The role of culture on the phenomenology of hallucinations and delusions, explanatory models, and help-seeking attitudes: A narrative review

Aim: Culture has been posited to be involved in the formation and maintenance of delusions and hallucinations. The extent of these differences and how they affect explanatory models of psychosis and help-seeking attitudes remains to be understood. This review aims to present a cultural formulation to account for psychosis onset, symptom maintenance, and help-seeking attitudes. Methods: A narrative review was conducted to summarize the existing evidence base regarding cross-cultural differences in hallucinatory and delusional prevalence, explanatory models, and help-seeking attitudes in First Episode Psychosis (FEP) and Non-FEP Schizophrenia samples. Results: Sixteen studies were eligible for inclusion. In terms of positive symptom specificity, cross-cultural differences were found. Specifically, auditory and visual hallucinations occurred most frequently in African patients, persecutory and grandiose delusions occurred at higher rates in African, Pakistani, and Latino patients, while delusions of reference were most prevalent in White-British groups. Three explanatory models were identified. Westerners tended to endorse a bio-psychosocial explanation, which was associated with increased help-seeking, engagement, and positive medication attitudes. Asian, Latino, Polish, and M (cid:1) aori patients endorsed religious-spiritual explanatory models, while African patients opted for a bewitchment model. The religious-spiritual and bewitchment models were associated with a longer duration of untreated psychosis (DUP) and poorer engagement with mental health services. Conclusions: These findings highlight the important influence of culture in the formation and maintenance of positive symptoms of psychosis, engagement, and help-seeking attitudes across different ethnic groups. The incorporation of cultural beliefs in formulation development could facilitate enriched CBTp practices and improved engagement amongst different cultural groups with Early Intervention Services.


| INTRODUCTION
Culture is a dynamic set of shared attitudes and customary practices that shape our perceptions of the world and exerts an influence on the form and content of psychosis symptoms (Jones et al., 2021;Luhrmann et al., 2015a;McLean et al., 2014). For example, there is substantial variability in the appraisal of symptoms of psychosis across cultural groups ranging from beliefs concerning God, Jinns, spirits, possession, black magic, and witchcraft (Bhikha et al., 2015;Burns et al., 2011;Carter et al., 2017;Jacob, 2014). These experiences may be appraised as distressing or as a divine spiritual gift depending on which cultural group one belongs to (Heffernan et al., 2016). Indeed, 15%-40% of the variability in psychosis symptoms has been accounted by cross-cultural differences (Stompe et al., 2006). This accounts for the growing interest towards understanding how cultural factors contribute to the formation and maintenance of psychosis, as well as engagement with treatment and services. Prevalence rates are 1.5-3-times elevated amongst Asian and African cultures compared to their European-American counterparts (Jongsma et al., 2019;Morgan et al., 2019). Despite these increased prevalence rates, only 30% of Asian and African clients complete treatments compared to 50% of White-British groups (Moore, 2018). Globalization has led to accelerating migration and increased cultural diversity and as such, to varying conceptualisations of mental illness and help-seeking behaviours. However, to date there appears to be an under-emphasis on the utilization of culture in the treatment of psychosis and a lack of understanding regarding how it may impact help-seeking and engagement with Early Intervention Services (EIS). Given the importance of reducing the Duration of Untreated Psychosis (DUP), this highlights the value of furthering our knowledge of cultural mediators of psychosis (Kane et al., 2016). Improved understanding of cross-cultural differences would be invaluable in facilitating culturally sensitive, personalized, and idiosyncratic symptom-specific formulations within Cognitive Behavioural Therapy for Psychosis (CBTp). It would also foster improved understanding of some of the barriers to engagement and concordance with interventions.
Models of Understanding (MOU) refers to culturally determined processes impacting mental health perceptions, causal psychosis attributions, and treatment expectations (Carter et al., 2017). Ethnocultural and idiosyncratic beliefs shape MOU for clients and social cultural communities. Mental health conceptualisations can influence help-seeking, treatment compliance, and meanings attributed to psychosis characteristics. Hence, an awareness of the disparities between western mental health concepts focusing on bio-psychosocial conceptualisations of mental illness and alternative explanatory frameworks such as spirituality, which is more prominently held in minority cultural groups could facilitate improved engagement in EIS (Larøi et al., 2014). This is the first review to synthesize the existing evidence-base regarding cross-cultural differences in psychosis symptom prevalence, explanatory models, and help-seeking behaviours with the view to proposing a culturally-specific formulation model outlining the influence of culture on hallucinatory and delusional onset, maintenance, and content.

