Vanuatu kastom healing

1. Introduction

Version followed: the Bislama healer categories used by Taylor and Elliott (Devpolicy, 20 February 2017) and Elliott and Taylor (Health Promotion International, 2021) for urban Espiritu Santo, supplemented by Tudal et al. (IRD and National University of Vanuatu, 2025) on kastom meresin; the psychotherapist's own work follows the DSM-5-TR Cultural Formulation Interview (CFI) and the explanatory-model approach of Kleinman, Eisenberg and Good (1978).

What it is. Vanuatu kastom healing is not a single therapy but part of a pluralist health landscape in which people combine Indigenous (kastom), Christian and biomedical care, often in sequence or at the same time [source: https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/]. Taylor and Elliott replaced the umbrella term "traditional healer", which they judged "locally inappropriate", with local categories: man blo prayer (loosely, Christian faith healer), man blo lif (loosely, herbalist), kleva (loosely, diviner), plus massage, bone setters and self-administered treatment [source: https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/]. Tudal et al. describe plant medicine as kastom meresin or lif meresin, report that it is recognised in Vanuatu's Health Practitioners Act, and describe kleva as people known for broad plant knowledge and for treating serious illness, sometimes associated with supernatural abilities; they note that the line between expert and non-expert is blurry [source: https://univ.edu.vu/attachments/article/410/Art.Vulgarisation_Final_EN.pdf]. The two sources therefore weight the meaning of kleva differently (diviner vs. expert healer); use the client's own meaning.

Scale and pattern of use. In a 2015 survey of 313 adults in Luganville, 81.5% had used a non-biomedical provider in the previous 12 months and 65.5% had used formal health services; faith healers had the highest user satisfaction (96.3% positive) and herbalists the lowest (85.4%); 62.6% considered sorcery (black magic) a plausible cause of diabetes, and for sorcery-related illness more respondents trusted faith healers (58%) than formal services (48.8%) [source: https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/]. Elliott and Taylor conclude that sociocultural, physical and clinical causes are "interwoven" in local understanding and that services should recognise and support local belief systems and informal care rather than ignore them [source: https://researchonline.jcu.edu.au/70377/].

Who holds authority. Healing knowledge is held by recognised healers and families; Tudal et al. focused on healers "recognised by their community" who treat people beyond their own families, found most knowledge-holders interviewed were women, and report that some knowledge is taboo or reserved for particular communities [source: https://univ.edu.vu/attachments/article/410/Art.Vulgarisation_Final_EN.pdf]. Christian faith healing sits within churches. Chiefs and community leaders hold local authority, including in disputes about sorcery [source: https://www.stopsorceryviolence.org/justice-in-vanuatu-for-sorcery-violence-crime/]. Research on kastom knowledge requires national permits from cultural, environmental and forestry bodies [source: https://univ.edu.vu/attachments/article/410/Art.Vulgarisation_Final_EN.pdf]. The specific published role of the Vanuatu Cultural Centre (Vanuatu Kaljoral Senta) and of national chiefly bodies in regulating healing knowledge is [uncertain: not verified].

How change happens (the psychotherapist's frame). A psychotherapist does not practise kastom healing, divination or faith healing. Change in the therapist's work comes from understanding the client's own explanatory model and help-seeking pathway, making biomedical and psychological care fit with it, collaborating with community-sanctioned healers when the client wishes, and protecting clients from harm, especially sorcery accusation [synthesis, built on https://researchonline.jcu.edu.au/70377/ and https://www.psychiatry.org/File%20Library/Psychiatrists/Practice/DSM/DSM-5-TR/APA-DSM5TR-CulturalFormulationInterview.pdf].

2. Framework

This is a framework for the psychotherapist's own culturally responsive work alongside kastom, church and biomedical care, not a manualised protocol. Phase structure, session counts and transition criteria below are [synthesis] unless cited.

Phase A: Engagement and cultural assessment (sessions 1-2). Goal: understand the problem in the client's terms and map every source of help already used. Tools: the core CFI (16 questions, four domains) and, where needed, its Informant Version and supplementary modules [source: https://www.psychiatry.org/File%20Library/Psychiatrists/Practice/DSM/DSM-5-TR/APA-DSM5TR-CulturalFormulationInterview.pdf]; Kleinman's explanatory-model questions [source: https://tanenbaum.org/wp-content/uploads/2021/06/HANDOUT-Explanatory-Models-of-Illness.pdf]. Time for the core CFI is not given in the APA forms; a field trial found later interviews "required less time" [source: https://experts.umn.edu/en/publications/feasibility-acceptability-and-clinical-utility-of-the-cultural-fo/]; any minute estimate is [uncertain]. Move on when you can state the client's name for the problem, perceived cause (including any kastom or sorcery attribution), all healers consulted, and current risks.

Phase B: Differential assessment and safety (sessions 1-3, in parallel). Goal: distinguish culturally shared belief and experience from disorder, and screen for medical urgency, physical-practice injury, herb or medicine interactions, and accusation-related risk. Move on when risk is assessed and urgent medical needs are routed.

Phase C: Shared formulation and plan (session 2-4). Goal: a formulation that holds the client's model and yours side by side, and a plan that states what each source of help (prayer, kastom medicine, clinic, therapy) is for. Move on when the client agrees the plan and any collaboration with healers has consent.

