give it because
he is facing death. If he says, ‘‘I want to
have the medication that I believe in from
my father,’’ or whatever, somebody should
be able to have that, because he would feel
comfortable at death now that that thing
has been done.dParticipant 8 (55-year-old
woman, urban setting, formally trained with
13 years’ experience)
There were descriptions of meeting specific
needs of the individuals as death approached,
Fig. 1. Coding frame.
389
which ranged from being given a favorite food
to being discharged to make a spiritual pil-
grimage. These needs vary across individuals,
and the person should be asked what will sat-
isfy them and allow them a good death.
.as a person comes to die, for each person it
is different. You have to listen to him.dPar-
ticipant 8 (55-year-old woman, urban setting, for-
mally trained with 13 years’ experience)
Death was described as a collective affair,
with the presence of the family at the deathbed
a significant factor in the quality of the death.
But if the family is not around, you will see
them really struggling. They can be very ill,
they can be critical, and you can think they
will die that day, but tomorrow they are still
struggling, waiting for their family to come.
After having seen the family, then they are
at peace with themselves and with the
family.dParticipant 7 (47-year-old woman, ur-
ban setting, formally trained with six months’
experience)
Their family gathering at the deathbed is
not only for the comfort of the dying person
but also for the chance to restore relation-
ships, express wishes for the family, and give
a verbal willd‘‘umnyolelo.’’
There was much emphasis placed on the spo-
ken word at the time of death in our sample.
390 Graham et al.
But for us, we can’t do anything, only talk,
only the talk that is going to take her away.d
Participant 1 (66-year-old woman, semi-rural set-
ting, formally trained with 26 years’ experience)
So you can see that the person is going to
die? I think you have to talk to that person
so that they can die peacefully.dParticipant
3 (42-year-old woman, peri-urban setting, trained
only by ancestors, with five years’ experience)
The verbal will, umnyolelo, and the taboos
placed on what should and should not be
said were prominent, and highlighted the
importance of speech at this time.
We are a culture of verbalization.dPartici-
pant 1 of Focus Group 1 (female healers, peri-
urban setting, formally trained)
There were different views about whether
talk at the deathbed should include truth-
telling about the imminent death or not.
Some healers expressed that honesty about
the impending death was a release.
Like maybe, the whole day you have watched
this person and seen that they are going to
godbut they can’t go. Then you must just
come and talk to her and say, you can go
now. You can say, this is the end of your
life, so you must go straightaway! And every-
one must be saying the same thing. You
can’t have this one over here saying, ‘‘Oh,
I’m so hurt, I don’t want her to go.’’ That
is wrong. You must only talk honestly. And
then you will see that person will go.dParti-
cipant 1 (66-year-old woman, semi-rural setting,
formally trained with 26 years’ experience)
Other healers thought that truth-telling was
unkind and would rob an individual of
a chance for a peaceful death.
You would never talk like that. You always give
hope. You always give hope. Because you want
them to die peacefully, and that might esca-
late the condition, which you don’t want. To
die peacefully, you must always give them
hope.dParticipant 2 of Focus Group 1 (female
healers, peri-urban setting, formally trained)
I don’t like to tell somebody that they are go-
ing to die because it’s not nice.dParticipant
3 (42-year-old woman, peri-urban setting, trained
only by ancestors, with five years’ experience)
Vol. 46 No. 3 September 2013
There was a recognition that the dying per-
son’s words carry much weight for continuity
after death. The contract with the dying has
gravity, and is binding because the person
will soon become an ancestor and will act in
the lives of those left behind.
If somebody dies and says they want some-
thing, you must do that thing, otherwise
you will have bad luck. Because it is her
wish. And if you don’t do that wish, she
will be cross with you. And her spirit is
now an ancestor.dParticipant 3 (42-year-old
woman, peri-urban setting, trained only by ances-
tors, with five years’ experience)
The five codes explaining the processes after
death are now described.
And after the person has died, you should
close the mouth, close the eyes, and stretch
out the arms and legs. You can also put a bit
of salt in the mouth to make the mouth soft
to close it if they died with their mouth
open.dParticipant 2 (woman older than 80
yearsdbirth date unknown, semi-rural setting,
formally trained with more than 40 years’
experience)
Rituals after death were recognized to differ in
structure according to different family cultures.