| METHODS
This review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Page et al., 2021). Methods and inclusion criteria were specified in advance and documented in a protocol registered with the PROS-PERO International Prospective Register of Systematic Reviews (registration number: CRD42022340067).

| Study selection
Trials of any design, length of follow-up, and any date of publication were considered. Eligible studies examined the influence of culture on any symptoms of psychosis and/or engagement with mental health services. Studies were also included if they reported prevalence rates for psychosis symptoms within specific cultural groups. Studies were eligible for inclusion if they were published in peer-reviewed journals, written in English, and investigated any psychosis sample (First Episode Psychosis [FEP] and non-FEP schizophrenia populations). To be included, participants had to be diagnosed with psychosis using a valid and reliable psychometric tool (e.g., DSM-IV; American Psychiatric Association, 1994). Non-clinical populations such as university student populations were included if their attitudes to psychosis and help-seeking were investigated and their cultural background was clearly documented. The study selection process is summarized in the PRISMA flow diagram (see Figure 1).

| Exclusion criteria
Studies written in languages other than English, unpublished/grey literature, conference abstracts, book chapters, and case studies were excluded from the review. Studies were excluded if patients had druginduced psychosis or psychosis due to organic causes.

Relevant literature was de-duplicated and imported into Microsoft
Excel. An Excel spreadsheet was created to record the characteristics of the included studies. The extracted data included different variables: authors name, study location, sample, diagnosis, study design, clinical setting, main findings, and clinical implications. Study screening was performed independently by one reviewer (MG) and was subsequently crosschecked by a second reviewer (CE-D). Any disagreements were resolved by a third reviewer (AG).

| Quality appraisal
The methodological quality of the included studies was assessed using the Joanna Briggs Institute Critical Appraisal Checklist for Qualitative Research (Joanna Briggs Institute;Lockwood et al., 2015). Studies were included if they met the quality points considered essential by the review team, namely a 'yes' response for questions 2-5 and 8-10, and consensus was agreed by the whole research team. All papers provided an explanation of their aims, justified their use of qualitative methods, and explicitly reported their findings. All 16 studies met the quality JBI assessment criteria and were therefore included in the present narrative synthesis (see Table 1). The JBI checklist for studies reporting prevalence data was also employed and studies were included if they achieved a 'yes' response for questions 1-9. All seven studies examining prevalence met the JBI assessment prevalence criteria and were thus included in the present review (see Table 2).

| Narrative synthesis
A structured narrative synthesis was conducted in accordance with the ENTREQ statement (Enhancing transparency in reporting the synthesis of qualitative research) (Tong et al., 2012). Data was tabulated, textual descriptions were summarized, and these findings were then grouped into clusters with shared themes, allowing for the identification and comparison of commonalities between studies and cultural groups (Arai et al., 2007;Barnett-Page & Thomas, 2009;France et al., 2019;Lachal et al., 2017). The results section has been structured based on the thematic main findings and an inductive approach was adopted for theory generation regarding the possible influence of culture on psychosis formation, symptom persistence, and engagement with services.

| RESULTS
Of the 66 papers initially identified through database screening, a total of 16 studies were included in the present review. Table 3 summarizes the characteristics and main findings of the included studies.

| Positive-symptom prevalence
Seven studies explored cross-cultural differences in the prevalence of positive symptoms of psychosis. Velthorst et al. (2011) examined atrisk mental-state (ARMS) patients and found a higher prevalence of negative symptoms in three ethnic minority groups; Moroccan-Dutch, Turkish-Dutch, and Surinamese compared to the majority Native-Dutch group. However, no statistically significant differences were found between these groups in terms of the prevalence of positive symptoms. This may be due to inclusion of an ARMS only sample, which is characterized by sub-threshold psychosis symptoms, thereby reducing the likelihood of identifying statistically significant crosscultural differences in psychosis symptom prevalence.     in the PP (42%) than the British-Pakistani (23%) or White-British (14%) groups.

|
Similarly, Campbell et al. (2017) examined the relationship between culture and delusional subtypes in the South-African Xhosa people and found that the most prevalent delusion was persecutory delusions. The majority (86%) believed that family or friends were planning to harm them due to jealousy via black magic or bewitchment. Two thirds of the sample reported grandiose delusions (e.g., having special powers, being the Son of God, Prophets, or healers), thought broadcasting, and thought control by evil spirits (all 59%). This was followed by somatic delusions (37%) with the head, skull, or brain being the focus of concern, and the least common delusional subtype pertained to religious delusions (26%) characterised by the belief that one has a special relationship with God.