Phase D: Treatment alongside other care (variable; typically 6-20 sessions of whatever evidence-based therapy fits the disorder). Goal: deliver psychological treatment while revisiting the help-seeking map at each review; the CFI modules may be repeated during treatment [source: https://psychiatry.org/File%20Library/Psychiatrists/Practice/DSM/APA_DSM5_Cultural-Formulation-Interview-Supplementary-Modules.pdf].

Phase E: Integration and ending. Goal: help the client make sense of healing experiences from all sources, plan relapse prevention that includes trusted community supports, and close with explicit routes back.

Safeguarding runs through every phase (see sections 4 and 5 and the Harm account).

3. Guiding principles

Pluralism is normal, not non-adherence. Expect concurrent and sequential use of prayer, kastom medicine and clinics; ask about all of them every review [source: https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/].
Use the client's categories. Say man blo prayer, man blo lif, kleva or the client's own term, not "traditional healer" [source: https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/].
Causes are interwoven. A client can hold sorcery, diet and genetics as causes at once; work with all of them rather than replacing one [source: https://researchonline.jcu.edu.au/70377/].
Neither affirm nor dismiss spiritual causation. Record sorcery, kastom or posen attributions as the client's explanation; do not confirm or argue [synthesis].
You are not a healer in this system. Never perform, imitate or arrange prayer healing, divination, plant medicine or massage; refer within community pathways when the client wishes [synthesis].
Respect reserved knowledge. Do not ask clients or healers to disclose taboo or community-reserved knowledge, plant identities or methods [source: https://univ.edu.vu/attachments/article/410/Art.Vulgarisation_Final_EN.pdf].
Cost and distance do not explain everything. Explore meaning, trust and moral expectations (e.g. concern that high fees interfere with healing) before assuming access barriers [source: https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/].
Never name or endorse a suspected sorcerer. Any action that identifies a person as the cause of illness or death can trigger violence [source: https://openresearch-repository.anu.edu.au/bitstream/1885/228693/1/DPA%20In%20Brief%2020216%20Forsyth%20et%20al.pdf].
Medicines go to the prescriber. Never advise starting, stopping or changing a medicine or remedy; route questions to the prescriber [synthesis].
Touch only with consent. For massage and bone-setting, the client's consent before and during, the right to stop, and never sexual or genital touch apply; physical injuries go to medical care [synthesis].
The CFI supports, not decides, diagnosis. It is not meant to be the only basis for a diagnosis [source: https://www.psychiatry.org/File%20Library/Psychiatrists/Practice/DSM/DSM-5-TR/APA-DSM5TR-CulturalFormulationInterview.pdf].
Collaboration has evidence when done carefully. Collaborative care between healers and health workers reduced psychosis symptoms in Nigeria and Ghana (COSIMPO) [source: https://pubannotation.org/docs/sourcedb/PubMed/sourceid/32861306]; adding medication in a Ghanaian prayer camp reduced symptoms but not days in chains [source: https://www.mentalhealthdataprizeafrica.aphrc.org/repository/paper/33405/joining-psychiatric-care-and-faith-healing-in-a-prayer-camp-in-ghana-randomised-trial]. Neither trial was in Vanuatu; transfer is [synthesis].
Safety overrides collaboration. Where a practice involves restraint, violence, accusation or delayed urgent care, safeguarding comes first [synthesis].

4. Intervention inventory

All interventions below are the psychotherapist's own work. None of them is kastom healing, divination or faith healing, and none describes how those are done. Example language is illustrative unless quoted.

4.1 Cultural humility and self-examination

Purpose / when: Before and throughout work with any ni-Vanuatu client or family; especially when you notice reactions to sorcery talk, prayer healing or plant medicine.
Steps: (1) Before the first session, note your own assumptions about sorcery, church healing and herbal medicine, and about Pacific Islanders as service users. (2) Remind yourself that most people combine sources of care and that faith healers are highly rated by their users [source: https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/]. (3) In supervision, review any moment you felt the urge to correct or endorse a belief. (4) Acknowledge openly what you do not know about the client's island, language and church [synthesis].
Example language (illustrative): "I don't know much about how things are done on your island. I'd like you to teach me what matters, and tell me if I get something wrong."
Watch for: Clients giving short or polite answers about healers (a sign they expect judgement); your own impatience with explanations that do not match yours. Working: the client volunteers details about prayer, kastom medicine or sorcery worries without being asked twice.

4.2 Core Cultural Formulation Interview (DSM-5-TR)

Purpose / when: First assessment with every client; the APA describes it as designed for "any clinician" with "any individual in any care setting" [source: https://www.psychiatry.org/File%20Library/Psychiatrists/Practice/DSM/DSM-5-TR/APA-DSM5TR-SectionIII.pdf].
Steps: (1) Use the official APA form; clinicians may reproduce it for use with their patients, but modification or translation needs APA permission [source: https://www.psychiatry.org/File%20Library/Psychiatrists/Practice/DSM/DSM-5-TR/APA-DSM5TR-CulturalFormulationInterview.pdf]. (2) Work through the four domains in order: cultural definition of the problem; cultural perceptions of cause, context and support (including cultural identity); cultural factors affecting self-coping and past help seeking; cultural factors affecting current help seeking, including the clinician-patient relationship [source: https://www.psychiatry.org/File%20Library/Psychiatrists/Practice/DSM/DSM-5-TR/APA-DSM5TR-CulturalFormulationInterview.pdf]. (3) Under past help seeking, ask specifically about each local category: church prayer, man blo prayer, man blo lif, kleva, massage, bone setters, leaves used at home [synthesis, categories from https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/]. (4) Record answers in the client's words, including Bislama terms. (5) Use the CFI to inform, not settle, diagnosis.
Example language (illustrative): "Sometimes people see a pastor, someone who uses leaves, or a kleva, as well as the clinic. Who have you or your family gone to for this?"
Watch for: The client switching to talk of family conflict or land when asked about causes (often relevant to sorcery fears). Working: you can describe the problem as the client would describe it to their family.