I would say that the culture in different fam-
ilies can be very different. Because if you be-
long to this clan, you might slaughter
something. But if I belong to another clan,
then that can be very wrong for that clan.
They do differ. The way they do their own
things. What is right for the one clan can
be poison for another.dParticipant 8 (55-
year-old woman, urban setting, formally trained
with 13 years’ experience)
As mentioned previously, the specific re-
quests of the individual also influence the
structure of the rituals.
If you were there and the person said, ‘‘Now
I can see that I am going to die. Don’t
slaughter anything, just give the people
maize only.’’ And you must do it the way
he said. And you must tell the others that I
was there with him, and he said to me that
he knew he was going to die. He said that
we must not slaughter anything, we must
Vol. 46 No. 3 September 2013 Traditional Healers at the End of Life
do this and this and this.dParticipant 2 of
Focus Group 2 (mixed male and female healers,
rural setting, method of training unknown)
Part of the purpose of the rituals after death
is to ease the passage of the spirit into the af-
terlife and to join the other ancestors. Cleans-
ing was described as an important aspect of
postdeath ritual. A form of cleansing needs
to take place to wash away the contagiousness
of death.
Yes, the day after the funeral, there are some
things that are supposed to be done. But it de-
pends on that family exactly what happens.
But something should be done to wash away
all that darkness, or else it becomes part of
the family. Especially in our culture.dPartici-
pant 8 (55-year-old woman, urban setting, formally
trained with 13 years’ experience)
If a person dies in a home, we use uqomboti,
the African beer, to cleanse the home. There
was darkness in this house. Because the cof-
fin came in here, there was a death of a per-
son, there was darkness in the house so we
have to do something. Other people are
cleansing the spade because in our culture,
we have to dig for the grave. So we do this
cleansing ceremony by making this African
beer. And we are cleansing the house be-
cause there is that darkness.dParticipant 3
(42-year-old woman, peri-urban setting, trained
only by ancestors, with 5 years’ experience)
Between the time of death of the person and
the time that the necessary rituals have been
performed, there is a period of hibernation
for the family.
When you have lost someone, then there is
a funeral, and everyone must be there. Say
this is on Saturday. Then on Sunday, you
take all the clothes and you must wash
them. Then we are going to wait for 5
daysdwe are here, all of us, and we are going
to wait for 5 days. Then we are going to have
that ceremony on Saturday. That ceremony
means that now everyone can go back to their
places, back to school, back to work, back to
their houses. That is the culturedexcept
with those people who don’t believe. But
everyone here agrees.dParticipant 1 of Focus
Group 2, (mixed male and female healers, rural
setting, method of training unknown)
391
And also we don’t go and visit other people’s
ceremonies or works until we have done this
thing, so we are respecting that person who
left us. So we don’t even go next door and
ask them for salt, or something to eat.dPar-
ticipant 3 (42-year-old woman, peri-urban setting,
trained only by ancestors, with 5 years’ experience)
The timing of this ceremony of ritual accom-
paniment to the ancestors varied, but most
healers described a ritual to release the spirit
to be with the other ancestors.
And then maybe a year or so after, if it is the
father of the family, then there must be
a cow that is slaughtered as a sort of send
off ritual for the person.dParticipant 7 (47-
year-old woman, urban setting, formally trained
with six months’ experience)
Like my grandfather, his cow which was
slaughtereddit was an ox which was slaugh-
tered on the funeral day. We call it siyamkha-
pha utatumkhulu. In English it means ‘‘we are
accompanying him to his ancestors.’’dParti-
cipant 3 (42-year-old woman, peri-urban setting,
trained only by ancestors, with 5 years’ experience)
Following the release ceremony, also at a var-
iable time, another ritual is performed to wel-
come the spirit home, to protect the
household and guard the family.
So one will find that you slaughter another
cow as a welcoming of the spirit of that per-
son, so that spirit must be able to come in
and guard over evil spirits or bad luck. I
don’t know how to explain it, ‘‘uvuyisa.’’