| Explanatory models and help-seeking behaviours
Three cross-cultural explanatory models were identified in the present review, namely (a) bewitchment, (b) religious-spirituality, and (c) the bio-psychosocial model.

| Bewitchment
Two studies reported that African patients tended to choose a bewitchment explanation to account for their psychosis symptoms (Campbell et al., 2017;Napo et al., 2012). Of their South-African patient sample, 72% believed that their schizophrenia was caused by bewitchment due to jealousy from others (Campbell et al., 2017). Similarly, Mali, West-African patients attributed schizophrenia primarily to external causes including magical inferences, evil eye, and witches (Napo et al., 2012). This aligns with a jealousy-induced witchcraft framework common in African communities, prompting Africans to seek traditional faith healers (TFH) in order to cleanse themselves of evil spirits and to repel perceived bewitchment practices.

| Religious-spirituality
Polish, M aori, Latin, and Asian cultural groups attributed psychosis to religious-spiritual explanations characterized by beliefs in Jinns and spiritual possession, increasing the likelihood of help-seeking initially through TFH (Pietkiewicz et al., 2021). Roman-Catholic Polish patients were found to adopt spiritual-possession explanations for psychosis symptoms and sought spiritual help-seeking through exorcism first (Pietkiewicz et al., 2021). It was similarly found that the M aori, New-Zealanders, attributed psychosis to spiritual explanations. This resulted in delayed-help seeking, as clients were apprehensive about disclosing their explanatory models due to concerns about services being dismissive (Taitimu et al., 2018).  (Bhugra et al., 1999). These authors posited that Asian patients who primarily endorse religious-spiritual explanations seek TFH initially and therefore delay help-seeking from services, subsequently increasing DUP.

| Bio-psychosocial
Western mental-health services tend to conceptualize mental illness using the bio-psychosocial model, which advocates that stress interacts with one's overall level of bio-psychosocial vulnerability to give rise to mental health difficulties. This explanation was indeed held by Western patients from Britain and America (Bhugra et al., 1999;Luhrmann et al., 2015a).
Two studies demonstrated that bio-psychosocial explanations tended to result in improved help-seeking (Bhugra et al., 1999;Caqueo-Urízar et al., 2015). Specifically, White-British groups (59%) preferred the bio-psychosocial model, emphasizing the role of genetics and one's social environment in contributing to symptom development (Bhugra et al., 1999). Additionally, Latin-American patients who endorsed bio-psychosocial explanations demonstrated improved helpseeking due to their positive medication attitudes and lower symptom severity in comparison to those adopting religious-spiritual explanations. These findings elucidate the importance of exploring clients' explanatory models, as they may serve as predictors of symptom severity, medication attitudes, and help-seeking (Caqueo-Urízar et al., 2015). It also highlights the value of sharing the biopsychosocial model conceptualisation of psychosis through psychoeducation in order to improve engagement with services and to reduce distress associated with spiritual-religious and bewitchment frameworks. However, the timing of which would need to occur after a comprehensive assessment of the client's cultural/religious beliefs and explanatory models, meaning and content of psychosis symptoms, and validation of the client's distress. This will enhance client engagement owing to their experiences not being dismissed or invalidated.
The bio-psychosocial model can then be proposed as an alternative explanatory hypothesis for their distressing experiences and through socratic questioning the clinician can explore collaboratively with the client whether this alternative bio-psychosocial conceptualisation is more helpful and less distressing that their pre-existing cultural explanatory famework (e.g. possession). If the client deems psychosis to be more distressing than their cultural/religious attribution, then an exploration of what psychosis means for them would be important.
For example, their primary appraisals may concern negative other-self evaluations such as a fear of rejection and exclusion due to mental illness (e.g. threat to marriage prospects), while their secondary appraisals may pertain to core beliefs about the self as a failure, weak/vulnerable, defective, inferior, bad, worthless, or unlovable.
Therefore, working with the appraisals of psychosis and beliefs about the self regarding having developed psychosis, whilst also building more adaptive beliefs about recovery and treatment would be clinically meaningful intervention targets. Only 12% correctly identified and labelled schizophrenia from the vignettes, with more than 10% using stigmatizing labels such as 'crazy' and 'mad'. Despite poor mental-health literacy, most of these students suggested appropriate help-seeking from psychiatrists (36%), psychologists (20%), physicians (14%), and counsellors (10%). Less than 10% suggested religious help-seeking or TFH, which contradicts the findings of two previous studies of African patients (Campbell et al., 2017;Napo et al., 2012). Cadge et al. (2019) found that White-British, Pakistani, Indian,

