The current status of culturally adapted mental health interventions: a practice-focused review of meta-analyses

Figures & data

Figure 1 PRISMA flow chart.

Notes: Moher D, Liberati A, Tetzlaff J, Altman DG; The PRISMA Group (2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA Statement. PLoS Med. 6(6):e1000097.³²

Table 1 Findings from review of meta-analysis of culturally adapted interventions

Author	Study characteristics	Participant characteristics	Results and methodological rigor	Cultural adaptation, moderators, comments
Griner and Smith^Citation14	N=76 studies Focus = Broad Type of review = Effect of interventions Design of studies = Experimental, quasi-experimental, no comparison group Intervention types = Not reported Settings = Not reported Outcomes = Ranging from retention to clinical outcome Economic evaluation = Not reported	N=25,225 Age = All ages Gender =55% female Background =31% African American, 31% Hispanic/Latino, 19% Asian American, 11% Native American, 5% European American, 3% not specified Target problem = Wide range Acculturation level = Reported Region (western, non-western) = Not reported	Effect size = d=0.45 (SE =0.04, p<0.0001) with a 95% CI of d=0.36–0.53 Heterogeneity analysis (Q(75) =459, p<0.0001) Risk of bias analysis = U Search strategy = Adequately described Inclusion/exclusion criteria = Adequately described (only English literature studies) Heterogeneity analysis Risk of bias analysis = Yes Search strategy = Adequately described Inclusion/exclusion criteria =	Cultural adaptation = Attempted but inadequately reported (not consistently conducted) Age (older age), ethnicity (interventions with groups that consisted of participants who were same race yielded effect sizes, which were four times more efficacious than did studies with groups of mixed racial composition), and language matching (studies in which therapists spoke the same [non-English] language as clients had effect sizes that were twice as high [d=0.49] than studies in which no descriptions regarding language matching were provided [d=0.21])
Huey and Polo^Citation17	Number =25 studies Focus = Minority youth with anxiety and other clinical problems Type = Effect Design of studies = Experimental Intervention types = Not reported Settings = Not reported Outcomes = Clinical outcomes Economic evaluation = Not reported	N=1,056 Age ≤18 years Gender = N/A Background = African American, Hispanic, Multiracial Hawaiian Target problems = Anxiety, attention-deficit/hyperactivity disorder, depression, conduct, substance use disorder, trauma-related syndromes, other clinical problems (no formal psychiatric diagnosis) Acculturation level = Not reported Region = N/A (Studies were included if at least 70% participants were from ethnic minority background)	Heterogeneity analysis = Effect size at 4–6 months (d=0.36), 1–1.17 years (d=0.28), 4 years (d=0.68), and 13.7 years (d=0.37). Effect sizes higher with no control conditions (d=0.58); cf placebo (d=0.51) and TAU (d=0.22) (13 probably efficacious, 17 possibly efficacious) Risk of bias analysis = Yes Search strategy = Described Inclusion/exclusion criteria = Adequately described	No moderating effect for ethnicity, problem type, clinical severity, diagnostic status, and culture-responsive treatment
Hodge et al^Citation16	N=21 studies Focus = Youth Type = Effectiveness Design of studies = Experimental, quasi-experimental, no comparison group Intervention types = Not reported Settings = Not reported Outcomes = Behavioral outcomes Economic evaluation = Not reported	Age ≤18 years Gender = Both Background = Latino, African American, or Native American Target problems = Behavioral problems Acculturation level = Not reported Region = Only USA	Effect size = Adapted interventions an aggregate standardized mean difference of 0.240 compared to those not participating in CSIs (Hedge’s g=0.239, 95% CI =0.139–0.339, p<0.001), kappa (0.84) Heterogeneity analysis = Yes Risk of bias analysis = + Some degree of publication bias found Search strategy = Adequately described Inclusion/exclusion criteria = Adequately described	No moderating effect of race or ethnicity Overall results suggest that CSIs are minimally effective with minority youths