But it isn’t the person coming home physi-
cally, it’s the spirit coming home.dPartici-
pant 7 (47-year-old woman, urban setting,
formally trained with six months’ experience)
In summary, the traditional healers de-
scribed practices necessary before and after
death that were necessary for a good death. Be-
fore death, they placed focus on the relief of
psychosocial suffering, importance of the spo-
ken word at the deathbed, and the value
placed on relationship and community with
the living, which continues in a spiritual way
after death. After death, the healers described
the importance of the post-death ritual. Al-
though the specifics of the rituals differed
from clan to clan, and across families, there
392 Graham et al.
were common elements of cleansing the
home, sequestration of the family, rituals to
release the spirit, and then to call the spirit
back home. Not adhering to the clan rituals,
or the wishes of the deceased, were described
to have considerable consequences for those
who remain alive.
Discussion
This study is the first to identify cultural
practices and beliefs before and after death
for African Xhosa patients. The finding that
traditional healers are reluctant to deal with
death in their professional capacity corrobo-
rated the findings of a previous smaller study
of six sub-Saharan traditional healers.28 The
family or biomedical establishments were iden-
tified as the units of care at the end of life,
although patients did consult the traditional
healers.
‘‘The struggle,’’ or suffering, was recognized
as an obstacle to a good deathdlargely de-
scribed in terms of emotional or spiritual suf-
fering. Often the issues were unresolved
relationships, a sense of loneliness, and a need
for spiritual connectedness. The clinical impli-
cation of this is to recognize the importance of
unspoken psycho-socio-spiritual needs of dying
patients. This supports data from African pa-
tients at the end of life who, when determining
quality of life, placed greater importance on
their spiritual well-being than on other
domains.29
The African philosophy of ubuntu is ex-
plained through the translation of umuntu ngu-
muntu ngabantuda person is only a person
through his interaction with other people.30
Consequently, if health care workers are able
to facilitate the gathering of family at the
deathbed, they may provide a meaningful
opportunity for restoration of relationships,
expression of wishes, and peaceful acceptance
of death.
The concept that a dying person will be-
come an ancestral spirit has an impact on
death transactions, with weight placed on the
need for continuity as the spirit continues to
act in the world beyond death. Family mem-
bers respect those demands about what should
happen after death, from respect for the per-
son and because not following through may
Vol. 46 No. 3 September 2013
have consequences. Here there is a common
theme of the dying person wanting their family
members to honor both specific rituals and
their decisions for the family after death.
The traditional belief that the dead continue
to live, unseen by the living, affects the choice of
language used to describe death.30 The general
isiXhosa and isiZulu words for death are
avoided in relation to people and are only
used for other forms of life. This reflects the
concept that people do not die but transition
from living to ancestor. Euphemism, therefore,
may not only represent a form of kind speech
but may reflect the deeper underlying philoso-
phy of the afterlife. For a clinician caring for
patients at the end of life, it is essential to have
some understanding of the depth and complex-
ity that words about death can hold.
A sense of spiritual connectedness underlies
the importance of being home in a spiritual
sense. Connecting with one’s roots, the place
of the family and of the ancestors, and the
place where your spirit will come out were all
cited as strong motivators not to die in an insti-
tutional setting but in a setting rich with rela-
tionship. This is similar to the good death
described by people of Kwahu-Tafo, Ghana.31
There was much emphasis placed on the
spoken word. The verbal will (umnyolelo) and
taboos on what should and should not be
said, highlighted the importance of speech,
and contrast with the British ‘‘nonexpressive
stereotype’’ of watchfulness, silence and silent
weeping described by Woodhouse.32 Group 1
described this emphasis on the spoken word
by describing themselves as a culture of
verbalization.
Sub-Saharan Africa has adapted palliative
care into its systems and cultures. Our data
offer new knowledge to truly deliver multidi-
mensional care that addresses the physical,
psychological, social, and spiritual needs of
patients and families.
There are a number of limitations to our
study. Given the snowball recruitment method,
there maybe sampling bias and only interested
healers were interviewed; however, the major-
ity of those contacted had no previous connec-
tion to the researcher that may have biased
them toward an interest in palliative or end-
of-life care. Because of access issues, it is diffi-
cult to sample this population in other ways.
