Sie sind auf Seite 1von 17

Review

Effectiveness of traditional healers in treating mental


disorders: a systematic review
Gareth Nortje, Bibilola Oladeji, Oye Gureje, Soraya Seedat

Lancet Psychiatry 2016: Traditional healers form a major part of the mental health workforce worldwide. Despite this, little systematic
3: 154–70 examination has been done of their effectiveness in treating mental illness or alleviating psychological distress. In this
See Online for podcast interview Review, we aim to fill this gap, with a focus on quantitative outcomes. We searched four databases and reference lists
with Gareth Nortje
for papers that explicitly measured the effectiveness of traditional healers on mental illness and psychological distress.
Department of Psychiatry, Eligible papers were assessed for quality, and outcomes and other details were extracted with the use of a standardised
Stellenbosch University,
Cape Town, South Africa
template. 32 eligible papers from 20 countries were included. The published literature on this topic is heterogeneous
(G Nortje FCPsych [SA], and studies are generally of poor quality, although some findings emerge more consistently. Some evidence suggests
Prof S Seedat PhD); and that traditional healers can provide an effective psychosocial intervention. Their interventions might help to relieve
Department of Psychiatry,
distress and improve mild symptoms in common mental disorders such as depression and anxiety. However, little
University of Ibadan, Ibadan,
Nigeria (B Oladeji FWACP, evidence exists to suggest that they change the course of severe mental illnesses such as bipolar and psychotic
Prof O Gureje DSc) disorders. Nevertheless, qualitative changes that are captured poorly by conventional rating scales might be as
Correspondence to: important as the quantitative changes reviewed here. We conclude by outlining the challenges involved in assessing
Dr Gareth Nortje, Department of the effectiveness of traditional healers.
Psychiatry, Stellenbosch
University,Francie van Zyl
Avenue, Cape Town, 7505,
Introduction (as is occurring in Indonesia, South Africa, Bali,
South Africa Traditional healers form a major part of the global Uganda, Papua New Guinea, New Zealand, Canada, and
g.nortje@gmail.com mental health workforce. A large amount of published the USA7,12,32–35).
literature spanning the past 50 years suggests that many Despite this widespread use of traditional healers,
people in developing countries seek out traditional their effectiveness in treating mental illness has never
healers for mental health complaints, sometimes in been systematically reviewed. Effectiveness is often
addition to using conventional psychiatric services.1–14 tacitly assumed, and only a few studies explicitly attempt
Traditional healers are appealing because they share a to quantify it.36 This Review aims to systematically
common perspective with their clients, and make use of collate and analyse the published literature measuring
knowledge, beliefs, and practices indigenous to the local the effectiveness of traditional healers in improving
culture.15 In developing countries in particular, these mental health.
shared beliefs typically include spiritual and religious This task is complicated by the fact that conceptions of
models of illness causation, which affect patterns of mental illness differ greatly between cultures, and it is not
help-seeking.16 In areas where formal psychiatric services always clear that psychiatric labels and outcomes are
are scarce or unaffordable, traditional healers are universally valid. The mind–body distinction implicit in
especially well used and provide a potentially valuable mental illness is not present in many other healing
source of mental health care.9,17,18 traditions, which make little distinction between physical
In addition to populations from developing countries, and emotional illness, spiritual problems, and the obstacles
notable minority populations in high-income and of daily living.37 For example, depression is often not
developed countries continue to use their own traditional viewed as a mental illness, and the strange behaviours of
healing systems. High rates of traditional healer psychosis are often explained in spiritual terms such as
attendance for mental health problems have been possession or bewitchment.16,31 Notions of wellness are
documented in North American Indians,19–21 Chinese linked to social and spiritual harmony and balance rather
immigrants in Canada,22 Pakistanis in Britain,23 than psychopathology.31,38 Consequently, concepts of
Bangladeshis in London,24 Turkish people in Germany,25 healing and the assessment of its efficacy are much
Hispanics in the USA,26 southeast Asian refugees in the broader and more nuanced in traditional healing than in
USA,27 and Muslims in the UK.28 conventional psychiatry. Application of the standard
Since patients with serious mental illnesses in criteria to assess outcomes would miss important
developing countries seldom receive formal psychiatric qualitative changes in meaning, attitude, relationships,
treatment,29 traditional healers will continue to have a self-image, and sick role.39,40 For example, the rituals of
substantial role in mental health-care delivery.30 symbolic healing used by traditional healers might work
Although the debate regarding the legitimacy and status on many levels, from catharsis and cognitive restructuring,
of traditional healers within existing health systems is to change in family structures, restoration of community
ongoing,31 many countries acknowledge the potential identity, and social cohesion.41 That said, patients do
usefulness of traditional healers for treating mental present to traditional healers with many psychosocial
health problems, and are attempting to incorporate complaints, and their symptoms often meet criteria for
these healers into their own formal health-care systems psychiatric caseness.31,42,43 Despite a myriad of culturally

154 www.thelancet.com/psychiatry Vol 3 February 2016


Review

specific idioms of distress, the prevailing view in psychiatry meaning. However, the definition excludes healing methods
suggests that many disorders are present and recognisable that rely mainly on physical, humoral, or quasi-mechanical
in every culture,16 although standard diagnostic criteria explanations for illness, such as acupuncture, chiropractic,
might not be universally valid.43,44 Additionally, traditional herbalism, homeopathy, traditional Chinese medicine,
healers are able to identify patients with putative mental Ayurveda, Qi-gong, Reiki, and Western medicine. We also
disorders, although they might name them and explain exclude the effects of personal religiosity and spirituality,
them differently.31,45 Therefore, it is reasonable to assume so-called distant healing where the patient is not directly
that patients with psychiatric disorders similar to those involved in the intervention, and western psychotherapies
diagnosed and treated by orthodox medicine attend and that incorporate religious elements.
are treated by traditional healers. Examination of outcomes Papers were selected for inclusion by two authors (GN
of these patients from the viewpoint of conventional and BO) independently. We included papers written in
psychiatry, while acknowledging that this is only part of the English that explicitly assessed the effectiveness of a
story, is therefore potentially useful. traditional healer on mental illness or psychological
distress. Traditional healers rarely use western psychiatric
Methods diagnoses, and this absence of diagnostic rigour is reflected
Search strategy, definitions, and selection criteria in the published literature. We therefore included papers
We aimed to include papers at the intersection of that used both psychiatric diagnoses and non-specific
three broad areas: traditional healing, effectiveness, and psychological or behavioural complaints. Furthermore,
mental disorder. The broad nature of each of these since traditional healers make little distinction between
concepts necessitated the use of several terms for each mind and body, and many somatic complaints brought to
concept in the search strategy. For example, in addition healers have psychological origins,7,47 we accepted studies
to “traditional healing”, we included “religious healing”, that assessed mixed complaints as long as they also
“indigenous healing”, “diviner”, and ten other related reported a psychological outcome. Studies of traditional
terms. Similarly, we searched for “mental illness” and healing often use measures of effectiveness that are more
related terms and specific psychiatric diagnoses. We holistic and subjective than those used in conventional
searched MEDLINE, PsycArticles, Scopus, and the Social psychiatric research. These quantitative ratings, such as
Sciences Citation Index on May 25, 2014 (for full details, “perceived improvement”, “effectiveness”, or “satisfaction”,
see appendix pp 1–2). We also manually searched the are less objective than formal rating scales, but could not See Online for appendix
reference lists of included papers and contacted key be ignored48 because they are personally and socially
authors for other potentially relevant papers. When relevant outcomes that can be quantified. Therefore, to be
necessary and possible, we requested these papers from as inclusive as possible, we accepted a broad range of
authors in at least two emails sent 1 week apart. outcome measures, ranging from validated psychiatric
Existing definitions of traditional healers are inadequate rating scales, through non-validated custom rating scales,
for use as inclusion criteria in a systematic review. For to the aforementioned subjective measures. However, we
example, the key criteria in the WHO definition of excluded papers that discussed only qualitative changes
traditional medicine is that it should be “based on attributed to the healing experience, such as personal
experiences indigenous to different cultures” and usually transformation or spiritual growth. We also excluded
has a “long history”.46 However, this definition does not anecdotes and case reports.
clarify the meaning of “indigenous” or “long history”, and
includes a heterogeneous range of healing methods with Data extraction
little in common. We preferred to use a criterion based on Data were extracted from eligible papers by two authors
qualities inherent to the healing modality, and chose an independently (GN and BO) using a fixed template, and
operational definition of traditional healers as “healers who any differences were decided by consensus. The quality of
explicitly appeal to spiritual, magical or religious included studies was assessed through the use of
explanations for disease and distress”. This definition six operationalised criteria: adequacy of controls, sampling
encapsulates a core feature underlying the meanings and methods, diagnostic rigour, follow-up procedures,
rituals of the healing modalities of interest, and was derived outcome assessment method, and reporting of results (see
by the authors after reading the published literature. appendix p 3). When samples included patients with both
Notably, healing modalities that fit this criterion tend to be physical and psychological complaints, we reported on the
very culturally specific and have a long history, as per the outcome for the psychological complaints only, or we
WHO definitions. This definition includes those healing reported the psychological outcomes for the whole sample.
practices that typically make use of sacred rituals, We anticipated that too much heterogeneity would exist
ceremonies, talismans, divination, and prayer during between studies to allow for any statistical generalisations.
diagnosis or treatment. Traditional healers who use physical Since the effectiveness of traditional healing might depend
treatments such as herbs or massage to complement on the severity of the complaint, we report the main
spiritual treatments are also included, since these physical findings stratified by the severity of the mental illnesses
treatments often have a notable magical or religious treated, where possible.

www.thelancet.com/psychiatry Vol 3 February 2016 155


Review

one case-control study,68 and nine retrospective


4128 records identified from database searches surveys.7,10,18,69–74 The studies varied substantially in terms of
1485 from MedLine
140 from PsycArticles the type and uniformity of interventions, patient
864 from Scopus recruitment and sampling methods, diagnostic spread,
1639 from Social Sciences Citation Index
follow-up protocols, and assessment methods. Most
interventions were administered to outpatients, in single
327 duplicates removed or repeated sessions. However, five studies52,59,63–65 report on
outcomes after extended admissions to traditional healing
shrines. Roughly a third of the papers reviewed used rating
3801 records after duplicates removed
scales to assess symptomatic improvement. Two-thirds of
the papers asked patients to rate the subjective effectiveness
2763 records rejected because about non-traditional healing modalities of the healing on a one-dimensional scale assessing
(eg, mindfulness, homoeopathy), about physical interventions, about a effectiveness, satisfaction, helpfulness, improvement, or a
physical illness (eg, cancer), non-empirical papers (eg, historical reviews),
or non-health related (political, economic) similar descriptive term.