4.3 CFI Informant Version and supplementary modules

Purpose / when: When family members are central to decisions, when the client is too unwell to answer, or when one domain needs depth. The Informant Version (17 questions) is used with a family member or friend [source: https://psychiatry.org/File%20Library/Psychiatrists/Practice/DSM/APA_DSM5_Cultural-Formulation-Interview-Informant.pdf]. Of the 12 supplementary modules, the most relevant here are Explanatory Model; Spirituality, Religion and Moral Traditions; Coping and Help-Seeking; Social Network; Psychosocial Stressors; Cultural Identity; Immigrants and Refugees; School-Age Children and Adolescents; Caregivers [source: https://psychiatry.org/File%20Library/Psychiatrists/Practice/DSM/APA_DSM5_Cultural-Formulation-Interview-Supplementary-Modules.pdf].
Steps: (1) Decide which module the core CFI points to (the form shows which module expands each subtopic) [source: https://psychiatry.org/File%20Library/Psychiatrists/Practice/DSM/APA_DSM5_Cultural-Formulation-Interview-Supplementary-Modules.pdf]. (2) Get the client's consent before interviewing family. (3) Interview informants separately when sorcery suspicion or family conflict is possible, so no one is put on the spot [synthesis]. (4) Repeat modules later in treatment if the help-seeking map changes.
Example language (illustrative): "Your aunt is helping you a lot. Would it be all right if I asked her how the family understands what is happening?"
Watch for: Informant accounts that name a person as responsible for the illness; treat this as a safety signal (see 4.11). Working: family and client accounts converge on a shared plan.

4.4 Explanatory-model inquiry (Kleinman)

Purpose / when: To understand what the client calls the problem, what caused it, and what treatment they expect; essential where sorcery (blak masik), kastom or posen are possible causes [sources: https://tanenbaum.org/wp-content/uploads/2021/06/HANDOUT-Explanatory-Models-of-Illness.pdf; https://www.cambridge.org/core/product/F99D9D36838A8207D377730DEB445F7B/core-reader].
Steps: Ask, in your own words, the eight areas: the name for the problem; its cause; why it started when it did; what it does and how it works; how severe it is and how long it will last; what treatment should be given; the main problems it has caused; what is feared most [source: https://tanenbaum.org/wp-content/uploads/2021/06/HANDOUT-Explanatory-Models-of-Illness.pdf]. Then: (1) note any locally loaded terms (e.g. sik suka for diabetes, which Taylor and Elliott note may lead people to see sugar as the only cause [source: https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/]); (2) summarise the model back; (3) say where your model overlaps and where it adds something.
Example language (illustrative): "What do you call this sickness at home? What do you think started it, and why now? What are you most afraid it will do?"
Watch for: Fear that is about being harmed by another person rather than about symptoms; shame about having used kastom medicine. Working: the client corrects your summary, a sign they feel safe to disagree.

4.5 Mapping the pluralist help-seeking pathway

Purpose / when: At intake and every review, to see which care is used for which purpose and in what order, and to spot delays in urgent care [synthesis, grounded in https://researchonline.jcu.edu.au/70377/].
Steps: (1) Draw a simple timeline with the client: first sign of the problem, each person consulted, what each said, what was done, and what changed. (2) Note costs and who paid, since spending on non-biomedical care can exceed clinic costs [source: https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/]. (3) Ask what each source is good for (e.g. prayer for the cause, clinic for the body). (4) Agree a plan that keeps what the client values and adds what is missing. (5) Do not ask about remedies' ingredients, preparation or methods.
Example language (illustrative): "Let's put on one line everyone who has helped so far. What did each one do that helped, and what is still not better?"
Watch for: Long gaps before medical care for serious symptoms; a belief that the clinic cannot help if sorcery is the cause, which many respondents in Luganville held [source: https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/]. Working: the client can say what each source of help is for.