| Non-clinical samples
and African-Caribbean students demonstrated poor knowledge of schizophrenia, with the Indian and Pakistani groups regarding schizophrenia as a 'made-up condition' therefore delaying help-seeking.
Cross-cultural differences were evident in the explanatory models employed. Pakistani students tended to endorse a religious explanatory model, attributing the pathogenesis of psychosis to Jinn possession, in line with Asian clinical populations (Bhugra et al., 1999).
In contrast, Indian students emphasized upbringing as a causal factor for psychosis, which is incongruent with the spiritual explanatory models advocated by Indian patients who deemed voices as kin offering guidance (Luhrmann et al., 2015a). Societal stigma was a barrier to help-seeking in both Asian and Pakistani cultures due to the perceived negative impact of psychosis on marriage prospects, leading to a higher DUP.
In a vignette-based study, Phalen et al. (2019) reported that the strength of religiosity impacted on voice-hearing experiences in American students. Highly religious Americans were less likely to attribute psychosis to positive voice-hearing experiences. Conversely, the same students were confident ascribing psychosis labels to negative voicehearing experiences, resulting in greater stigmatization. These authors suggested that perceptions of voice hearing are impacted by culturally relevant religious etiological explanations. Similarly, American patients (60%) reported more frequent violent voice commands compared to Indian patients (10%) (Lebovitz et al., 2021). However, twice as many Indian patients (N = 14) to American patients (N = 7) reported hearing voices of a sexual theme, which generated feelings of shame in comparison to the American patients who considered these voices pleasurable. These authors suggested that differing cultural explanations influenced voice content and ensuing client distress.

| Cultural formulation for delusions and hallucinations
Based on the findings of the present review we propose the following 4 | DISCUSSION

| Prevalence
There was evidence for cross-cultural differences in the prevalence and content of hallucination subtypes, with AH being most frequently reported across cultures. Latinos frequently reported hearing advisory-commanding voices of religious content compared to Americans (Yamada et al., 2006). Religiously-themed content may be culturally syntonic with Latino culture and advisory-type content may be reflective of Latino's family interdependence (Larøi et al., 2014). Family in Latino culture is a protective factor against negative AH appraisal, reducing stress sensitivity, and normalizing AH as guidance from external agents such as God or family (Gonyea et al., 2016).
Thus, prompting greater AH reporting and less distressing voice hearing experiences.
VH prevalence rates between European-American and Asian cultures were conflicting. VH were more prevalent in non-Western cultures in Ghana, Nigeria, and Latinos (Bauer et al., 2011;Yamada et al., 2006). VH of Jinns were most prevalent in PP as opposed to British-Pakistani or White-British groups (Khan & Sanober, 2016;Suhail & Cochrane, 2002). This can be explained by the social rekindling theory, suggesting that cultural expectations impact VH content and subsequent appraisal (Luhrmann et al., 2015b). Pakistani culture is influenced by Islam, whereby Jinn are conceptualized as invisible beings, appearing in various forms that are capable of possession. The dominance of religious-spiritual and bewitchment models in non-Western cultures therefore fosters religious interpretations. Endorsement of religious-spiritual explanations in Pakistan results in greater acceptance of seeing spirits, leading to an increased willingness to report them, and a lower threshold for experiencing religious VH.
Cross-cultural differences were found in the prevalence of delusional subtypes. Grandiose delusions were most prevalent in PP and African cultures comprising of beliefs of being God, a Prophet, or healers. Religious delusions are of clinical significance as they are associated with increased symptom-severity, higher rates of self-harm, and are held with stronger conviction than non-religious delusions (Bhavsar & Bhugra, 2008;Lilford et al., 2020). These culturally-specific beliefs may be attributed to prominent economic and social class disparities observed in Pakistani and African societies, resulting in difficulty achieving upward social mobility. Therefore, grandiose delusions may serve to bolster self-esteem and aspirations of success. Indeed, associations have been found between lower social class and grandiose delusions, highlighting the impact of sociocultural background on influencing delusion phenomenology (Sajid et al., 2011).
Alternatively, one study found that European-Americans were twice as likely to exhibit grandiose delusions than Latinos and African-Americans (Yamada et al., 2006). Enhanced prevalence rates may reflect an emphasis on individualism in European-American cultures, which is positively associated with an enhanced opinion of oneself (Burton et al., 2021). This is culturally dystonic with Latino and African collectivist cultures. This is exemplified by the finding that U.S. born F I G U R E 2 Cultural formulation for delusions and hallucinations.
Delusions of reference were most common amongst White-British as opposed to Pakistani or African patients (Suhail & Cochrane, 2002