Quality of papers
1038 abstracts considered
The methodological quality of the papers was generally
poor. Many aspects fell short of the standards usually
827 papers rejected from abstracts needed to assess the effectiveness of a psychiatric
or article skimming intervention (table 1). Convenience sampling used in
many papers is vulnerable to selection bias.12,51–53,56–58,69
211 papers for full-text 142 papers from reference Study designs lacked randomisation and control groups.
examination chaining Many papers used informal or unclear diagnostic
procedures7,12,42,49,54,57–60,70,75 or omitted diagnoses altogether
and listed only patient complaints.40,55,61,62,76 Few papers
353 papers for full-text
provided adequate descriptions of the procedures used
consideration for outcome assessment or provided a complete report of
results. Many papers did not account for the large
321 not eligible because contained only descriptive or qualitative data,
numbers of patients lost to follow-up.49,55,59,61,76
about personal religiosity, about exclusively physical illness or
substance addiction, or non-empirical (eg, reviews, opinion, Severe mental illness
theoretical)
Eight papers reported outcomes of mainly patients with
psychosis (table 2). The most rigorous of these, by Abbo
32 eligible papers and colleagues63 and Sorketti and colleagues64 used
validated rating scales to measure the change in psychotic
Figure: Search strategy symptoms of patients treated by traditional healers in
Uganda and Sudan, respectively. Both studies reported a
Results significant reduction in psychotic or manic symptoms
Literature search over the course of 3–6 months. Abbo and colleagues63
Our initial literature search found 3801 unique records noted that 80% of their sample attended local primary
(figure). This number was reduced to 211 potentially health clinics concurrently and probably received some
eligible papers, which were combined with a further form of conventional care. Sorketti and colleagues64
142 papers identified by reference chaining. Of these reported that many harmful treatments, such as beating
353 papers that needed detailed full-text examination, and chaining, were sometimes used. Harding and
64 (mostly published pre-1995) were unavailable through colleagues’ 1973 study65 of patients with psychosis
our university. We requested copies of 32 of the admitted to a traditional healer’s compound in Nigeria
unavailable papers from the authors, of whom 16 supplied reported significant rates of improvement that were
copies for consideration. Many insightful and excellent similar to those achieved with the psychiatric treatment
papers about the qualitative effects of traditional healing of the time. However, these studies are naturalistic and
on mental health were excluded because they did not uncontrolled studies, so comparison with other
provide any quantitative data. treatments or the normal course of illness is impossible.
In a controlled study, Halliburton52 compared outcomes
Eligible papers of severe mental disorder when treated by conventional
The 32 eligible papers from 20 countries consisted of psychiatric care, religious healing, or Ayurveda in India,
two randomised controlled follow-up studies,49,50 six and found a modest improvement that was similar across
non-randomised controlled follow-up studies,42,51–55 all three treatments. He suggested that different
14 uncontrolled naturalistic follow-up studies,12,40,56–67 treatments might be suitable for different patients, and

156 www.thelancet.com/psychiatry Vol 3 February 2016


Review

Country Study design Controls Sampling Diagnostic Assessment Outcome Reporting of Sample homogeneity
rigour procedures measures results
Controlled follow-up studies
Kleinman and Gale Taiwan Controlled follow-up study Adequate Adequate Poor Poor Adequate Complete Mixed physical and
(1982)54 psychological diagnoses
Koss (1987)55 Puerto Rico Controlled follow-up study Adequate Adequate Very poor Adequate Poor Complete Mixed social, emotional, and
behavioural complaints
Patel et al (1998)42 Zimbabwe Controlled follow-up study Adequate Adequate Poor Good Good Complete Mixed psychiatric disorders
Dixon (1998)49 UK Randomised controlled Good Adequate Poor, unclear Adequate Good Selective, Mixed physical and
follow-up study unclear psychological complaints
Mehl-Madrona USA Controlled follow-up study Poor Unclear Unclear Poor, unclear Poor, unclear Incomplete Mixed physical and
(1999)51 psychological complaints
Halliburton (2004)52 India Controlled follow-up study Adequate Poor Very poor Poor Adequate Incomplete Mixed severe psychiatric
disorders
Hurst et al (2008)53 USA Controlled follow-up study Poor Very poor Very poor Poor Poor Incomplete Mixed psychological complaints
Boelens et al (2009)50 USA Randomised controlled Good Adequate Unclear Adequate Good Complete Single psychiatric diagnosis
follow-up study
Case–control study
Salib et al (2001)68 Egypt Case–control study Adequate Adequate Adequate Unclear Adequate Complete Single psychiatric diagnosis
Uncontrolled naturalistic cohort studies
Harding (1973)65 Nigeria Prospective follow-up study None Adequate Poor Adequate Poor Complete Mixed psychiatric diagnoses, all
psychotic
Kleinman and Sung Taiwan Ethnographic follow-up None Adequate Very poor Poor Adequate Partial Mixed physical and
(1979)61 study psychological complaints
Finkler (1980)40 Mexico Ethnographic follow-up None Poor Very poor Adequate Adequate Partial, Mixed physical and
study selective psychological complaints
Salan et al (1983)12 Indonesia Ethnographic follow-up None Very poor Poor Poor Adequate Partial Mixed physical and
study psychological complaints
Peltzer and Machleidt Malawi Ethnographic follow-up None Poor Poor Poor, unclear Poor Partial, Mixed psychosocial and
(1992)59 study selective psychiatric complaints
Wirth (1995)60 USA Naturalistic prospective None Adequate Very poor, Poor Good Complete Mixed physical and
cohort unclear psychological complaints
Lemelson (2004)67 Indonesia Ethnographic follow-up None Very poor Good Poor Poor Partial Specific psychiatric diagnosis
study
Zacharias (2006)57 Mexico Naturalistic prospective None Very poor Poor Poor Poor Partial, Mixed psychiatric disorders
cohort selective
Schiff and Moore Canada Follow-up study None Poor Not applicable Poor Adequate Complete Unclear
(2006)56
Lee et al (2010)62 Singapore Ethnographic follow-up None Poor Very poor Poor Poor Partial Mixed physical and
study psychological
Boelens et al (2012)66 USA Prospective crossover None Poor Unclear Poor Good Complete Specific psychiatric diagnosis
follow-up study
Abbo et al (2012)63 Uganda Naturalistic prospective None Adequate Good Good Good Complete Specific psychiatric diagnoses
cohort
Sorketti et al (2012)64 Sudan Naturalistic prospective None Adequate Good Good Good Complete Specific psychiatric diagnoses
cohort
Mainguy et al (2013)58 Canada Naturalistic follow-up study Poor Very poor Very poor Poor Poor Incomplete Mixed physical and
psychological complaints
(Table 1 continues on next page)

argued that a greater availability of distinct forms of if they did not, independently of the presence of religious
treatment could improve mental health outcomes at the beliefs or religious delusions. However, reverse
population level. Although controlled, the study had a causality—the possibility that early relapse led to
small sample size, substantial loss to follow-up, unclear increased use of religious healing—could explain this
standards of conventional treatment, and an absence of association. Increased relapse rates were most
rigour in diagnosis and assessment. The Egyptian case– pronounced for patients who had attended a religious
control study by Salib and Youakim68 showed that elderly healer for exorcism or witchcraft rituals. Three
patients with schizophrenia relapsed more frequently if retrospective surveys of patients with mainly severe
they engaged in religious healing during remission than mental illnesses reported previous experiences of

www.thelancet.com/psychiatry Vol 3 February 2016 157


Review

Country Study design Controls Sampling Diagnostic Assessment Outcome Reporting of Sample homogeneity
rigour procedures measures results
(Continued from previous page)
Retrospective surveys
Bhana (1986)71 South Africa Retrospective survey None Poor Adequate Adequate Adequate Complete Mixed psychiatric disorders
Campion and Bhugra South India Retrospective survey None Poor Adequate Good Adequate Complete Mixed psychiatric disorders
(1997)73
Gadit (1998)18 Pakistan Retrospective survey None Poor, Poor, unclear Poor, unclear Unclear Partial Mixed psychiatric disorders
unclear
Ensink and Robertson South Africa Retrospective survey None Poor Unclear Good Adequate Partial Mixed psychiatric disorders
(1999)70
Mirza et al (2006)69 Pakistan Retrospective survey None Very poor Poor Poor Poor Complete Mixed psychiatric disorders
Kurihara et al (2006)7 Indonesia Retrospective survey None Poor Adequate Adequate Poor Partial Mixed psychiatric disorders
MacLaren et al Solomon Retrospective survey None Poor Poor Poor Poor Partial Single culture-bound syndrome
(2009)72 Islands
Salem et al (2009)74 United Arab Retrospective survey None Poor Adequate Adequate Poor Complete Mixed psychiatric disorders
Emirates
Mbwayo et al (2013)10 Kenya Retrospective survey None Unclear Adequate Adequate Poor Partial Mixed physical and psychiatric
disorders

Table 1: Included studies with quality assessments

attending a traditional healer for mental illness. In these with the shamans was largely attributable to a single healer
surveys, patients were generally dissatisfied with and that research nurses were overtly ambivalent to
traditional healing,70 which they found to be less effective traditional healing, both of which could have potentially
than conventional treatment.71 However, in a survey from skewed results. In Zimbabwe, Patel and colleagues42
Bali7 two-thirds of patients reported at least an compared outcomes between traditional medical
improvement after seeing religious healers, although practitioners, primary health-care clinics, and private
non-psychotic patients reported more improvement than general practitioners, using a validated questionnaire at 1
those with psychosis. Notably, recruitment at and 12 months, and reported generally high rates of
conventional psychiatric services in these studies improvement in symptoms of common mental disorders
probably excludes patients who were helped the most by irrespective of the type of treatment. Although the
traditional healers. differences between practitioners did not reach significance,
In general therefore, the studies of major mental patients attending traditional healers tended to report more
illness suggest that acute psychotic episodes might improvement than those attending primary health-care
improve somewhat over time under the care of traditional clinics. In Puerto Rico, Koss55 asked patients with similar
healers. Studies that assessed outcome at least 6 months complaints attending either spiritist healers or receiving
after the healing show that initial modest improvements psychotherapy in primary care to rate their expected degree
are often maintained.52,63,65 However, little evidence exists of problem resolution or symptom alleviation before the
to suggest that this improvement is different from the session. After the sessions, patients attending spiritist
natural course of the illness. The only study using healers rated outcomes significantly higher than those
conventional psychiatric care as a control showed a attending psychotherapists, with the difference largely
non-significant trend towards a worse outcome for the being explained by pre-existing expectations. These three
traditional healer patients.52 Somewhat alarmingly, controlled studies suggest that traditional healers might be
attending religious healers might increase the risk of as helpful as conventional primary care services for
relapse in chronic schizophrenia.68 psychosocial and neurotic complaints, although the results
vary by culture and comparison group. In uncontrolled
Common mental disorders studies, small samples of patients with mainly neurotic
13 studies analysed outcomes of mainly common mental disorders were followed up after attending dang-ki
disorders (eg, depression and anxiety) and other shamans in Taiwan61 and Singapore,62 and curanderos in
psychosocial complaints treated by traditional healers. In a Mexico.57 Patients reported generally high rates of perceived
controlled study, Kleinman and Gale54 reported that “effectiveness”, “helpfulness”, or “improvement”,
Taiwanese patients attending a tang-ki shrine mostly with respectively. Interestingly, two of these studies showed that
somatic complaints owing to psychiatric or psychosocial subjective reports did not necessarily correlate with
problems (ie, somatisation) were less satisfied with their symptomatic improvement or problem resolution.61,62 In
treatment and outcome than were patients attending the Taiwanese study, patients reported reluctance to criticise
western medical services matched for demographic and the treatment in case they prejudice the gods against
illness variables. However, they noted that dissatisfaction them,61 while the Mexican study suggested that less severe

158 www.thelancet.com/psychiatry Vol 3 February 2016


Review

complaints improved more than did severe ones.57 depression and anxiety ratings in depressed patients,
A controlled study of repeated individual prayer sessions in although the study had poor sampling, unclear diagnosis,
the USA50,66 reported significant improvements in and poor follow-up procedures.