4.6 Cultural concepts of distress and the normative-versus-illness differential

Purpose / when: When a client describes sorcery attack, spirit involvement, trance or possession, or fear of being harmed by someone through posen, so that shared cultural belief is not mistaken for psychosis and real illness is not missed.
Steps: (1) Treat sorcery attribution as an explanation shared by many people in the community (62.6% in one Luganville survey considered sorcery a plausible cause of diabetes [source: https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/]), not by itself as a delusion. (2) Ask whether family and church members share the belief and the degree of conviction. (3) Use DSM-5-TR cultural concepts of distress as the manual frames them: idioms, explanations or perceived causes, and syndromes [source: https://www.psychiatry.org/File%20Library/Psychiatrists/Practice/DSM/DSM-5-TR/APA-DSM5TR-SectionIII.pdf]; no Vanuatu-specific concept is among the DSM-5-TR examples reported in secondary sources [source: https://www.unmc.edu/bhecn/_documents/dsm5tr_bw_6_slides_11_pages.pdf] [uncertain]. (4) For trance or possession: dissociative identity disorder can present as possession, but not when the experience is a normal part of broadly accepted cultural or religious practice [source: https://pmc.ncbi.nlm.nih.gov/articles/PMC9199574]; in DSM-5-TR the substance is retained though the exact wording is [uncertain] [source: https://umsystem.pressbooks.pub/aswbprep/?p=1299]. Dissociative trance (other specified dissociative disorder) likewise excludes trance that is a normal part of collective cultural or religious practice [source: https://pmc.ncbi.nlm.nih.gov/articles/PMC9199574]. In ICD-11, 6B62 trance disorder and 6B63 possession trance disorder require the state to be involuntary, unwanted and outside accepted cultural or religious practice, with significant distress or impairment [sources: https://www.findacode.com/icd-11/code-1807490186.html; https://pmc.ncbi.nlm.nih.gov/articles/PMC9199574]. (5) Look for features that go beyond the shared frame: idiosyncratic content no one else shares, bizarre elaboration, functional collapse, mood or sleep change, substance use, medical causes. (6) Consult a cultural broker from the client's community before concluding [synthesis]. A DSM-5-TR note on culturally sanctioned beliefs and psychotic symptoms is [uncertain: not verified].
Example language (illustrative): "Many people in your family think this could be sorcery. Do they see the same signs you do? Is there anything you notice that others don't?"
Watch for: Over-pathologising a shared belief; under-recognising psychosis because sorcery talk is common; fear of a named person (a safety flag, 4.11). Working: you can say which parts are shared and which are distress or illness.

4.7 Interpreters and cultural brokers

Purpose / when: When the client prefers Bislama, French, an island language or a mix; when assessing belief versus illness; when planning collaboration with healers or chiefs.
Steps: (1) Offer a professional interpreter; avoid using children or the person suspected in a family dispute. (2) Brief the interpreter on confidentiality and on not adding opinions about sorcery. (3) Ask a cultural broker (e.g. a trusted community health worker or church worker the client accepts) for context, with the client's consent. (4) Check that the broker has no stake in local disputes [synthesis throughout].
Example language (illustrative): "Would you like someone who speaks Bislama with us? You can choose someone from outside your community if that feels safer."
Watch for: The client censoring themselves in front of a broker from the same island; interpreters softening sorcery content. Working: richer, more specific answers once language support is in place.

4.8 Asking about healers and collaboration or referral when the client wishes

Purpose / when: When the client values a man blo prayer, man blo lif, kleva, massage or bone setter and wants their care joined up with yours.
Steps: (1) Ask without judgement what each healer offers and how the client feels after seeing them. (2) If the client wants contact, get written consent stating what may be shared (minimum necessary; never details that could identify someone as suspected of sorcery). (3) Approach the healer through the client or family, as a fellow carer, to agree roles: e.g. prayer and spiritual support from the church, symptom and risk monitoring from you and the clinic. (4) Agree warning signs for which the healer will send the client to the clinic. (5) Respect that some knowledge is reserved; do not ask about plants or methods [source: https://univ.edu.vu/attachments/article/410/Art.Vulgarisation_Final_EN.pdf]. (6) Review jointly over time. Basis: Tudal et al. recommend collaboration between traditional healers and health professionals [source: https://univ.edu.vu/attachments/article/410/Art.Vulgarisation_Final_EN.pdf]; the WHO Traditional Medicine Strategy 2014-2023 aims at safe, effective use and integration into health systems [source: https://www.who.int/activities/implementation-of-the-WHO-traditional-medicine-strategy-2014-2023]; COSIMPO (Gureje et al., The Lancet 2020; Nigeria and Ghana) found collaborative shared care between healers and primary care reduced psychosis symptoms more than usual care, and harmful practices fell from 57% to 9% in the intervention arm [sources: https://pubannotation.org/docs/sourcedb/PubMed/sourceid/32861306; https://centreforglobalmentalhealth.org/node/839]; in a Ghanaian prayer camp, adding medication lowered symptoms but did not significantly reduce days in chains (Br J Psychiatry 212(1)) [source: https://www.mentalhealthdataprizeafrica.aphrc.org/repository/paper/33405/joining-psychiatric-care-and-faith-healing-in-a-prayer-camp-in-ghana-randomised-trial]. Formal collaboration programmes in Vanuatu are [uncertain: not verified]; steps (2)-(4) are [synthesis].
Example language (illustrative): "Your pastor's prayers clearly matter to you. Would it help if we agreed together what the church does, what the clinic does and what we do here?"
Watch for: Any healer whose involvement includes identifying who caused the illness, restraint, beating or advice to stop clinic treatment: stop collaboration and move to safeguarding (4.11). Working: the client reports no conflict between sources of care and attends both.

4.9 Remedies, medicines and the prescriber

Purpose / when: Whenever the client uses leaves, plant medicine or kava and also takes, or is offered, prescribed medication.
Steps: (1) Ask neutrally whether the client uses leaves or remedies, without asking for ingredients or preparation. (2) Tell the client that some plants can affect the liver or interact with medicines; Tudal et al. judged two commonly cited child remedies possibly toxic (soursop and burao) and highlighted the hepatotoxicity risk of wael kava [source: https://univ.edu.vu/attachments/article/410/Art.Vulgarisation_Final_EN.pdf]. (3) With consent, inform the prescriber or GP. (4) Do not advise starting, stopping or changing any remedy or medicine; route to the prescriber. (5) For children's remedies, raise this with the child's doctor [synthesis].
Example language (illustrative): "Lots of families use leaves alongside clinic medicine. I'm not the right person to say what's safe to combine, but your doctor is. Shall we tell them what you use?"
Watch for: Yellowing skin, abdominal pain, unusual tiredness or confusion (refer medically the same day); stopping prescribed medicine because a healer advised it. Working: the prescriber knows about all remedies.