We recognize that focus groups may introduce
Vol. 46 No. 3 September 2013 Traditional Healers at the End of Life
a social desirability bias; therefore, the use of
individual interviews to expand the themes
and responses was included. The interviewees’
use of the translator was not consistent across
the interviews, and this resulted in differences
for the translation process from isiXhosa to
English. It also is recognized that the culture
is dynamic and changing, and that although
the views of traditional healers may represent
the views of many people within Xhosa culture,
not all Xhosa individuals would subscribe to
the views held by traditional healers.
In light of the original findings, we propose
a number of recommendations. We propose
first that clinicians ask the important questions
around the needs and wishes of the person at
the end of life: specific requests; the degree of
information they wish to have about their
impending death; and the needs of the family
as the unit of care. Second, it should be
accepted that the choices of the individual
may differ from those of the health care prac-
titioner. Third, end-of-life care planning
should include psychosocial needs and expec-
tations of both the family and community.
Although health care providers are often
under-resourced and under pressure to pro-
vide care to a large numbers of patients, funda-
mental care in line with belief and practice at
the end of life should be provided.
Further research needs to be directed at com-
paring the findings with other cultural settings
and geographical areas. The needs of Xhosa
patients should be assessed by studying Xhosa
patients directly, and further study may help de-
fine the role of the traditional healer at the end
of life as part of the patient’s care and support
team. Training programs need to be developed
that up-skill traditional healers in end-of-life
care; that educate palliative care practitioners
about traditional health approaches; and that
facilitate increased collaboration between the
two sectors.
Disclosures and Acknowledgments
The authors have no conflict of interest to
declare, and no funding was received for this
study.
The authors would like to thank the commu-
nity members who assisted in recruitment, the
traditional healers who took part, and Kate
393
Sherry and Nicholas Graham for their insight-
ful input during the analysis and write-up
phases.
References
1. Statistics South Africa. Mortality and causes of
death in South Africa 2008: Findings from death
notification. Pretoria: Statistics South Africa, 2010.
2. Wilson D, Naidoo S, Bekker L, Cotton M,
Maartens G. Handbook of HIV medicine. Cape
Town: Oxford University Press Southern Africa, 2002.
3. Bradshaw D, Pillay-Van Wyk V, Laubscher R, et
al. Cause of death statistics for South Africa: chal-
lenges and possibilities for improvement. Burden
of Disease Research Unit, Medical Research Coun-
cil, South Africa. November 2010. Available from
http://www.mrc.ac.za/bod/cause_death_statsSA.pdf.
Accessed October 13, 2012.
4. Sepulveda C, Marlin A, Yoshida T, Ullrich A. Pal-
liative care: the World Health Organization’s global
perspective. J Pain Symptom Manage 2002;24:
91e96.
5. Harding R, Selman L, Agupio G, et al. The prev-
alence and burden of symptoms amongst cancer pa-
tients attending palliative care in two African
countries. Eur J Cancer 2011;47:51e56.
6. Harding R, Selman L, Agupio G, et al. Preva-
lence, burden and correlates of physical and psycho-
logical symptoms among HIV palliative care patients
in sub-Saharan Africa: an international multicenter
study. J Pain Symptom Manage 2012;44:1e9.
7. Harding R, Higginson IJ. Palliative care in sub-
Saharan Africa. Lancet 2005;365:1971e1977.
8. Harding R, Selman L, Agupio G, et al. Valida-
tion of a core outcome measure for palliative care
in Africa: the APCA African Palliative Outcome
Scale. Health Qual Life Outcomes 2010;8:10.
9. Selman L, Siegert RJ, Higginson IJ, et al. The
‘‘Spirit 8’’ successfully captured spiritual well-being
in African palliative care: factor and Rasch analysis.
J Clin Epidemiol 2012;65:434e443.
10. Selman L, Siegert RJ, Higginson IJ, et al. The
MVQOLI successfully captured quality of life in
African palliative care: a factor analysis. J Clin Epide-
miol 2011;64:913e924.
11. Cassidy J, Davies D. Cultural and spiritual issues
in palliative care. In: Doyle D, Hanks G, Cherny N,
Calman K, eds. Oxford textbook of palliative medi-
cine, 3rd ed. Oxford: Oxford University Press, 2006:
949e960.