Study Setting and Patient selection Sample size and Diagnostic Assessment Results Comments
location intervention* characteristics procedure and procedures and
diagnoses* outcome measures
Controlled follow-up studies
Kleinman Tapei, Shamanic treatment Consecutive patients 112 shaman patients Clinical interview by Semi-structured For somatisation Results possibly biased
and Gale Taiwan at tang-ki shrines with first attendance (28 with psychiatrist. interviews done complaints, TAU in favour of physicians.
(1982)54 (no details given) vs for current illness. somatisation) and Somatisation 3–4 weeks after more satisfactory Dissatisfaction with
TAU at private Controls matched by 118 physician patients category included treatment. Patients than shamanic shamans was largely
physician or general socioeconomic and (25 with mainly depression, subjectively assessed treatment (81% vs due to only one of
hospital clinic illness variables somatisation), mostly anxiety, chronic pain, satisfaction with 58% were at least three shamans studied
lower-middle or and conversion treatment and change satisfied)
upper-lower class disorders in symptoms
Koss Puerto Spiritist healers First new patients of 46 therapist patients No diagnoses listed. Patients rated pre- Spiritists were rated Higher expectations of
(1987)55 Rico working with the day attending and 54 spiritist Patient complaints treatment outcome as more effective spiritist patients were
individuals or groups trained therapists or patients. Spiritist included nervousness, expectations and than therapists associated with better
vs supportive and spiritist healers patients younger. family problems, subsequent outcome (p<0·05), but became outcomes. Almost half
problem-solving Similar occupational visions, and insomnia on a Likert scale non-significant after of patients did not rate
psychotherapy at and socioeconomic correction for their outcome, which
community clinics status in pre-treatment is a potential source
and a psychiatric the two treatment expectations of bias
hospital groups
Patel et al Zimbabwe Traditional medical Consecutive patients Traditional medical Nurses and GPs Shona Symptom Caseness at baseline, Traditional medical
(1998)42 practitioners vs scoring >8 on Shona practitioners: n=67; confirmed caseness Questionnaire 2 months, and practitioners’ diagnosis
primary health care Symptom primary health care: for common mental administered at 12 months was of kufungisisa (roughly
clinics and GPs in Questionnaire were n=72; general disorders (depressive baseline, 2 months 100%, 38%, and 41%, equivalent to common
urban Harare eligible “cases” practitioner: n=60. and anxiety disorders, and 12 months to respectively. No mental disorders) was
35% male, 46% not details not given) assess caseness significant difference significantly related to
formally employed, between treatment improved outcome.
mean age 34·5 years groups Belief in witchcraft
associated with poorer
outcome
Dixon UK Semi-rural general Consenting patients Healer: n=30; Chronic symptoms Patient interview with Median symptom Research nurse not
(1998)49 practice. 10 weekly were divided controls: n=27. for >6 months, research nurse (not rating and HADS masked. Possible
40-min sessions of alternately to healing 32% male, diagnostic procedure masked to group scores at 3 and difference in ages of
brief discussion, or control group by modal ages: not specified. allocation) assessing 6 months decreased controls and
application of healing the research nurse healer group Depression and stress overall symptoms significantly more intervention
hands over body >65 years, accounted for 20% of (1–10) and HADS, at (p<0·05) in healing participants. Unusual
while visualising with controls 45–54 years complaints baseline and at 3 and group than in control choice of summary
music vs conventional 6 months group statistics. High loss to
care from GP follow-up. Selective
reporting of results
Mehl- USA Intensive Native Intervention patients Intensive Diagnostic procedure 5-year follow-up by Perceived Control results not
Madrona American healing contacted author to intervention: n=107; not specified. semi-structured improvement of the reported. Methods
(1999)51 experience lasting request Native emergency room: Diagnoses: depression telephone interview 11 depression and unclear. Small sample
7–21 days, with healing. n=100; modal age: or bipolar disorder in of 76% of intervention bipolar disorder size. Possibility of social
follow-up phone calls Matching of controls 30–49 years 11 patients group. Follow-up of patients: “cure” 7, desirability and
vs patients with not described controls not described “better” 3, “no expectancy bias
equivalent diagnoses change” 1, “worse” 0
attending a hospital
emergency room for
TAU
Halliburton Kerala, 3 religious healing Random or Religious healing: Diagnoses of severe Interviewed at healing All three groups Substantial loss to
(2004)52 India centres (Hindu, convenience n=33; allopathic: mental disorders such centres (or postal showed modest follow-up. Small
Muslim, and sampling of n=35; Ayurvedic: as schizophrenia was questionnaire if improvements, no sample sizes. Diagnoses
Christian) with inpatients n=32. Of these assumed, on the basis discharged), at significant differences and assessment
reputations for 100 patients, 53 were of patients being baseline (n=53), and procedures informal
healing mental inpatients and the treated as inpatients 6–9 months later and poorly described.
disorders vs focus of this paper at the healing centres (n=16). Psychiatric standard of
2 allopathic Researcher rated care not outlined. High
psychiatric hospitals “degree of decline or risk of rater bias
and an Ayurvedic improvement” on
hospital 7-point scale
(Table 2 continues on next page)

www.thelancet.com/psychiatry Vol 3 February 2016 159


Review

Study Setting and Patient selection Sample size and Diagnostic Assessment Results Comments
location intervention* characteristics procedure and procedures and
diagnoses* outcome measures
(Continued from previous page)
Hurst et al USA Prayer sessions lasting Conference attendees Participants: n=78 No formal diagnoses Custom questionnaire Intervention group Substantial loss to
(2008)53 6–7 h as part of “Steps were invited for selected for inclusion; made. before and after reported deductions follow-up. Control
to Freedom in Christ” interview, then controls: n=40 Selected for inclusion healing session in symptom scores of group and assessment
at a conference vs selected if reported if “significant (n=78). Repeat 42–49% (p<0·005) procedures not
controls who psychological depression (including questionnaires and improvements in described. Unclear if
volunteered at problems suicidality), anxiety or returned by functioning of controls had any
church, had not other typical self-addressed 14–45% (p<0·05). psychological
attended conference, presenting problems” envelope 7 days and Controls had complaints
and had no 3–4 months later non-significant
intervention (n=33) changes
Boelens USA Six prayer sessions Recruited from Prayer intervention: Diagnosed in HAM-D, HAM-A, Significant reduction Unusual effect sizes for
et al lasting 60 min over medical physicians n=27; controls: n=36 outpatient setting LOT6, and DSES7 (p<0·01) in HAM-D, prayer group: HAM-D
(2009)50 6 weeks, conducted through posters, (no intervention) with DSM-IV before, after, and at HAM-A, DSES and effect size 3·0, HAM-A
by lay-preacher with some directly from depressive disorder 1 month following significant increase in effect size 4·1. Beliefs
participant. community (most also with intervention LOT after and expectations of
No explicit anxiety). No further intervention and at participants not stated.
psychotherapy or details provided 1 month follow-up in Self-selection bias likely.
touch involved prayer group. No Control condition
changes in control poorly matched
group
Case–control studies
Salib and Egypt Private psychiatric Patients with 20 cases and Clinical diagnosis of Outcome was acute Cases: 85% relapsed Cases were defined by
Youakim hospital. Cases schizophrenia drawn 20 controls. All on schizophrenia by relapse of and required hospital increased and excessive
(2001)68 received religious from register of maintenance hospital psychiatrist schizophrenia admission; controls: personal religious
healing before study psychiatric hospital, antipsychotics. satisfying ICD-10 necessitating hospital 60% relapsed and activity and religious
and during follow-up. age >60 years Matched for age, sex, criteria. Mean admission following required admission healing such as
Controls received no and illness duration duration of illness stable remission (p<0·05) attendance at healing
spiritual healing and 21 years (SD 8) period, over rituals including
had TAU 18-month period witchcraft or exorcism.
Reverse causality might
bias results
Uncontrolled naturalistic cohort studies
Harding Nigeria Traditional healer’s All patients with 24 patients. 50% Researcher interview Assessed during the At 3-month and Outcomes compared
(1973)65 compound in western psychosis who were male, mean age and observation. No admission, and 6-month follow-ups, favourably to outcomes
Nigeria. Treatments unwell enough to 33 years diagnostic schedule followed up with at least a third were in Western psychiatric
included need admission to specified. informal interview at fully recovered, at hospitals in
confinement, the healer’s Schizophrenic 42%, patients’ homes least a third had slight contemporaneous
chaining (for compound during a manic 33%, depressed 3 months and incapacity, and less reports
violence), herbal period of 6 months 21%, other 4% 6-months after than a third were
remedies, divination, discharge. No details severely incapacitated
chanting, communal of outcome
work. Modal stay assessments provided
6–10 weeks
Kleinman Taiwan Shamanic treatment Consecutive clients 19 patients, of whom No formal diagnoses Semi-structured Mixed results. Some patients followed
and Sung at tang-ki shrine. Brief attending a tang-ki 6 had psychological made. Patient interview at clients’ 5/6 rated treatment shaman’s assuredness
(1979)61 shamanic rituals, shrine over 3 days problems: women complaints labelled by homes 2 months after “effective” or of effectiveness, or
followed by aged 22, 25, 54, and researchers as initial visit, exploring “partially effective”. feared prejudicing the
meditating or 61 years; man aged somatisation (n=3), perceived Patients’ evaluation gods. Incomplete
praying. Other cult 31 years; and boy aged anxiety (n=1), bad effectiveness of of effectiveness and reporting of results.
members assist in 10 years behaviour (n=1), and healing and perceived symptom change Loss to follow-up might
healing depression or symptom change. sometimes differed bias findings towards
personality (n=1) 7 clients lost to better outcomes
follow-up
Finkler Mexico Brief ritual interaction Every fifth patient 107 patients. No formal diagnoses Initial assessment at Failures 35·5%, Only successful
(1980)40 with a curer at was eligible. 15% male and 27% made. A third of temple by researcher, successes 25·3%, outcomes reported in
Spiritualist temple in Follow-up sample children complaints had with home follow-up inconclusive 20%, detail, unclear how
rural Mexico. Ritual lived nearby, overtly psychological 7–13 days and other 19%. Cornell many unsuccessful
includes light touch, consented, and were component 30–45 days later. Medical Index scores psychological
blessing, symbolic not regular attenders (eg, nerves, Cornell Medical Index1 improved to similar outcomes. Some
cleansing, brief nightmares, feeling and subjective degrees in both patients reported
history taking, angry, seeing dead assessment of the perceived successes success despite
healer’s trance, people, “cerebro”) treatment received and failures continuing symptoms
and herbs
(Table 2 continues on next page)