4.10 Massage and bone-setting: touch, consent and physical safety

Purpose / when: When the client has had, or plans, massage or bone-setting for injury, pain or a condition they link to distress. The therapist does not touch or manipulate; this is a screening and safety conversation.
Steps: (1) Ask what was done in general terms and how the client felt during and after. (2) Explore consent: whether the client could say stop, whether touch stayed away from intimate areas, whether they felt pressured. (3) Screen for warning signs after bone-setting and refer medically. A systematic review of 12 studies (833 participants, 691 complications) reports limb- and life-threatening complications of traditional bone-setter treatment, including mortality, and notes that bone setters' engagement with orthodox services and training can reduce complications [source: https://journals.sagepub.com/doi/ref/10.1177/00494755221122543]. That review covers low- and middle-income countries generally; Vanuatu-specific data are [uncertain]. (4) If any touch was sexual or non-consensual, follow 4.11 [synthesis].
Example language (illustrative): "When someone helps you with massage, you are always allowed to say stop, and no one should touch private parts of your body. Did it feel like that for you?"
Watch for: Swelling, colour change, coldness, numbness or rising pain in a limb after tight binding or setting; fever; deformity (urgent medical review). Disclosure of sexual touch. Working: the client knows the consent rule and seeks medical review for injuries.

4.11 Safeguarding and accusation-risk assessment

Purpose / when: Whenever an illness or death is attributed to a person's sorcery, a diviner or relative has named someone, the client fears being accused, or any practice involves restraint, beating, starvation, sexual contact or withholding urgent care. Research in Papua New Guinea found accusations followed a trigger (death 50%, sickness 37%) in 99% of cases, with pre-existing conflict in 82%, and that 19% of victims were identified by a diviner (glasman/glasmeri) [source: https://openresearch-repository.anu.edu.au/bitstream/1885/228693/1/DPA%20In%20Brief%2020216%20Forsyth%20et%20al.pdf].
Steps: (1) Ask directly whether anyone has been named as the cause of the illness or a recent death, and whether anyone is angry with or threatening that person. (2) Ask whether the client, or anyone in the family, has been accused before, is from another community or has little schooling; in PNG data these raised the risk of violence, as did multiple accused and accusations of causing death [source: https://openresearch-repository.anu.edu.au/bitstream/1885/228693/1/DPA%20In%20Brief%2020216%20Forsyth%20et%20al.pdf]. (3) If anyone is at imminent risk, follow your emergency and police procedures; in Vanuatu the Supreme Court has said people should complain to "the relevant authorities" rather than act themselves [source: https://www.stopsorceryviolence.org/justice-in-vanuatu-for-sorcery-violence-crime/]. (4) Apply child-protection duties where children are accused, harmed or witness violence. (5) Never repeat or record the name of a suspected sorcerer in a way that could be shared; document risk, not the accusation's content [synthesis]. (6) Do not mediate the accusation yourself; involve police, protection services or recognised human rights defenders where they exist [synthesis].
Example language (illustrative): "When someone dies or falls sick, sometimes people start to say a person caused it. Has anyone been blamed? Is anyone in danger because of that?"
Watch for: Funerals and the days after an unexpected death; talk of meetings to find who is responsible; a client who has been asked to confess. Working: a written safety plan exists and named supports know their role.

4.12 Working with sorcery fear and blame in therapy

Purpose / when: When the client is distressed by fear of sorcery attack, guilt over suspected wrongdoing, or anger at someone they believe harmed them.
Steps: (1) Validate the fear as real distress without confirming or disputing its cause. (2) Separate the questions: what the client believes caused the problem; what helps the symptoms now; what keeps everyone safe. (3) Use the client's own protective resources (prayer, church support, family) to reduce fear, and standard evidence-based techniques for anxiety, sleep and low mood. (4) Explore the conflict behind the fear (land, jealousy over money) as a problem in its own right, since such conflicts underlie most accusations in PNG data [source: https://openresearch-repository.anu.edu.au/bitstream/1885/228693/1/DPA%20In%20Brief%2020216%20Forsyth%20et%20al.pdf]. (5) If the client plans to confront or harm a suspected person, treat this as a risk-to-others situation (4.11) [synthesis].
Example language (illustrative): "Whatever the cause, you've been frightened for weeks and not sleeping. Let's work on how to get you some rest, and also think about how to keep everyone safe while this is going on."
Watch for: Rumination fixed on one person; family pressure to act; the client asking you whether sorcery is real (answer with respect for their view and your role). Working: fear reduces and the client can talk about the conflict without naming a culprit as the solution.