12. Doyle D. Introduction. In: Doyle D, Hanks G,
Cherny N, Calman K, eds. Oxford textbook of palli-
ative medicine, 3rd ed. Oxford: Oxford University
Press, 2006:3e4.
13. African Palliative Care Association. APCA stan-
dards for providing quality palliative care across
394 Graham et al.
Africa. Available from http://www.hospicecare.com/
standards/APCA_Standards_AW.pdf: 2012. Accessed
July 2012.
14. Selman L, Harding R, Agupio G, et al. Spiritual
care recommendations for people receiving pallia-
tive care in sub-Saharan Africa. 2010. Available
from http://www.csi.kcl.ac.uk/spiritualcare.html?
searched¼spiritualþcareþrecommendations&advse
arch¼exactphrase&highlight¼ajaxSearch_highlight
þajaxSearch_highlight1. Accessed July 2012.
15. World Health Organization. Traditional medi-
cine strategy. 2002. Available from http://www.
who.int/medicinedocs/en/d/Js2297e/#Js2297e. Ac-
cessed July 2012.
16. Pew Forum on Religion and Public Life. Tolerance
and tension: Islam and Christianity in sub-Saharan
Africa. 2010. Available from http://www.pewforum.
org/executive-summary-islam-and-christianity-in-sub-
saharan-africa.aspx. Accessed August 2012.
17. Felhaber T, Mayeng I. South African primary
health care handbookdcombining Western and tra-
ditional practices. The Medical Research Council
Traditional Medicines Research Group of the Uni-
versity of Cape Town and the University of the West-
ern Cape. Cape Town: Copycat Communications,
1999.
18. Meissner O. The traditional healer as part of
the primary health care team? S Afr Med J 2004;
94:901e902.
19. Gqaleni N, Mbatha N, Mkhize T, et al. Educa-
tion and development of traditional health practi-
tioners in isiZulu to promote their collaboration
with public health care workers. Alternation 2010;
17:295e311.
20. Abdool Kariem S, Ziqubu-Page T, Arendse R.
Bridging the gap: potential for a health care part-
nership between African traditional healers and bio-
medical personnel in South Africa. S Afr Med J
1994;84:1e16.
21. Baleta A. South Africa to bring traditional
healers into mainstream medicine. Lancet 1998;
352:554.
Vol. 46 No. 3 September 2013
22. Peltzer K, Mngqundaniso N, Petros G.
A controlled study of an HIV/AIDS/STI/TB inter-
vention with traditional healers in KwaZulu-Natal,
South Africa. AIDS Behav 2006;10:683e690.
23. World Health Organization. WHO policy per-
spectives on medicine: Traditional medicinedgrow-
ing needs and potential. 2002. Available from
http://whqlibdoc.who.int/hq/2002/WHO_EDM_
2002.4.pdf. Accessed August 2012.
24. Statistics South Africa. Census 2001 key results.
2001. Available from http://www.statssa.gov.za/
census01/html/Key%20results_files/Key%20results.
pdf. Accessed August 2012.
25. Bryman A. Social research methods. Oxford:
Oxford University Press, 2001.
26. Gantley M, Tissier J. An introduction to qualita-
tive methods for health professionals. In: Carter Y,
Shaw S, Thomas C, eds. Master classes in primary
care research. London: Royal College of General
Practitioners, 1999.
27. Mays N, Pope C. Qualitative research in health
care. Assessing quality in qualitative research. BMJ
2000;320:50e52.
28. Hewson MG. Traditional healers in southern
Africa. Ann Intern Med 1998;128(12 Pt 1):
1029e1034.
29. Selman LE, Higginson IJ, Agupio G, et al. Qual-
ity of life among patients receiving palliative care in
South Africa and Uganda: a multi-centred study.
Health Qual Life Outcomes 2011;9:21.
30. Elion B, Strieman M. Clued up on culture. A
practical guide for all South Africans, 3rd ed.
Cape Town: One Life Media, 2005.
31. van der Geest S. Dying peacefully: considering
good death and bad death in Kwahu-Tafo, Ghana.
Soc Sci Med 2004;58:899e911.
32. Woodhouse J. A personal reflection on sitting at
the bedside of a dying loved one: the final hours of
life. Int J Palliat Nurs 2005;11:28e32.