160 www.thelancet.com/psychiatry Vol 3 February 2016


Review

Study Setting and Patient selection Sample size and Diagnostic Assessment Results Comments
location intervention* characteristics procedure and procedures and
diagnoses* outcome measures
(Continued from previous page)
Salan and Indonesia Ritual healing in Patients selected 90 patients with an Clinical assessment by Semi-structured Improvement Diagnoses were unclear
Maretzki 9 healers’ offices or “somewhat even age distribution physician. No interviews and Cornell reported in 60% and and not rigorous.
(1983)12 houses. Various arbitrarily” if willing from <20 years to schedule. 61% had Medical Index1 57% of psychological Incomplete reporting of
methods including to participate in 60 years. 38% male, psychological or immediately after the and major disorders results. Possible
holy water, Koranic interviews and fill in and all lower or psychiatric problems, healing session, and at respectively, although response bias due to
talismans, herbs, questionnaires lower-middle class 15% had major home 2–4 weeks later, half were regarded as influence of
palpation, exorcising mental disorders especially “views self-limiting. Great Western-trained
rituals, divination, (mean duration concerning the variation between clinicians. Possible bias
and trance 4·7 years) outcome of individual healers due to convenience
treatment” sampling
Peltzer and Malawi Traditional healing Patients selected if 51 patients. No details Diagnostic procedure Unstructured 51% “improved”, Unclear and incomplete
Machleidt community for semi-structured given not described. interview including 28% “not improved”, data reporting, text
(1992)59 inpatients. Daily tasks interview revealed Neurotic and subjective 22% “no follow-up” does not agree with
included chores, mental disorder. No personality disorders improvement rated by tables. Possibility of
prayer, music, dance, details of selection 35%, psychosocial patients and relatives investigator bias
possession, rituals, procedure given disorders 33%, at next visit to healing
ancestor veneration, schizophrenia 28%, centre or elsewhere
sacrifices. Mean affective disorder 4% within 1 year.
duration 8 months 11 patients lost to
follow-up
Wirth USA Multiple 15–30 min Consecutive new 48 patients, of whom Independent medical General Well Being Significant No clinician ratings.
(1995)60 healing sessions over patients attending 6 had mental diagnoses were used Questionnaire,2 Health improvement on all Choice of non-
3 weeks in Espiritista Espiritista healer disorders. 64% white. “when available”. Perceptions measures (p<0·05) dominant healing
healer’s office. 38% Christian, 15% Mostly chronic Questionnaire,3 with including worry/ paradigm could lead to
Idiosyncratic methods Jewish, and 15% physical disease. postal questionnaire concern, sickness self-selection bias. 31%
based on spiritual Atheist Mental disorders follow-up 18–25 days orientation, anxiety used other treatment
power, laying on of included anorexia, later. Patient and depression. concurrently. Perceived
hands, and religious depression, and expectations assessed Improvement cure 12·5%, significant
icons manic depression with 3-item correlated with Improvement 52%,
questionnaire expectations slight improvement
(no details) 25%, no change 10%
Lemelson Bali, Treatment by Muslim Patients with 22 patients. Obsessive Diagnoses of Subjective 1 patient noticed Potential reporting bias
(2004)67 Indonesia dukuns, including obsessive-compulsive compulsive disorder obsessive-compulsive improvement, improvement in since no rating scales
religious ceremonies, disorder and/or mean age 39·9 years disorder and/or assessed by interviews symptoms, which were used. Potential
massage, incense, Tourette’s syndrome (SD 16). Tourette’s Tourette’s syndrome over 20-month was temporary. No sampling bias skewed
sorcery, charms, and recruited from syndrome mean age made by psychiatrist duration of study. other improvements towards protracted and
sacrifices. Some psychiatrists and not reported according to DSM-IV No details provided. noted severe cases recruited
patients attended neurologists in Bali and ICD-10 No Y-BOCS outcomes from specialists
several times definitions, using reported
items from the CIDI
and Y-BOCS
Zacharias Mexico Curanderos Inclusion required 8 patients. 37% male. ICD-10 criteria. Semi-structured Improvement after Patient selection and
(2006)57 diagnosed and presence of 7 adults (mean age Procedure not interview treatment: rating procedures
treated by divination, psychological 30 years), 1 child (age described. pre-treatment and 4 complete, 3 partial, poorly described,
tactile contact, disorder, degree of 9 years) Adjustment disorders post-treatment and 1 no data. At potential for selection
symbolic rituals, self-reflection, n=4, chronic pain n=1, 6 months’ follow-up, 6 months: and rater bias. 6-month
spiritual practices, and availability, and panic n=1 (both with including rating of 6 complete follow-up not
occasional catharsis. consent anxiety/depression), degrees of improvement, 2 no described. Long
Multiple sessions over substance improvement from data. Less severe follow-up and mostly
weeks to months dependence n=1, “deteriorated” to disorders improved self-limiting conditions
schizophrenia n=1 “fully or very more than did
improved” schizophrenia
Schiff and North Sweat lodge All participants who 42 patients. 59% No diagnoses or Health screening No significant No diagnoses
Moore American ceremony as practised arrived early to fill in indigenous First complaints noted questionnaire (SF-364) changes in mental or mentioned
(2006)56 Native by First Nation’s baseline Nation people, and Heroic Myth physical health scores (self-selected sample
healing People of Canada. questionnaire were 41% not indigenous. Index5 immediately on SF-36, probably who may have
(USA) Lasts up to 3 h and invited to take part 75% had attended a before and after because of ceiling attended for general
culminates in a feast. ceremony before ceremony (including effect and small wellbeing). Ratings
Ceremony not feast). Long-term sample size. Innocent, done 3 h apart,
described out of outcomes not tested Sage and Warrior potential for
respect to elders scales increased expectancy bias
significantly on
Heroic Myth Index
(Table 2 continues on next page)

www.thelancet.com/psychiatry Vol 3 February 2016 161


Review

Study Setting and Patient selection Sample size and Diagnostic Assessment Results Comments
location intervention* characteristics procedure and procedures and
diagnoses* outcome measures
(Continued from previous page)
Lee et al Singapore Chinese dang-ki Patients in lobby of 21 patients, of whom No formal diagnoses Three ethnographic Post-treatment: Inconsistent reporting,
(2010)62 healers in three shrine who gave 9 had psychosocial made. Presenting interviews at the 5 helpful, 3 not sure no details of 1 month
selected shrines. consent and issues (reported here). problems included shrine immediately but hopeful, and outcomes. Only 3 of
Healing involved presented with health Mean age 44 years, anxiety, depression, before and after the 1 unhelpful. At 9 patients thought
trance, divination, or psychosocial 33% with a nightmares, bad luck, healing, and by phone 1 month, “as a whole their problems were
advice, divine writings problems high-school unemployment, and 1 month later. the consultations solved. Worry of
and symbols, and education, and interpersonal Focused on research were considered offending the gods
talismans 66% employed problems questions including helpful” might bias ratings at
perceived helpfulness shrine
of dang-ki
Boelens USA This study is the Initial recruitment as 63 initially enrolled, Diagnosed in HAM-D, HAM-A, Decreases in HAM-D, Adjunctive
et al one-year follow-up of for the 2009 study. 48 of whom received outpatient setting LOT6, and DSES7 HAM-A, and DSES psychotropic
(2012)66 Boelens and Control participants prayer intervention. with DSM-IV repeated 1 year maintained at 1 year medication was
colleagues’ 2009 could choose to enter Mean age 48 years, depressive disorder following (p<0·01). Increase in allowed before and
study50 (described prayer intervention all women (most also with intervention. LOT maintained at during study. Beliefs
above). The prayer anxiety). No further 4 participants lost to 1 year (p<0·01) and expectations of
group and crossover details provided follow-up participants not stated.
prayer group received Self-selection bias likely
no further
interventions during
the year
Abbo et al Uganda Traditional healers at Patients attending 115 patients aged SRQ-25 and MINI. All Assessment at >20% improvement Ratings were blinded.
(2012)63 their shrines in two traditional healer >18 years, mean age had psychosis baseline, 3 months, in all scales at 3 and 80% of patients
districts of eastern shrines who 34·5 years (SD 8), (psychotic depression and 6 months 6 months, all attended local
Uganda. Treatment answered yes to any 56% male patients 40%, mania 36%, (follow-up rates of significant. Greatest psychiatric clinics
not specified screening questions schizophrenia 24%) 87% and 78% improvement for concurrently. Clinic
for psychosis were respectively) using mania, smallest in attendance decreased
included PANSS, YMRS, schizophrenia caseness at 3 months
MADRS, CGI, GAF, and but increased caseness
COMPASS at 6 months, compared
with non-attendance
Sorketti Sudan 10 randomly selected All inpatients 129 patients, mean Schizophrenia PANSS assessed on Reduction in PANSS Potentially harmful
et al traditional healing receiving treatment age 29 years (range diagnosed on MINI. admission and from 118·36 on interventions included
(2012)64 shrines. Treatments in the shrines with a 16–55), 71% male, Mean duration of discharge. Mean admission to food restriction,
included admission, diagnosis of 25% with secondary untreated psychosis duration of admission 69·36 on discharge. chaining, isolation,
reciting holy verses, schizophrenia education or higher, 15·8 months 4·5 months, median Reduction in total beating, and stopping
inhaling holy smoke, 54% not working <2 months and all subscales was psychiatric treatment
and holy water (see statistically (in 20%)
also the Comments significant
column)
Mainguy Canada Canadian Aboriginal Participants 155, of whom 35 had Diagnostic procedure Phone contact at 5-year outcomes of Unclear if all
et al healing, including contacted author psychiatric not specified. regular intervals psychiatric 155 patients were
(2013)58 sweat lodge requesting traditional complaints. Diagnoses: anxiety (usually every complaints: cured 15, followed up over
ceremony, Aboriginal healing Participants were disorders 12%, mood 6 months for 5 years). better 12, the same 7, 5 years. Loss to
storytelling, daily non-Aboriginal, disorders 8%, other Medical outcomes worse 1, dead 0 follow-up not
prayer ceremonies, median age psychiatric disorders rated on 5-point Likert mentioned
and sacred pipe, 40–49 years 2% scale from +2 (cured)
followed by to 0 (no different) to
debriefing −2 (dead).
(Table 2 continues on next page)

Three retrospective surveys recruited patients with studies, such as the timing, location, and the wording of
mainly neurotic disorders from the shrines of healers or questions. Lemelson67 noted that traditional healing is
directly from the community.10,18,69 In these surveys, often thought to help individuals to reframe and make
treatments by shamans and religious healers for mainly sense of their illnesses, but that this highly subjective
depression in Pakistan achieved only modest ratings of process might not necessarily improve underlying
satisfaction,18,69 by contrast with a Kenyan survey that neuropsychiatric symptoms. He therefore studied the
indicated that 95% of patients were satisfied with the effectiveness of Balinese traditional healers in treating the
traditional healing they had received.10 The wide variation symptoms of obsessive compulsive disorder or Tourette’s
in results could represent real differences in effectiveness disorder, which are less responsive to non-specific
of the respective healers, or might be caused by the therapies than common psychosocial problems. Of
sensitivity of subjective ratings to procedural aspects of the 22 patients with these disorders who had repeatedly