4.13 Working with family, church and community leaders

Purpose / when: When decisions are made collectively, when the client's church is central, or when chiefs or community leaders are involved in a dispute linked to illness.
Steps: (1) With consent, identify who in the family makes health decisions. (2) Invite them to a family meeting focused on the client's recovery plan. (3) Where church leaders are involved, ask what they offer and agree how prayer and clinic care fit. (4) Engage community leaders only through the client and family and never on questions of who caused an illness; in a 2014 Vanuatu case, community leaders organised meetings that led to the killing of two accused people [source: https://www.stopsorceryviolence.org/justice-in-vanuatu-for-sorcery-violence-crime/]. (5) PNG's 2014 Sorcery National Action Plan asks health workers to explain disease processes clearly to patients and their immediate families [source: https://stopsorceryviolence.org/the-sorcery-national-action-plan]; do the same for mental illness [synthesis].
Example language (illustrative): "Your family clearly cares a lot. Today I'd like us to agree how everyone can help her get well, and what each of you will do if she gets worse."
Watch for: Family members steering the meeting towards blame; a leader asking the therapist to confirm a cause. Working: the family speaks of illness and recovery rather than culprits.

4.14 Supporting timely care and adherence within pluralism

Purpose / when: When co-existing physical illness (e.g. diabetes) or severe mental illness needs regular clinic care and the client also uses faith or kastom healing. Taylor and Elliott note that these dynamics affect timely presentation, adherence and healthy lifestyles [source: https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/].
Steps: (1) Ask what the client believes the clinic can and cannot help with. (2) Frame clinic care as compatible with prayer or kastom medicine where the client sees it that way. (3) Agree specific signs that mean go to hospital now. (4) Use motivational interviewing and practical problem-solving for attendance. (5) Explain medical terms that may mislead (e.g. a term implying sugar is the only cause) in plain words [synthesis; term example from https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/].
Example language (illustrative): "You've told me prayer helps you feel strong. Many people pray and also take their clinic medicine. How would it be for you to do both?"
Watch for: Stopping medication after a healing experience; delay in presenting with chest pain, infection or limb injury. Working: attendance is stable and the client can name hospital warning signs.

4.15 Ni-Vanuatu clients living or working abroad

Purpose / when: For clients seen outside Vanuatu, including temporary workers and students, who may lack access to their usual healers and church.
Steps: (1) Use the CFI Immigrants and Refugees and Cultural Identity modules [source: https://psychiatry.org/File%20Library/Psychiatrists/Practice/DSM/APA_DSM5_Cultural-Formulation-Interview-Supplementary-Modules.pdf]. (2) Ask about contact with home: phone prayer, remedies sent from home, family news of illness or death, and accusations happening at home. (3) Identify local Pacific church or community support the client trusts. (4) Address work-visa fears that stop help-seeking [synthesis].
Example language (illustrative): "Being far from home with this is hard. Who do you talk to back home about it, and is there anyone here from your church or island you trust?"
Watch for: Distress after news of a death at home and talk of who is to blame; isolation. Working: the client has at least one trusted local support.

4.16 Integration and ending

Purpose / when: Near the end of therapy, or after the client has had a significant healing, church or kastom experience.
Steps: (1) Invite the client to tell what each source of help meant. (2) Link changes to the client's own efforts and supports, not only to one source. (3) Write a relapse plan that names trusted people (healer, pastor, family, clinic) and when to use each, plus hospital warning signs. (4) Leave a clear route back to you [synthesis].
Example language (illustrative): "Looking back, what helped most: the prayers, the leaves, the clinic, our talks? What will you do first if this starts again?"
Watch for: Attribution of all recovery to one source in a way that would stop needed medicine later. Working: the plan includes biomedical, spiritual and family elements the client endorses.

5. Handling problems

The sources cited describe health-seeking and harm patterns, not therapy problems; all responses below are [synthesis] unless cited.

Client minimises or hides healer use. Likely expectation of judgement. Normalise pluralism with the survey finding that most people use both kinds of care [source: https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/]; ask again later; never criticise a healer the client trusts. Say: "Most families I see use both. Nothing you tell me about that will change how I treat you." (illustrative)
Client asks you to say whether sorcery is real. State your role without dismissing theirs. Say: "I can't answer that from my training. What I can do is help with the fear and the sleeplessness, and help keep everyone safe." (illustrative)
Stuck: client believes therapy cannot help a sorcery-caused illness. Offer therapy for the effects (sleep, fear, grief, conflict) rather than the cause; ask what their pastor or healer would think of it; propose a time-limited trial with review.
Non-adherence to medication or clinic after a healing experience. Explore the meaning of the healing; do not advise on medicines; with consent bring the prescriber in; agree warning signs (4.14).
Overwhelm or acute fear of attack. Grounding, sleep and safety planning; involve trusted church or family supports; assess for psychosis and suicidality using standard risk assessment, including lethal-means counselling (storage and removal) where relevant.
Rupture: client feels judged after you referred medically. Name it, repair, and restate that medical referral adds to, rather than replaces, their other care.
Family pressure to name a culprit, or a client planning confrontation. Treat as risk to others; follow 4.11; do not mediate the accusation; involve authorities where there is danger [source: https://www.stopsorceryviolence.org/justice-in-vanuatu-for-sorcery-violence-crime/].
Client is the accused. Prioritise physical safety, housing and separation from threat; assess trauma, shame and suicidality; document injuries; connect with protection services, police and human rights defenders where available (see Harm account).
Healer advises stopping clinic care, uses restraint or harmful practice. Pause collaboration, explain concerns to the client, follow safeguarding duties. Collaboration trials report harmful practices can fall with shared care but do not disappear [source: https://pubannotation.org/docs/sourcedb/PubMed/sourceid/32861306]; a prayer-camp trial found medication did not significantly reduce days in chains [source: https://www.mentalhealthdataprizeafrica.aphrc.org/repository/paper/33405/joining-psychiatric-care-and-faith-healing-in-a-prayer-camp-in-ghana-randomised-trial].
Physical injury after massage or bone-setting. Same-day medical review for the warning signs in 4.10 [source: https://journals.sagepub.com/doi/ref/10.1177/00494755221122543]; if touch was sexual or non-consensual, safeguarding and support.
Interpreter or broker conflict of interest. Change interpreter; prefer someone from outside the client's community.
Cost conflicts. Clients may spend more on healers than on clinic visits and may hold that high fees interfere with healing [source: https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/]; discuss fees and financial strain openly, without criticising choices.
Therapist discomfort with sorcery content. Use supervision; check that discomfort is not pushing you towards either dismissal or over-identification.