| Explanatory models and help-seeking behaviour
The bio-psychosocial model was the predominant explanatory model for psychosis in European-American cultures. This model was associated with improved help-seeking, positive medication attitudes, and lower symptom severity (Caqueo-Urízar et al., 2015). Indeed, the endorsement of bio-psychosocial explanations of psychosis resulted in greater CBTp compliance and reduced relapse risk (Carter et al., 2017).
Indeed, south asian students tended to attribute psychosis to evil spritis, punishment from God, or to previous misdeeds, rather than to a mental-health difficulty (Mirza et al., 2019). In highly religious Western cultures such as Poland, seeking religious-interventions is widespread, as being possessed is less stigmatizing than having psy- Religious-spiritual explanatory models were associated with postponed help-seeking owing to initial TFH contact and lower service-engagement (Caqueo-Urízar et al., 2015). In Asian cultures, religious-spiritual beliefs relieves clients of potential shameful community stigma associated with psychosis. As being possessed is more socially-acceptable than having psychosis, this leads to pursuing faithhealers or exorcists (Kopeyko et al., 2018). Individuals may thus adopt religious-spiritual models to avoid stigma associated with traditional bio-psychosocial explanations. If unaddressed, stigma postpones helpseeking and worsens prognosis, highlighting the necessity for stigmareduction interventions.
An explanation for delayed help-seeking is that clients are concerned that their religious-spiritual explanations will be dismissed, prompting them to seek help from TFH; further increasing DUP and symptom severity (Taitimu et al., 2018). As mental-health services are underpinned by a bio-psychosocial model, there is disparity between western mental-health conceptualisations and religious-spirituality frameworks, which needs to be addressed (Mohr et al., 2012). Indeed, 70% of clinicians have reported poor knowledge of religion, with 37% found to be uncomfortable discussing religion within their psychological formulations (Rosmarin et al., 2013). This further highlights the importance of occupying a position of curiosity and integrating spirituality within CBTp practice in order to increase engagement in first episode psychosis.

| Non-clinical samples
Conflicting findings were found comparing explanatory models in non-clinical university students with clinical samples. Pakistani students tended to endorse a religious-explanatory model with beliefs in Jinn possession, similar to clinical samples (Bhugra et al., 1999).
Highly-religious Americans withheld stigmatizing labels to positive

| Voice-hearing experiences
Cultural variations in voice-hearing experiences and content were also found. Americans were more likely to hear violent command hallucinations instructing them to harm themselves or others (Lebovitz et al., 2021). Voice content may reflect cultural norms, as violence is more prominent in America with greater exposure to gun-related violence in the media (Shulman et al., 2021). This might explain AH content characterized by violence, resulting in distressing voice-hearing experiences. Hence, negative voice-content may reflect wider socialcultural contexts.
Conversely, Indians heard more sexual-related voices compared to Americans. Marriage is a significant source of identity within Indian communities with social restrictions placed on sexual relationships outside of marriage. Thus, sexual voices may be viewed as a form of punishment and are consequently negatively appraised, resulting in increased shame, and heightened familial and community stigma.
Whereas, in America, sex is less taboo and more openly discussed (Pinto, 2014). Thus cultural context appears to contribute to the form and content of voice-hearing experiences across cultures. Thus, clinicians would benefit from exploring the cultural-meanings ascribed to voices in order to culturally-adapt CBTp practices.