162 www.thelancet.com/psychiatry Vol 3 February 2016


Review

Study Setting and Patient selection Sample size and Diagnostic Assessment Results Comments
location intervention* characteristics procedure and procedures and
diagnoses* outcome measures
(Continued from previous page)
Retrospective surveys
Campion South Psychiatric outpatient Consecutive patients 198, of whom 89 had ICD-10 diagnoses by Semi-structured “Subjective 90% of patients had
and Bhugra India clinic. Previous attending during a previously consulted researcher: mood interview with improvement in stopped religious
(1997)73 religious healing 3-month period religious healers. disorder 27%, patients and relatives symptoms” rated by treatment by the time
included chanting 81% Hindu, 9% neurotic or stress at the outpatient 33% of Muslim of psychiatric
mantras, thayath Muslim, 10% Christian disorder 23%, clinic, including patients, 35% of consultation. Hospital
(amulets), ingestion psychotic disorders questions about the Christian patients, sampling might
or sprinkling of ash or 23%, substance use patients’ contact with and 23% of Hindu introduce bias
holy water, 12%, other 15% religious healers patients
take-home objects,
and sacrifices
Bhana South Psychiatric outpatient Indian patients who 86, all of whom had Diagnosis from Semi-structured Doctors: effective Timing of previous
(1986)71 Africa facility. Previous had previously first consulted hospital notes: interview by 70%, not effective healer contact not
spiritual healing attended a spiritual spiritual healers schizophrenia 56%, psychologist with 26%; healers: specified. Hospital
included application healer depression 22%, patient and family effective 53%, sampling might
of ashes, cloves, lime, epilepsy 14%, member, including not effective 40%; introduce bias.
amulets or holy water, behaviour patients’ perceptions only the doctors 40% believed that both
prayer or fasting, and problems 7% of “effectiveness” helped 26%; only the doctors and healers had
other rituals and healers helped 8% helped
sacrifices
Gadit Pakistan Shrines, mosques or Patients visiting “About 400”. No other Diagnosed by Semi-structured Patient reports: 30% Timing of interview
(1998)18 residences of shamans. Selection details given psychiatrist using interview by “entirely satisfied” and treatment not
100 shamans in not described DSM-IIIR criteria. psychiatrist in with shamanic specified
Karachi districts. Diagnoses: depression shamanic place of treatment; 70%
Treatments included 80%, psychosomatic healing (details not “mixed feelings”
amulets and disorders 5%, epilepsy given). Patient
talismans, holy water, 7%, psychosis 6%, interviewed at time of
recitations, and rituals no disorder 2% visit to shaman (no
details given)
Ensink and South Two psychiatric Random sample of 62, of whom 36 had Diagnoses by Semi-structured For diviners and faith Generally low
Robertson Africa hospitals in Cape first admission consulted diviners or psychiatric services questionnaire healers: carers satisfaction with
(1999)70 Town. Details of African patients faith healers in the (no details given). exploring 30% satisfied, traditional healing.
previous healing past 12 months Most had acute “satisfaction” with all 70% uncertain or Diviners in particular
interventions not psychosis, organic services used. Main dissatisfied; patients criticised harshly.
described syndromes, caregiver and patient 47% satisfied, Psychiatric services
parasuicide, or mood interviewed at 53% uncertain or rated as more
disorder. No details patient’s home. dissatisfied satisfactory. Hospital
given Timing of interview sampling could bias
unclear selection
Mirza et al Pakistan Mental health Convenience sample 150, of whom 72 had Diagnosed by Semi-structured For religious or faith GPs received highest
(2006)69 consultation day in of attending villagers consulted religious or registrar-grade interview, including healers: 20–23% ratings of effectiveness.
village in rural who had faith healers psychiatrist or higher. rating of effectiveness response for Ratings were
Pakistan. Healers seen psychological Depression 68%, of previous depression, 22% retrospective, with no
included religious complaints and gave other 27% (mainly treatments on 4-point response for other apparent time limit on
healers (use Islamic consent epilepsy) Likert scale, then complaints when previous healers
teachings) and faith collapsed to “response were consulted
healers (for removing or not”
black magic or
harmful spells)
(Table 2 continues on next page)

consulted dukuns (Indonesian shamans), only one patient contact with a traditional or spiritual healer.49,51,57,58,66
noted a temporary improvement. He suggests that the By contrast, less malleable symptoms such as obsessions
psychosocial techniques used by traditional healers are or tics might be less amenable to such support.67
ineffective in treating neuropsychiatric symptoms.
Although wide variation exists, traditional healers in Culture-bound syndromes
general could be modestly helpful in the treatment of Surprisingly, the only study to assess the outcomes of
common mental or neurotic disorders that are most culture-bound syndromes as diagnosed by traditional
amenable to psychosocial interventions. In particular, healers is a retrospective survey by MacLaren and
patients with mild depression and anxiety tend to colleagues.72 They recruited patients in remote hamlets
respond favourably to repeated, regular, intense in the Solomon Islands who had been treated by

www.thelancet.com/psychiatry Vol 3 February 2016 163


Review

Study Setting and Patient selection Sample size and Diagnostic Assessment Results Comments
location intervention* characteristics procedure and procedures and
diagnoses* outcome measures
(Continued from previous page)
Kurihara Bali, Sampling at Consecutive new 54, of whom 47 had Diagnosed by Once psychiatrically 29·8% “much Treatments included
et al Indonesia psychiatric hospital, patients at psychiatric attended traditional psychiatrist using stable in own home, improved”, sprinkling or ingesting
(2006)7 interview in patients’ hospital healers ICD-10 criteria. 80% patients rated 38·3% “improved”. of holy water, chanting
home. Previous with psychotic “treatment effect” of (No data reported for mantras, advice,
healing interventions symptoms, 57% the traditional healer other outcomes). religious ceremonies,
by Hindu (95%) or admitted to hospital. on 5-point Likert Non-psychotic use of oil, herbal
Muslim (5%) healers 40 schizophreniform, scale: “much complaints rated medicine, and massage.
17 with mood, stress, improved”, significantly more Hospital sampling
or other disorders “improved”, “no improved than could introduce bias.
change”, “worse”, or psychotic complaints Ratings were
“much worse” retrospective
MacLaren Solomon Treatment by Patients who had 20 patients, 30 family Patients were Single discussion with 50% reported Buru possession
et al Islands community been possessed in the members diagnosed by patients and family complete cure, symptoms include
(2009)72 traditional healers in past and their family traditional healers members, rating 30% reported mutism or
five remote hamlets members. Selection with buru possession subjective temporary relief with over-talkativeness,
of the Solomon process not described (culture-bound improvement. later relapse, and 20% suicidal ideas, delusions,
Islands. Includes: food syndrome, see No details given reported no relief aggression, or social
taboos, divination, Comments column). isolation, lasting 1 day
and ritual smearing Psychiatric diagnoses to several weeks, rarely
with crushed lime and not given becoming permanent
leaves
Salem et al United Psychiatric outpatient Consecutive new 106, of whom 45% Consultant Semi-structured Of those who Patients with psychosis
(2009)74 Arab department and outpatients and all had consulted faith psychiatrist made interview with consulted faith visited faith healers
Emirates inpatient unit. admissions over healers before diagnosis using patients after seeing healers, 45% reported more frequently and
Interventions by 9 months in 2003 attending psychiatric ICD-10. Psychotic psychiatrist. Custom some improvement made more
Islamic faith healers services disorders 23%, questionnaire but symptoms supernatural
included prayers and depression 25%, included details of recurred later, 47% no attributions of illness
herbs anxiety 26%, other previous contact with improvement, and than depressed or
non-psychotic healers and views on 8% got worse anxious patients.
disorders 26% outcome. No further Hospital sampling
details given could introduce bias.
Ratings were
retrospective
Mbwayo Kenya 59 traditional healers Selection of patients 305, of whom 196 MINI-plus Nursing students did 95% responded Traditional healers
et al in 3 districts around attending healers not had a mental disorder administered by “in-depth interviews” reported satisfaction diagnosed only
(2013)10 Nairobi. Treatments specified diagnosed using trained lay with patients with traditional 27 patients with mental
included counselling, MINI-plus interviewers. Major including “patients’ healer services illness (mostly
herbs for ingesting, depressive disorder satisfaction with psychosis) and 15 with
washing or inhaling, 20%, post-traumatic services offered by witchcraft or
divination, and stress disorder 15%, traditional healers”. possession, compared
removal of bewitched generalised anxiety No further details with 196 diagnosed
items disorder 10%, other given using MINI
anxiety disorders
10%, alcohol misuse
9%, schizophrenia 7%

TAU=treatment as usual. GP=general practitioner. HADS=Hospital Anxiety and Depression Scale. HAM-A=Hamilton Rating Scale for Anxiety. HAM-D=Hamilton Rating Scale for Depression. LOT=Life Orientation
Test. DSES=Daily Spiritual Experiences Scale. ICD-10=International Classification of Diseases 10th edition. CIDI=Comprehensive International Diagnostic Interview. Y-BOCS=Yale–Brown Obsessive Compulsive
Scale. SQR-25=Self Reporting Questionnaire 25.10 MINI=Mini International Neuropsychiatric Interview. PANSS=Positive And Negative Symptoms Scale. YMRS=Young Mania Rating Scale.
MADRS=Montgomery-Asberg Depression Rating Scale. CGI=Clinical Global Impression. GAF=Global Assessment of Functioning. COMPASS=COMPASS Mental Health Index.9 CGI=Clinical Global Impression.
MDD=major depressive dsorder. MINI=Mini International Neuropsychiatric Interview. *Treatments or diagnoses are ordered from most to least common.