Harm account

This section concerns sorcery and witchcraft accusations in Melanesia, in which divination and community processes for finding who caused an illness or death can feed violence. It is not an account of kastom healing as a whole, most of which involves no accusation.

What was actually done

The usual sequence (PNG data, 2016-2020). A death (50%) or sickness (37%) acted as a trigger in 99% of recorded cases, usually where tension already existed (82%), most often land disputes (35%) or jealousy over money and goods (26%). Victims were identified by people believed to have been harmed or their relatives (39%), by being seen near the person (22%), or by a diviner, glasman/glasmeri (19%; 46% in Jiwaka Province). In 30% of cases more than one person was accused; others were named by the first accused, were family members, or had tried to defend the first accused. Accusations followed locally dominant narratives, e.g. sanguma in Enga, where women were accused of taking or eating a heart, and poisoning (poisenman) in Bougainville, where victims were mostly men [source: https://openresearch-repository.anu.edu.au/bitstream/1885/228693/1/DPA%20In%20Brief%2020216%20Forsyth%20et%20al.pdf].
Diviners' role. A PNG National Research Institute issues paper (Forsyth et al., 2021) addresses glasman/glasmeri as catalysts of accusation and violence [source: https://openresearch-repository.anu.edu.au/items/c00070a7-b21b-45fe-b205-baf4e72b8d29/full]; its detailed findings are [uncertain]. A University of the Sunshine Coast ACPIR project (2023-24) on diviners and traditional healers in Pacific accusations notes glasman/glasmeri "are frequently reported to contribute" to accusations in PNG and that empirical research is limited [source: https://usc.edu.au/research/australian-centre-for-pacific-islands-research/research-students/acpir-srs/understanding-the-role-of-diviners-and-traditional-healers-in-witchcraft-and-sorcery-accusations-in-the-pacific]. Whether and how often kleva in Vanuatu identify suspects is [uncertain].
Vanuatu, Akham (Akaham) Island, South Malekula, November 2014. After three deaths, community leaders held three meetings to identify people believed to practise witchcraft; seven men were accused, five admitted and two denied; a community leader ordered the five to kill the two and supplied weapons, and with villagers' support at the community hall the two were tortured and killed [source: https://www.stopsorceryviolence.org/justice-in-vanuatu-for-sorcery-violence-crime/]. The killings reportedly took place in front of onlookers including children; police detained more than 40 people, including church pastors and customary chiefs [source: https://cathnews.co.nz/2014/11/21/sorcery-pastors-chiefs-arrested-hangings-vanuatu]. In April 2015 the Supreme Court (Chief Justice Vincent Lunabek) sentenced the five to 15 years and seven organisers to one year; chiefs and community leaders were not charged, which the judge criticised [source: https://www.stopsorceryviolence.org/justice-in-vanuatu-for-sorcery-violence-crime/].
Regional pattern. Forsyth (2013) describes public torture and killing of accused people across Melanesia, often of women "in highly sexualised ways", sometimes starting at funerals, with police at times standing by; men as well as women are accused [source: https://cdn.auckland.ac.nz/assets/humanrights/Research/HRWorkingpaper20-Forsyth.pdf]. Amnesty International names the Solomon Islands and Vanuatu alongside PNG as Melanesian countries affected by sorcery-related violence [source: https://www.amnesty.org.uk/files/glasgow_2014_media_briefing_-_sorcery_related_violence_0.pdf]; Solomon Islands case data are [uncertain].