Smith et al^Citation20	N=65 Focus = Broad Type = Effect Design of studies = Experimental and quasi-experimental Intervention types = Not reported Settings = Not reported Outcomes = Clinical Economic evaluation = Not reported	N=8,620 Age =24.4 years (55–73) Gender =55% female Background =39% Asian American, 32% Hispanic/Latino(a), 20% African American, 4% Native American, 1% White/European American, and 4% other Target problems = Various Acculturation level = Experimental, quasi-experimental, no comparison group Region = Experimental, quasi-experimental, no comparison group	Effect size = Weighted average effect size was d=0.46 (95% CI =0.36–0.56) (effect size using “trim and fill” methodology;^Citation33 the recalculated value was d=0.27 [95% CI =0.16–0.38] [effect sizes ranged from −0.97 to 2.80]) Heterogeneity analysis substantial heterogeneity across studies, I^2=74%; Q(64)=247, p<0.001 Risk of bias analysis = Outcome evaluations by therapists effect sizes (d=0.09), cf by the clients (d=0.45) or external observers (d=0.45) Search strategy = Adequately described Inclusion/exclusion criteria = Adequately described	Age (average age) was strongly associated with the magnitude of effect sizes within studies, and ethnicity (studies with Asian American had an average effect size of more than twice that of 14 studies of African American participants [d=0.47], Hispanic/Latino(a) [d=0.47], and Native American participants [d=0.22]) Specific cultural group versus mixed ethnic group (interventions delivered to a specific cultural group were much more effective [d=0.51, 95% CI =0.40–0.63] than interventions delivered to mixed groups [d=0.18, 95% CI =−0.08 to 0.44])
Benish et al^Citation12	N=59 (21 studies that compared adapted versus unadapted therapy) Focus = Adapted versus unadapted therapy Type = Effectiveness Design of studies = Experimental, quasi-experimental, no comparison group Intervention types = Not reported Settings = Not reported Outcomes = Various Economic evaluation = Not reported	N=1,242 Age = Mean age 26 years Gender =46% female Background = African American (21.6%), Asian American (14.3%), non-Puerto Rican Latino/Hispanic (33%), and Puerto Rican Latino/Hispanic (26.7%) Target problems = Various Acculturation level = Experimental, quasi-experimental, no comparison group Region = USA	Effect size = For analysis 1 aggregated effect size d=0.41 for primary measures and 0.33 for all measures. For analysis 2 aggregated effect size d=0.32 for primary measures Heterogeneity analysis Risk of bias analysis = Yes Search strategy = Adequately described Inclusion/exclusion criteria = Adequately described	The conditional model with myth adaptation as a Level 2 predictor revealed that adaptation of the myth was significant (p=0.04), with a coefficient of 0.21 relative to culturally adapted treatments without illness myth adaptation
Chowdhary et al^Citation13	N=16 studies Focus = Depressive symptoms in adult population Type = Effect and process of adaptation Design of studies = RCTs (randomized controlled trials) Intervention types = CBT, IPT, PS, PST, DOT, PE Settings = Various (group, classes, individual, community, clinic, etc.) Outcomes = Ranging from retention to clinical outcome Economic evaluation = Not reported	N=4,162 Age = Studies conducted in adults (aged > or equal to 19 years) Gender = N/A Background = N/A Target problems = Depressive disorder Target diagnoses/symptoms = Depressive disorders Acculturation level = Not reported Region = Not reported	Effect size = SMD =−0.72, 95% CI =−0.94 to −0.49 Heterogeneity analysis significant heterogeneity (χ^2=146, df=15, p<0.001, I^2=90%) Risk of bias analysis = Funnel plot was asymmetrical suggesting publication bias. Studies with no active treatments had higher effect than those comparing an active group Search strategy = Adequately described Inclusion/exclusion criteria = Adequately described (excluded studies that used a new PT that was developed specifically for the ethnic group and was not an adaptation of an existing PT)	Cultural adaptation – MRC Guidelines* and Bernal & Saez-Sanriago’s (2006) model to describe nature of adaptation (majority of adaptations were made in the dimensions of language, context, and therapist delivering the treatment)