| Study strengths
As the prevalence data was quantitative, and the relationships between culture and psychosis symptomatology and engagement with mental health services were qualitative, a mixed research synthesis was employed (Sandelowski et al., 2006). This has been deemed advantageous for narrative reviews, as the combined data can provide a more detailed synthesis of the data available regarding a given topic and therefore can be deemed a strength of the current review. This narrative synthesis is unique because no previous reviews have investigated such an extensive range of cross-cultural differences in positive symptom prevalence, explanatory models, and Included studies were heterogeneous in terms of outcome measures, prevalence assessments, explanatory models, and help-seeking.
This therefore makes comparisons of cross-cultural findings difficult.
Additionally, as all the studies examining the impact of culture on psychosis symptomatology were qualitative, statistical analyses were not possible.
As narrative reviews rely on narrative rather than statistical synthesis, it has been deemed a more interpretative and subjective approach than systematic reviews. However, it has been argued that these methodological approaches serve different purposes. Systematic reviews aim to summarize quantitative evidence to address a narrowly focused research question, while narrative reviews synthesize qualitative data from heterogeneous sources to derive meaning and theory generation. It has thus been argued that these methodologies serve differing yet complementary approaches to data synthesis (Greenhalgh et al., 2018).

| Clinical implications
This review highlights that stigma within cultural communities can prevent service-engagement, increasing DUP. If unaddressed, crosscultural differences in stigma may impede timely diagnosis and treatment, reducing the likelihood of recovery and remission.
To reduce stigma, community-based strategies such as clinicians working collaboratively alongside religious-community leaders to improve psychosis understanding, and to promote service engagement may help with the normalization and de-stigmatization of psychosis within such communities. In EIS settings, psychoeducation addressing the specific cultural impact of internalized stigma can help families to cope with stigma related to locally held beliefs and decrease expressed emotions (Woods et al., 2015). Thus, discussions regarding culture, spirituality, and religion, as well as perceived stigma, could be helpfully incorporated within CBTp assessments to aid in formulation development. It may also enhance self-disclosure and engagement.
Current findings emphasize the importance of incorporating religious-spiritual explanatory models, prominent in Asian and African cultures, into interventions to improve help-seeking. As only 47% of clients report spirituality-related discussion in CBTp, there remains uneasiness expressing spirituality within conventional-mental-health services (Yamada et al., 2006). Consequently, clients seek help from TFH as clinicians often have limited knowledge regarding incorporating spirituality-religion into formulations (Milner et al., 2020). Culturalcompetence training informed by faith-healers is thus necessary as religious attributions of psychosis are present in certain cultures.
Cultural-religious terms could also be utilized reflecting culturallynormative expressions of distress in CBTp (Rathod et al., 2019). This would help to clarify if one's perceptions are true hallucinations or prevalent in religion, thereby reducing misdiagnosis risk (Adeponle et al., 2012). In a recent systematic-review, integrating religionspirituality into CBTp was more effective than CBTp alone for it validated clients' religious-spiritual viewpoints rather than dismissing them as maladaptive (Turner & Hodge, 2020

| Future research
As most of the available evidence was conducted in the UK and USA, future research should explore cross-cultural differences across wider European demographics. This can determine the generalizability of findings in other contexts including low-medium-income countries.
Additional research is necessary to further refine tentative symptomspecific formulation models incorporating the impact of culture, religion, spirituality, and stigma on psychosis formation and symptom persistence. This would accelerate cultural adaptations by predicting cultures help-seeking methods based on their chosen explanatory model, hallucinatory, and delusional subtype.

| CONCLUSION
This is the first narrative review to synthesize the existing evidencebase surrounding cross-cultural differences in positive psychosis symptom prevalence, explanatory models, and help-seeking attitudes.
The contribution of cultural factors to the phenomenology of hallucinations and delusions and the identification of explanatory models across cultures has enabled the development of a culturally informed formulation model for psychosis.
Religious-spiritual and bewitchment explanatory models were primarily endorsed by Asian and African cultures, respectively while European-Americans opted for a bio-psychosocial framework. The religious-spiritual and bewitchment causal explanations were found to contribute to delayed help-seeking with EIS services, as TFH were sought instead.
Given the cultural variations found in this review, clinicallymeaningful modifications can be facilitated in assessments and formulations to address the incongruity between non-Western clients' explanatory model of psychosis and the bio-psychosocial model underpinning CBTp. Stigma-reduction interventions could further address the culturally-specific impact of stigma in order to validate marginalized ethnicities' understanding of psychosis. As the findings demonstrate the prominence of religious-spiritual explanations in Asian and African cultures, services would benefit from integrating religion-spirituality within CBTp formulations. Through these servicelevel and therapeutic adaptations, engagement and acceptability could be enhanced in cultural-minority groups.