Table 2: Studies assessing the effectiveness of traditional healers—summary of findings

traditional healers for buru (possession; symptoms experimental design.77 In an ethnographic study, the
listed in table 1). Subjective rates of permanent cure researcher, usually an anthropologist, embeds herself
were 50%, with 30% finding temporary relief and 20% within the social context of an interaction to explore its
reporting no improvement. meaning and social function in an open and unstructured
manner.78 Diagnoses are necessarily often unclear or
Mixed diagnoses mixed. Despite the different emphasis, some ethnographic
Several earlier studies adopted an ethnographic approach, studies report quantitative outcomes. Finkler40 in Mexico
which typically has a greater focus on qualitative than critically assessed the success rates of spiritualist healers.
quantitative outcomes and uses a less rigorous Her sample with mixed physical and psychological

164 www.thelancet.com/psychiatry Vol 3 February 2016


Review

complaints reported more perceived failures (36%) than recommend psychosocial interventions as first-line
successes (25%), although incomplete reporting of treatment for common mental disorders, but only as
outcomes prevents a detailed analysis. Salan and adjunct therapy for major psychiatric disorders such as
Maretzki’s12 sample of patients attending Indonesian psychoses and mania.79
traditional healers with a wide variety of complaints The reviewed studies provide some insights into the
reported high rates of improvement (60%) for psychological mechanisms that enable positive outcomes for traditional
and psychiatric problems. They also found some healers healing. Two studies55,60 reported a significant effect for
performed consistently better or worse than others. patient expectancy and belief on outcome. Similarly,
However, poor reporting and vague diagnoses make two studies12,54 showed that individual healers can differ
detailed analysis difficult. Peltzer and Machleidt59 followed substantially in their effectiveness. Indeed, positive client
up patients with mixed psychiatric disorders who had been expectations and the personal qualities of the healer are
admitted to a traditional healing centre in Malawi. core ingredients of traditional healing and Western
Follow-up within the year after discharge showed that half psychotherapy alike, suggesting that these disciplines
of the patients and their relatives felt that the patient’s might share some therapeutic mechanisms.80–82 In
condition had improved, although almost a third denied traditional healing especially, ritual and ceremony can
any improvement, with the rest of the sample lost to encourage successful outcomes by increasing expectancy
follow-up. A follow-up study by Wirth60 that recruited and the belief in the authority of the healer.83 Interestingly,
mainly non-indigenous white patients attending an belief in the power of the ceremony and the healer’s
Espiritista healer in California, USA, used a validated scale authority might be helpful even in the absence of a
to show improvements in general wellbeing, which were shared cultural heritage, as shown in the two studies that
predicted by the patients’ pre-healing expectations. report favourable outcomes for non-indigenous
Two controlled studies of spiritual healing in the UK49 and patients.58,60 Going further, a German study showed that a
a Native American healing ceremony51 recorded significant contrived spiritual ceremony with no overt meaning was
improvement in depression ratings and subjective found to improve participants’ mental, spiritual, and
improvement, respectively, compared with controls. physical quality of life.76 The role of ritual, metaphor, and
However, both studies had notable flaws, including control meaning in traditional healing is described well
conditions that were poorly matched to the intensity of the elsewhere.36,41,83
active interventions.
In four retrospective surveys,70,71,73,74 patients with mixed Limitations of the published literature
psychiatric disorders were recruited from psychiatric Overall, the quality of studies is poor when viewed
clinics or hospitals and asked about their experiences of through the lens of rigorous psychiatric research. Most
traditional healing. Perhaps unsurprisingly, these trials do not include a control group, and therefore to
patients typically reported some ambivalence or frank assess whether any changes over time differ from the
dissatisfaction with their previous traditional healing, natural course of the illness is difficult. This situation is
although sampling and assessment factors might have particularly problematic for psychiatric illnesses in
biased these findings. which the natural course of symptoms can wax and
wane. Spontaneous improvement might therefore be
Discussion wrongly attributed to the healing intervention. In many
We found 32 studies that reported quantitatively about studies, patients attending traditional healers are a self-
the effectiveness of traditional healers in treating selected group with high expectations, which limits the
mental disorders and psychological distress. Although generalisability of the results to the general
substantial shortcomings exist in the published literature, population.40,50,51,53,56–58,60,66 For example, in Mehl-Madrona
some generalisations can be made. Acute relapses of and colleagues’ controlled follow-up study of
major mental illnesses such as schizophrenia and bipolar Native-American healing,51 patients actively sought
disorder tend to improve with time under the care of alternative healing and elected to travel from out-of-
traditional healers, although improvements might state for the intervention, whereas control patients
represent the natural course of the illness. The small merely attended local hospitals. Patient diagnoses are
amount of available evidence suggests that traditional often vague and subjective, and diverse psychiatric
healing methods have little effect on the long-term diagnoses are frequently grouped together, sometimes
course of chronic illnesses such as schizophrenia and with physical complaints. Outcome assessments often
obsessive-compulsive disorder. Common mental use non-validated rating scales or interviews. Ill-defined
disorders, such as depression, anxiety, and somatisation, subjective measures (eg, “satisfaction” or “helpfulness”),
and interpersonal and social difficulties, seem to be more although useful, are prone to different interpretations
likely to respond to traditional healing interventions, by patients and researchers and are especially
which might be at least as effective as primary psychiatric vulnerable to response and expectancy biases. Patients
care in the countries studied. These findings are might attempt to produce socially desirable responses
consistent with modern psychiatric guidelines that in their interactions with both researchers7,12,42,54,70 and

www.thelancet.com/psychiatry Vol 3 February 2016 165


Review

traditional healers.54,61,62 Researchers themselves might Design


give biased ratings in studies without masked control Control groups are essential, but to identify a suitable
interventions.49,51,59,67,75 Many studies do not adequately control to compare with traditional healing is difficult.
describe their assessment procedures and might not Traditional healing is embedded in culture, and
sufficiently acknowledge and minimise these sources evaluations of its effectiveness are only likely to be valid
of bias. Data reporting across studies is inconsistent for people within that particular culture. Culturally
and patchy, with studies often omitting actual scores foreign control interventions, such as a waitlist group,
from assessment scales and not providing reasons for psychotherapy, medication, or non-specific support,
loss to follow-up. Finally, the healing interventions suffer the disadvantage of not meeting important
themselves are heterogeneous and not standardised— cultural needs and expectations. Withholding culturally
both between and within studies—making it difficult to expected treatment for the sake of a control condition is
tease out which components of the intervention also likely to increase distress and might not be viewed
contribute to effectiveness. Notably, the most as ethical.63,90 Some studies in this review42,52,54,55 therefore
methodologically flawed studies often find the most forego randomisation and give patients the choice of
unambiguous and impressive effects.49,53,58,84 allocation to traditional healing or a control condition of
conventional treatment. Although this approach
Limitations of this Review introduces selection bias, these studies usefully assess
Other limitations of this Review include the wide the effectiveness of traditional healers, for those who
heterogeneity of studies, which allows only broad choose them, compared with conventional treatments,
generalisations to be made. Despite some commonalities for those who choose them. This finding can inform
across different healing modalities, such as a shared clinicians about whether an intervention works for
spiritual perspective, cultural endorsement, and patients who choose it.90,91 Since patient expectations and
extensive use of rituals and ceremony, substantial beliefs are not easily changed or controlled for, these
differences in practice remain. For example, healing patient factors should be incorporated into evaluations
interventions might be brief or long, personal or of treatment effectiveness as treatment–patient
impersonal, inpatient or outpatient, and involve private interactions.55,60 This concept is not new: treatment–
ritual or public ceremony. A review cannot do justice to genotype–phenotype interactions are increasingly
each study, and interested readers are referred to the important considerations in clinical psychiatry.92
individual studies. Second, our literature search, Standardisation of treatment is similarly poor when
although extensive, might not be exhaustive since this studying traditional healing. An empathic traditional
topic extends across several disciplines and is not well healer, like a good psychotherapist, will adjust and tune
defined. Older studies in particular might not be their approach to the nuances of the client’s problem,
indexed in electronic databases and are more often rather than applying a standard treatment dictated by the
unavailable for potential inclusion. Finally, this Review symptoms.89,93 In addition to this within-session
does not discuss the potential harms of traditional variability, treatment styles and techniques differ from
healing, which would normally form part of a review of region to region, and even from healer to healer within
effectiveness. Although numerous case reports exist the same village.12,94 Indeed, as is the case for studies of
describing harms such as beating, deprivation, and psychotherapy, different healers working within the
chaining, the studies included here did not report on same paradigm can achieve very different results.12,15,54
these harms. Manualisation of an intervention, or indeed placing most
forms of constraint on it, would probably interfere with
Challenges of assessing outcomes of traditional healers the process and outcome, and is strongly resisted by
Assessment of the outcomes of traditional healing is traditional healers themselves.84 Therefore, we should in
fraught with difficulties, and some have argued that principle extend our analysis to treatment–patient–healer
non-western healing should not be subjected to western interactions to obtain optimum predictive value.
standards of scientific evidence at all.85,86 Traditional
healing is an extreme example of a complex intervention, Implementation
which does not submit easily to the standard models of Patients routinely use both conventional medical services
evidence-based medicine even in relatively well-resourced and traditional healing when both are available, and to
ideal settings.87–89 Therefore, it is not surprising that the prevent such an approach would be both impractical and
studies reviewed here often fall short of rigorous unethical, which makes it difficult to distinguish
standards of acceptable evidence. Nevertheless, in the traditional from conventional healing effects.63,66 Language
interest of evidence-based practice, evaluations must be and cultural barriers might be substantial for non-native
made, however problematic they might be. Moreover, investigators.39,75 The choice of dialect and style of
assessment of traditional healing presents particular and translated questionnaires could prime nuances of status
unique challenges related to trial design, implementation, and hierarchy which can affect results,67 and to ask
and evaluation. participants about certain demographic variables, such as

166 www.thelancet.com/psychiatry Vol 3 February 2016


Review

age, might be a spiritual taboo.95 Other investigators have showed that subjective impressions of effectiveness do
found that to gain access to the private and sacred rituals not necessarily correlate with symptomatic
of traditional healing can be difficult for outsiders.38,56,96,97 improvement.55,61,62,102 Clearly, patients’ assessments of
More importantly, once trust and confidence are gained, effectiveness do not necessarily align with scores on
the presence of the researchers might affect the process psychiatric rating scales. Patients also judge success
and outcome of treatment.54 The healing experience of according to a healing system’s ability to provide internally
sacred and private rituals could be undermined by consistent and satisfying explanations for sickness and
outside scrutiny, and scientific researchers might bias misfortune,55,103 or by the ability to maintain social
responses.39,56,57 These effects are likely to be significant in connections and social structure.102 Implicit in the
traditional healing, which mainly relies on more malleable psychiatric approach is the assumption that successful
psychological and social processes.38 treatment involves a restoration to a previous level of
Profoundly different understandings of health and wellbeing or functioning.104 Waldram makes the useful
illness between traditional healers and modern medicine observation that many traditional healing paradigms,
confound attempts to create a patient group that conforms however, facilitate transformation of the person from
to both psychiatric and traditional diagnostic categories. illness to wellness, which might not involve curing the
Traditional diagnosis aims to find the ultimate meta- disease.104 Although the degree of restoration can be
physical cause of a disturbance, such as spiritual neglect or quantified in principle, transformation implies a
witchcraft, whereas psychiatric diagnosis focuses on the qualitative change that is difficult to measure (although
symptoms and behaviour, remaining steadfastly agnostic this has been attempted105). Such transformation from
regarding the ultimate cause. There is at best a tenuous illness to wellness, or from one state of illness to another,
link between diagnoses generated by these different is a crucial aspect of the so-called performative efficacy of
systems.37,98,99 This tension is evident in the reviewed traditional healing, by contrast with the fastidious efficacy
studies, which generally made little attempt to single out a of medicine that demands specific biological changes.106
psychiatric diagnosis for study and often grouped physical The performance of symbolic rituals rich in meaning and
and mental presenting complaints together. metaphor facilitate the patient’s transition, independent
Acknowledging this, we cast a wide net in this Review and of any change in symptoms.41,83 The healer’s ritualised
included studies with heterogeneous diagnoses or divination and explanation, followed by the prescribed
complaints. We believe that our approach is justified ceremonies, might substantially relieve distress and
because a high proportion of complaints brought to prime positive expectations. Similarly, of course, the
traditional healers are psychological in nature47 and many conventional doctor’s ritualised examination and
patients with psychiatric problems present to traditional explanation, followed by the prescribed remedy, might
healers with somatised physical complaints.55,98,100 considerably relieve distress and prime positive
expectations long before any medication is swallowed.
Evaluation
Evaluation of the effectiveness of an intervention depends Conclusions
critically on the concept of effectiveness. Modern Despite the limitations of the published literature, it
psychiatry typically interprets effectiveness in terms of seems reasonable to conclude that many people,
symptom improvement as measured on a validated rating especially those with less severe complaints and positive
scale. This simplification strives for specificity, objectivity, expectations, derive subjective benefit from attending
and quantifiability, removing ambiguity and clutter while their chosen traditional or spiritual healers. Value seems
focusing on what are believed to be the core features. It to exist in attending a traditional healer for those who
focuses on malfunctioning biological or psychological choose to do so and who find the process meaningful.
processes—the disease—rather than the personal and This benefit might occur without a concomitant
social experience that constitutes the illness (using improvement in symptoms. The effective component of
Kleinman’s distinction101). Although undoubtedly healing is often unclear, although intensive regular social
powerful, this view makes particular culture-specific interventions generally achieve superior outcomes to
assumptions about what is most important. In doing so, brief single interventions. That said, little evidence is
it might ignore aspects of effectiveness that are especially available to suggest that traditional healers are effective
valuable and relevant in other cultures.41 Less specific and in changing the course of major psychiatric illnesses
more subjective outcomes, such as “satisfaction” or such as schizophrenia or obsessive-compulsive disorder.
“helpfulness”, as assessed by some of the reviewed Additionally, harms could be associated with traditional
studies, might capture other aspects of effectiveness that healing, which we have not reviewed here. For patients
are important to the patient. These outcomes could be who have cultural and spiritual beliefs not in line with
useful indicators of the patient’s self-perceived health those of conventional psychiatry, the possibility exists
status and ability to function, which are important social for more holistic care, and potential synergies, if
and economic outcomes.40 The distinction between illness collaboration between healing systems is facilitated. This
and disease is shown in some of the reviewed studies that topic should be an area of active research.