How it caused harm

Violence and death. Accusations become violent more often when there are multiple accused, when death (rather than sickness) is alleged, when the accusation fits the dominant local narrative, and when the accused is from another community, unschooled or previously accused; in Enga the chance of violence nearly doubled after a prior accusation [source: https://openresearch-repository.anu.edu.au/bitstream/1885/228693/1/DPA%20In%20Brief%2020216%20Forsyth%20et%20al.pdf]. In Vanuatu, two accused men died of strangulation and heavy blood loss after torture [source: https://www.stopsorceryviolence.org/justice-in-vanuatu-for-sorcery-violence-crime/].
Spread to families. Accusation often extends to relatives and to those who defend the accused (family-based accusations reached 91% of multi-accused cases in Jiwaka) [source: https://openresearch-repository.anu.edu.au/bitstream/1885/228693/1/DPA%20In%20Brief%2020216%20Forsyth%20et%20al.pdf].
Children as witnesses. The Vanuatu killings were reportedly watched by children [source: https://cathnews.co.nz/2014/11/21/sorcery-pastors-chiefs-arrested-hangings-vanuatu]. PNG's 2014 plan calls for mechanisms for cases involving children [source: https://stopsorceryviolence.org/the-sorcery-national-action-plan].
Delay of care. When sorcery is suspected, more people in Luganville trusted non-biomedical providers for diagnosis (53%) and treatment (54.2%), and fewer (48.8%) thought formal services effective [source: https://devpolicy.org/knowledge-sickness-healing-medical-pluralism-health-seeking-vanuatu-20170220/]; this can delay treatable illness [synthesis]. Bone-setting complications include limb- and life-threatening harm when medical care is delayed [source: https://journals.sagepub.com/doi/ref/10.1177/00494755221122543].
Psychological sequelae. Survivors' trauma, displacement and stigma, and the effects on witnesses, are expected but specific clinical data for Vanuatu are [uncertain]; PNG's plan prioritises counselling and repatriation support for victims [source: https://stopsorceryviolence.org/the-sorcery-national-action-plan].
Who is most at risk. People already in dispute with the bereaved family, outsiders, the unschooled, those accused before, relatives of the accused, and (by narrative) women in some provinces and men in others [source: https://openresearch-repository.anu.edu.au/bitstream/1885/228693/1/DPA%20In%20Brief%2020216%20Forsyth%20et%20al.pdf].

Recognising it today

Features in a client's story: an unexpected death or illness followed by talk of who caused it; family or community meetings to find the person responsible; a diviner, healer or relative having named someone; long-standing land or money disputes; a person pressed to confess; a relative avoiding hospital because the cause is believed to be sorcery [synthesis, from patterns in https://openresearch-repository.anu.edu.au/bitstream/1885/228693/1/DPA%20In%20Brief%2020216%20Forsyth%20et%20al.pdf].
Phrases and rationales: someone is said to be the cause of a death; confession or punishment framed as needed to stop further deaths; the claim that state law does not deal with sorcery, which the Vanuatu Supreme Court rejected ("does not approve of killing of sorcerers") [source: https://www.stopsorceryviolence.org/justice-in-vanuatu-for-sorcery-violence-crime/].
Red flags in a colleague's or service's practice: confirming or speculating on who caused an illness; recommending divination; passing identifying details to community members; treating accusation as a family matter not needing safeguarding; dismissing a client's fear as "just superstition" so that risk is never assessed [synthesis].
How legitimate care differs: legitimate healers, pastors and therapists address the sick person's needs without naming a culprit, and refer to clinics for serious symptoms; bone setters who engage with orthodox services reduce complications [sources: https://univ.edu.vu/attachments/article/410/Art.Vulgarisation_Final_EN.pdf; https://journals.sagepub.com/doi/ref/10.1177/00494755221122543]; the distinction between healers who do and do not name culprits is [synthesis].

What to do instead and how to help those harmed

Instead of culprit-finding: keep the focus on the sick person's care, explain disease processes to families (as PNG's plan asks health workers to do) [source: https://stopsorceryviolence.org/the-sorcery-national-action-plan], and use the explanatory-model, family and safety work in 4.4, 4.11-4.13.
For accused people and survivors: immediate safety and medical care; trauma-informed assessment (PTSD, depression, suicidality, grief); practical help with relocation, housing and income; restoring social ties through trusted church or community members; careful records of injuries; links to legal help and protection [synthesis]. PNG's 2014 plan lists care and counselling, repatriation support and service-provider training among its five areas [source: https://stopsorceryviolence.org/the-sorcery-national-action-plan]; a National Action Plan to Combat Sorcery Accusation-Related Violence, a report titled "No One Should Stand Alone" and a community curriculum were launched in PNG in September 2025 [source: https://www.theinternationalnetwork.org/latest/pngactionplan]. Forsyth notes human rights defenders support survivors largely without structured support [source: https://cdn.auckland.ac.nz/assets/humanrights/Research/HRWorkingpaper20-Forsyth.pdf].
For accusers and families: grief work after the triggering death; addressing the underlying dispute; clear statements that violence is a crime [synthesis].
Legal context: in PNG, the Sorcery Act 1971 was repealed in 2013 and a Criminal Code provision (s299A) makes intentionally killing a person over a sorcery accusation wilful murder [source: https://stopsorceryviolence.org/the-sorcery-national-action-plan]. Forsyth (2013) reports that harsher sentences had not reduced violence [source: https://cdn.auckland.ac.nz/assets/humanrights/Research/HRWorkingpaper20-Forsyth.pdf]. In Vanuatu, killings over sorcery are prosecuted as intentional homicide [source: https://www.stopsorceryviolence.org/justice-in-vanuatu-for-sorcery-violence-crime/]; the Penal Code [Cap 135] s 151 (Witchcraft) makes it an offence to practise witchcraft or sorcery with intent to cause harm to another person, with up to 2 years' imprisonment [source: https://vanuatucustoms.gov.vu/images/legislations/Customs/Others/Penal_CodeCAP_135.pdf].
Clinician duties: follow your jurisdiction's emergency, police, child-protection and adult-safeguarding procedures; report imminent risk; keep confidentiality except where the law or risk requires disclosure; never give identifying information about a suspected person to anyone who might act on it [synthesis].

Sources

← Previous · XI-125Tongan and Fijian healers Next · XI-127 →Cunning folk