www.thelancet.com/psychiatry Vol 3 February 2016 167


Review

Contributors 17 Farooqi YN. Traditional healing practices sought by Muslim


SS and OG conceived the Review. GN wrote the first and final drafts. psychiatric patients in Lahore, Pakistan. Int J Disabil Dev Educ 2006;
GN and BO did the systematic search, selected papers, and extracted 53: 401–15.
data. SS, OG, and BO contributed comments and suggestions to the 18 Gadit AA. Shamanic concepts and treatment of mental illness in
paper. All authors reviewed the paper before submission. Pakistan. J Coll Physicians Surg Pak 1998; 8: 33–35.
19 Beals J, Novins DK, Whitesell NR, Spicer P, Mitchell CM,
Declaration of interests Manson SM. Prevalence of mental disorders and utilization of
OG has received grants from the National Institute of Mental Health to mental health services in two American Indian reservation
study collaboration between conventional and alternative medicine and populations: mental health disparities in a national context.
conventional biomedicine. SS, GN, and BO declare no competing Am J Psychiatry 2005; 162: 1723–32.
interests. 20 Hartmann WE, Gone JP. Incorporating traditional healing into an
urban American Indian health organization: a case study of
Acknowledgments
community member perspectives. J Couns Psychol 2012;
This Review was supported by the National Institute of Mental Health 59: 542–54.
grant supporting Partnerships for Mental Health Development in
21 Walls ML, Johnson KD, Whitbeck LB, Hoyt DR. Mental health and
Sub-Saharan Africa (grant number 1U19MH098718–01), the South substance abuse services preferences among American Indian
African Research Chair in Post-Traumatic Stress Disorder hosted by people of the northern Midwest. Community Ment Health J 2006;
Stellenbosch University and funded by the Department of Science and 42: 521–35.
Technology, and the National Research Foundation, South Africa. These 22 Lin TY. Psychiatry and Chinese culture. West J Med 1983;
funders had no role in the methods, preparation, writing, or decision to 139: 862–67.
publish the Review. We thank Joan Koss-Chioino for valuable comments 23 Furnham A, Raja N, Khan UA. A cross-cultural comparison of
on an early draft of the report. British and Pakistani medical students’ understanding of
schizophrenia. Psychiatry Res 2008; 159: 308–19.
References
1 Alem A, Jacobsson L, Araya M, Kebede D, Kullgren G. How are 24 Dein S, Alexander M, Napier AD. Jinn, psychiatry and contested
mental disorders seen and where is help sought in a rural Ethiopian notions of misfortune among east London Bangladeshis.
community? A key informant study in Butajira, Ethiopia. Transcult Psychiatry 2008; 45: 31–55.
Acta Psychiatr Scand Suppl 1999; 397: 40–47. 25 Assion HJ, Dana I, Heinemann F. Folk medical practices in
2 Familiar I, Sharma S, Ndayisaba H, Munyentwari N, Sibomana S, psychiatric patients of Turkish origin in Germany.
Bass JK. Community perceptions of mental distress in a Fortschr Neurol Psychiatr 1999; 67: 12–20 [in German].
post-conflict setting: a qualitative study in Burundi. 26 Padilla R, Gomez V, Biggerstaff SL, Mehler PS. Use of
Glob Public Health 2013; 8: 943–57. curanderismo in a public health care system. Arch Intern Med 2001;
3 Galabuzi C, Agea JG, Fungo BL, Kamoga RMN. Traditional 161: 1336–40.
medicine as an alternative form of health care system: 27 Buchwald D, Panwala S, Hooton TM. Use of traditional health
a preliminary case study of Nangabo sub-county, central Uganda. practices by Southeast Asian refugees in a primary care clinic.
Afr J Tradit Complement Altern Med 2010; 7: 11–16. West J Med 1992; 156: 507–11.
4 Girma E, Tesfaye M. Patterns of treatment seeking behavior for 28 Khalifa N, Hardie T. Possession and jinn. J R Soc Med 2005;
mental illnesses in Southwest Ethiopia: a hospital based study. 98: 351–53.
BMC Psychiatry 2011; 11: 138. 29 Demyttenaere K, Bruffaerts R, Posada-Villa J, et al, and the WHO
5 Kapur RL. The role of traditional healers in mental health care in World Mental Health Survey Consortium. Prevalence, severity, and
rural India. Soc Sci Med Med Anthropol 1979; 13B: 27–31. unmet need for treatment of mental disorders in the World Health
6 Khan TM, Sulaiman SAS, Hassali MA, Anwar M, Wasif G, Organization World Mental Health Surveys. JAMA 2004;
Khan AH. Community knowledge, attitudes, and beliefs towards 291: 2581–90.
depression in the state of Penang, Malaysia. 30 Ae-Ngibise K, Cooper S, Adiibokah E, Akpalu B, Lund C, Doku V,
Community Ment Health J 2010; 46: 87–92. and the Mhapp Research Programme Consortium. ‘Whether you
7 Kurihara T, Kato M, Reverger R, Tirta IGR. Pathway to psychiatric like it or not people with mental problems are going to go to them’:
care in Bali. Psychiatry Clin Neurosci 2006; 60: 204–10. a qualitative exploration into the widespread use of traditional and
faith healers in the provision of mental health care in Ghana.
8 Latypov A. Healers and psychiatrists: the transformation of
Int Rev Psychiatry 2010; 22: 558–67.
mental health care in tajikistan. Transcult Psychiatry 2010;
47: 419–51. 31 Sorsdahl K, Stein DJ, Flisher AJ. Traditional healer attitudes and
beliefs regarding referral of the mentally ill to Western doctors in
9 Makanjuola AB, Adelekan ML, Morakinyo O. Current status of
South Africa. Transcult Psychiatry 2010; 47: 591–609.
traditional mental health practice in Ilorin Emirate Council area,
Kwara State, Nigeria. West Afr J Med 2000; 19: 43–49. 32 Durie M. Maori knowledge and medical science: The interface
between psychiatry and traditional healing in New Zealand.
10 Mbwayo AW, Ndetei DM, Mutiso V, Khasakhala LI. Traditional healers
In: Incayawar M, Wintrob R, Bouchard L, eds. Psychiatrists and
and provision of mental health services in cosmopolitan informal
traditional healers: unwitting partners in global mental health.
settlements in Nairobi, Kenya. Afr J Psychiatry 2013; 16: 134–40.
Chichester: John Wiley & Sons, Ltd, 2009.
11 Razali SM. Complementary treatment of mental illness in
33 Sorsdahl K, Stein DJ, Flisher AJ. Predicting referral practices of
Southeast Asia. Int Med J 2000; 7: 189–91.
traditional healers of their patients with a mental illness: an
12 Salan R, Maretzki T. Mental health services and traditional healing application of the theory of planned behaviour. Afr J Psychiatry 2013;
in Indonesia: are the roles compatible? Cult Med Psychiatry 16: 35–40.
1983; 7: 377–412.
34 Macfarlane JE, Alpers MP. Treatment-seeking behaviour among the
13 Varghese KJ, Gopal B, Thomas TM. Revisiting psychotherapeutic Nasioi people of Bougainville: choosing between traditional and
practices in Karnataka, India: lessons from indigenous healing western medicine. Ethn Health 2009; 14: 147–68.
methods. Int J Health Promot Educ 2011; 49: 128–36.
35 Manson SM. Mental health services for American Indians and
14 Wig NN, Suleiman MA, Routledge R, et al. Community reactions to Alaska Natives: need, use, and barriers to effective care.
mental disorders. A key informant study in three developing Can J Psychiatry 2000; 45: 617–26.
countries. Acta Psychiatr Scand 1980; 61: 111–26.
36 Csordas TJ, Lewton E. Practice, performance, and experience in
15 WHO. General guidelines for methodologies on research and ritual healing. Transcult Psychiatry 1998; 35: 435–512.
evaluation of traditional medicine. 2000. http://whqlibdoc.who.int/
37 Csordas TJ, Storck MJ, Strauss M. Diagnosis and distress in Navajo
hq/2000/WHO_EDM_TRM_2000.1.pdf (accessed April 13, 2014).
healing. J Nerv Ment Dis 2008; 196: 585–96.
16 Bhikha AG, Farooq S, Chaudhry N, Husain N. A systematic review
38 Portman TAA, Garrett MT. Native American healing traditions.
of explanatory models of illness for psychosis in developing
Int J Disabil Dev Educ 2006; 53: 453–69.
countries. Int Rev Psychiatry 2012; 24: 450–62.

168 www.thelancet.com/psychiatry Vol 3 February 2016


Review

39 Devisch R, Dimomfu L. Le Roy J, Crossman P. A community-action 65 Harding T. Psychosis in a rural West African community.
intervention to improve medical care services in Kinshasa, Congo: Soc Psychiatry 1973; 8: 198–203.
mediating the realms of healers and physicians. in: applying health 66 Boelens PA, Reeves RR, Replogle WH, Koenig HG. The effect of
social science: best practice in the developing world. London: prayer on depression and anxiety: maintenance of positive influence
Zed Books, 2001. one year after prayer intervention. Int J Psychiatry Med 2012; 43: 85–98.
40 Finkler K. Non-medical treatments and their outcomes. 67 Lemelson RB. Traditional healing and its discontents: efficacy and
Cult Med Psychiatry 1980; 4: 271–310. traditional therapies of neuropsychiatric disorders in Bali.
41 Kirmayer LJ. The cultural diversity of healing: meaning, metaphor Med Anthropol Q 2004; 18: 48–76.
and mechanism. Br Med Bull 2004; 69: 33–48. 68 Salib E, Youakim S. Spiritual healing in elderly psychiatric patients:
42 Patel V, Todd C, Winston M, et al. Outcome of common mental a case-control study in an Egyptian psychiatric hospital.
disorders in Harare, Zimbabwe. Br J Psychiatry 1998; 172: 53–57. Aging Ment Health 2001; 5: 366–70.
43 Kleinman A. Culture and depression. N Engl J Med 2004; 69 Mirza I, Mujtaba M, Chaudhry H, Jenkins R. Primary mental
351: 951–53. health care in rural Punjab, Pakistan: providers, and user
44 Steel Z, Silove D, Giao NM, et al. International and indigenous perspectives of the effectiveness of treatments. Soc Sci Med 2006;
diagnoses of mental disorder among Vietnamese living in Vietnam 63: 593–97.
and Australia. Br J Psychiatry 2009; 194: 326–33. 70 Ensink K, Robertson B. Patient and family experiences of
45 Incayawar M. Efficacy of Quichua healers as psychiatric psychiatric services and African indigenous healers.
diagnosticians. Br J Psychiatry 2008; 192: 390–91. Transcult Psychiatry 1999; 36: 23–43.
46 WHO. WHO traditional medicine strategy: 2014–2023. 2002. 71 Bhana K. Indian indigenous healers. S Afr Med J 1986; 70: 221–23.
http://www.who.int/medicines/publications/traditional/trm_ 72 MacLaren D, Asugeni J, Asugeni R, Kekeubata E. Incorporating
strategy14_23/en/ (accessed May 11, 2014). sociocultural beliefs in mental health services in Kwaio, Solomon
47 Robertson BA. Does the evidence support collaboration between Islands. Australas Psychiatry 2009; 17 (suppl 1): S125–27.
psychiatry and traditional healers? Findings from three South 73 Campion J, Bhugra D. Experiences of religious healing in
African studies. S Afr Psychiatry Rev 2006; 9: 87–90. psychiatric patients in south India.
48 Cloninger CR. The science of well-being: an integrated approach to Soc Psychiatry Psychiatr Epidemiol 1997; 32: 215–21.
mental health and its disorders. World Psychiatry 2006; 5: 71–76. 74 Salem MO, Saleh B, Yousef S, Sabri S. Help-seeking behaviour of
49 Dixon M. Does ‘healing’ benefit patients with chronic symptoms? patients attending the psychiatric service in a sample of United
A quasi-randomized trial in general practice. J R Soc Med 1998; Arab Emirates population. Int J Soc Psychiatry 2009; 55: 141–48.
91: 183–88. 75 Raguram R, Venkateswaran A, Ramakrishna J, Weiss MG.
50 Boelens PA, Reeves RR, Replogle WH, Koenig HG. A randomized Traditional community resources for mental health: a report of
trial of the effect of prayer on depression and anxiety. temple healing from India. BMJ 2002; 325: 38–40.
Int J Psychiatry Med 2009; 39: 377–92. 76 Hewson P, Rowold J, Sichler C, Walter W. Are healing ceremonies
51 Mehl-Madrona LE. Native American medicine in the treatment of useful for enhancing quality of life? J Altern Complement Med 2014;
chronic illness: developing an integrated program and evaluating its 20: 713–17. . DOI:10.1089/acm.2013.0248.
effectiveness. Altern Ther Health Med 1999; 5: 36–44. 77 Barry CA. The role of evidence in alternative medicine: contrasting
52 Halliburton M. Finding a fit: psychiatric pluralism in south India biomedical and anthropological approaches. Soc Sci Med 2006;
and its implications for WHO studies of mental disorder. 62: 2646–57.
Transcult Psychiatry 2004; 41: 80–98. 78 Reeves S, Kuper A, Hodges BD. Qualitative research
53 Hurst GA, Williams MG, King JE, Viken R. Faith-based intervention methodologies: ethnography. BMJ 2008; 337: a1020.
in depression, anxiety, and other mental disturbances. South Med J 79 National Institute for Clinical Excellence. NICE Guidelines.
2008; 101: 388–92. www.nice.org.uk (accessed Nov 16, 2014).
54 Kleinman A, Gale JL. Patients treated by physicians and folk 80 Johansson P, Høglend P, Hersoug AG. Therapeutic alliance
healers: a comparative outcome study in Taiwan. mediates the effect of patient expectancy in dynamic psychotherapy.
Cult Med Psychiatry 1982; 6: 405–23. Br J Clin Psychol 2011; 50: 283–97.
55 Koss JD. Expectations and outcomes for patients given mental 81 Frank JD, Frank JB. Persuasion and healing: a comparative study of
health care or spiritist healing in Puerto Rico. Am J Psychiatry 1987; psychotherapy. Baltimore: Johns Hopkins University Press, 1993.
144: 56–61. 82 Krippner S. Shamans as healers, counselors, and psychotherapists.
56 Schiff JW, Moore K. The impact of the sweat lodge ceremony on Int J Transpers Stud 2012; 31: 72–79.
dimensions of well-being. 83 Csordas TJ. Elements of charismatic persuasion and healing.
Am Indian Alsk Native Ment Health Res 2006; 13: 48–69. Med Anthropol Q 1988; 2: 121–42.
57 Zacharias S. Mexican Curanderismo as ethnopsychotherapy: 84 Mehl-Madrona L. What traditional indigenous elders say about
a qualitative study on treatment practices, effectiveness, and cross-cultural mental health training. Explore (NY) 2009; 5: 20–29.
mechanisms of change. Int J Disabil Dev Educ 2006; 53: 381–400. 85 Tangwa GB. How not to compare Western scientific medicine
58 Mainguy B, Valenti Pickren M, Mehl-Madrona L. Relationships with African traditional medicine. Dev World Bioeth
between level of spiritual transformation and medical outcome. 2007; 7: 41–44.
Adv Mind Body Med 2013; 27: 4–11. 86 Omonzejele PF. African concepts of health, disease, and treatment:
59 Peltzer K, Machleidt W. A traditional (African) approach towards the an ethical inquiry. Explore (NY) 2008; 4: 120–26.
therapy of schizophrenia and its comparison with western models. 87 Craig P, Dieppe P, Macintyre S, Michie S, Nazareth I, Petticrew M.
Ther Communities 1992; 13: 229–42. Developing and evaluating complex interventions: the new Medical
60 Wirth DP. The significance of belief and expectancy within the Research Council guidance. BMJ 2008; 337: a1655.
spiritual healing encounter. Soc Sci Med 1995; 41: 249–60. 88 Haynes B. Can it work? Does it work? Is it worth it? The testing of
61 Kleinman A, Sung LH. Why do indigenous practitioners healthcare interventions is evolving. BMJ 1999; 319: 652–53.
successfully heal? Soc Sci Med 1979; 13B: 7–26. 89 Seligman ME. The effectiveness of psychotherapy. The Consumer
62 Lee B-O, Kirmayer LJ, Groleau D. Therapeutic processes and Reports study. Am Psychol 1995; 50: 965–74.
perceived helpfulness of dang-ki (Chinese shamanism) from the 90 Howard L, Thornicroft G. Patient preference randomised
symbolic healing perspective. Cult Med Psychiatry 2010; 34: 56–105. controlled trials in mental health research. Br J Psychiatry 2006;
63 Abbo C, Okello ES, Musisi S, Waako P, Ekblad S. Naturalistic 188: 303–04.
outcome of treatment of psychosis by traditional healers in Jinja 91 Kowalski CJ, Mrdjenovich AJ. Patient preference clinical trials: why
and Iganga districts, Eastern Uganda – a 3- and 6 months follow up. and when they will sometimes be preferred. Perspect Biol Med 2013;
Int J Ment Health Syst 2012; 6: 13. 56: 18–35.
64 Sorketti EA, Zainal NZ, Habil MH. The treatment outcome of 92 McMahon FJ, Insel TR. Pharmacogenomics and personalized
psychotic disorders by traditional healers in central Sudan. medicine in neuropsychiatry. Neuron 2012; 74: 773–76.
Int J Soc Psychiatry 2013; 59: 365–76.

www.thelancet.com/psychiatry Vol 3 February 2016 169


Review

93 Westen D, Novotny CM, Thompson-Brenner H. The empirical 100 Clarke K, Saville N, Bhandari B, et al. Understanding psychological
status of empirically supported psychotherapies: assumptions, distress among mothers in rural Nepal: a qualitative grounded
findings, and reporting in controlled clinical trials. Psychol Bull theory exploration. BMC Psychiatry 2014; 14: 60.
2004; 130: 631–63. 101 Kleinman A. Concepts and a model for the comparison of medical
94 Gureje O, Nortje G, Makanjuola V, Oladeji BD, Seedat S, Jenkins R. systems as cultural systems. Soc Sci Med 1978; 12: 85–95.
The role of global traditional and complementary systems of 102 Finkler K. Non-medical treatments and their outcomes. Part two:
medicine in the treatment of mental health disorders. focus on adherents of spiritualism. Cult Med Psychiatry
Lancet Psychiatry 2015; 2: 168–77. 1981; 5: 65–103.
95 Helsel DG, Mochel M, Bauer R. Shamans in a Hmong American 103 Stock R. Traditional healers in rural Hausaland. GeoJournal
community. J Altern Complement Med 2004; 10: 933–38. 1981; 5: 363–68.
96 Al-Krenawi A, Graham JR. Spirit possession and exorcism in the 104 Waldram JB. Transformative and restorative processes: revisiting
treatment of a bedouin psychiatric patient. Clin Soc Work J 1997; the question of efficacy of indigenous healing. Med Anthropol 2013;
25: 211–22. 32: 191–207.
97 Sexton R, Stabbursvik EAB. Healing in the Sámi North. 105 Bishop FL, Barlow F, Walker J, McDermott C, Lewith GT. The
Cult Med Psychiatry 2010; 34: 571–89. development and validation of an outcome measure for spiritual
98 Patel V, Simunyu E, Gwanzura F. The pathways to primary mental healing: a mixed methods study. Psychother Psychosom 2010;
health care in high-density suburbs in Harare, Zimbabwe. 79: 350–62.
Soc Psychiatry Psychiatr Epidemiol 1997; 32: 97–103. 106 Kaptchuk TJ. The placebo effect in alternative medicine: can the
99 Saeed K, Gater R, Hussain A, Mubbashar M. The prevalence, performance of a healing ritual have clinical significance?
classification and treatment of mental disorders among attenders of Ann Intern Med 2002; 136: 817–25.
native faith healers in rural Pakistan.
Soc Psychiatry Psychiatr Epidemiol 2000; 35: 480–85.

170 www.thelancet.com/psychiatry Vol 3 February 2016

Das könnte Ihnen auch gefallen