Rojas-García et al^Citation19	N=11 studies (16 used in the review) Focus = Perinatal depression in socially disadvantaged population Type = Effect Design of studies = RCTs, quasi-experimental studies with a comparison group Intervention types: CBT, interpersonal therapies, modified CBT Settings = Not reported Outcomes = Various Economic evaluation = Not reported	N=864 Mean age = 18 years Gender = Male and female Background = Various Target problems = Perinatal depression (limited to OECD [Organization for Economic Co-operation and Development] countries) Acculturation level = Experimental, quasi-experimental, no comparison group Region = OECD countries	Effect size = 0.44, 95% CI =−0.67 to −0.22 Heterogeneity analysis Risk of bias analysis = Yes Search strategy = Adequately described Inclusion/exclusion criteria = Described	Individually administered interventions (higher effectiveness [−0.79, 95% CI =1.16 to −0.43]), cf group-administered interventions (−0.30, 95% CI =−0.53 to 0.07), interventions conducted in hospitals/clinics (0.61, 95% CI =−0.90 to −0.32) and at home (−0.68, 95% CI =−1.20 to −0.17); those conducted after birth (−0.82, 95% CI =−1.26 to −0.38); those based on psychoeducation (−0.81, 95% CI =−1.23 to −0.39) and IPT (−0.51, 95% CI =−0.92 to −0.10), and those that were culturally adapted (−0.54, 95% CI =−0.78 to −0.29) had more effect
Rojas-García et al^Citation18	N=15 studies (13 interventions) Focus = Depression, low socioeconomic status (SES) Type = Effect Design of studies = RCTs and quasi-experimental studies with nonrandomized comparison groups (control groups had to receive usual care or enhanced usual care) Intervention types = Psychotherapy, tele-psychiatry (counseling), cognitive behaviour therapy Settings = Primary, secondary, or tertiary care Outcomes = At least one measure of depressive symptoms Economic evaluation = Not reported	N=2,261 Age = 13–70 Gender = Male and female Background = Various Target problems = Limited to OECD countries Acculturation level = Experimental, quasi-experimental, no comparison group Region = OECD countries	Effect size = For short term (upto 3 months) −0.58, 95% CI =−0.74 to −0.41, and for long term −0.42, 95% CI =−0.63 to −0.21 Heterogeneity analysis Risk of bias analysis = + Publication bias Search strategy = Adequately described Inclusion/exclusion criteria = Described	Culturally specific training for providers/therapists and booster sessions improved the effectiveness of the interventions. Use of peer providers and the provision of extra services helped to facilitate adherence to the intervention Interventions that targeted low-SES ethnic minorities were less effective than those addressed to low-SES populations who were not from ethnic minorities Interventions adapted for participants from the same race were 4 times more effective than those interventions provided to those of mixed races
Sutton^Citation22 (dissertation)	N=49 studies Focus = Latino participants Study type = Effect Design of studies = Experimental, quasi-experimental, no comparison group Intervention types = Not reported Settings = Experimental, quasi-experimental, no comparison group Outcomes = Various Economic evaluation = Experimental, quasi-experimental, no comparison group	N = Unknown Age = U Gender = U Background = Latinos Target problems = Various Acculturation level = Experimental, quasi-experimental, no comparison group Region = USA	Effect size = For mental health outcomes d=0.58 (CI =0.33–0.83) using the random-effects model and d=0.37 (CI =0.31–0.43) using the fixed-effects model Heterogeneity analysis Q(48)=697 was statistically significant (p<0.001) Risk of bias analysis = + Full details of risk bias analysis process not provided Search strategy = Adequately described Inclusion/exclusion criteria = Described	The type of cultural adaptation (ie, surface, deep, or combined structure) does not appear to moderate outcomes. Acculturation, the immigration status of the study participants, the primary language of intervention, treatment style and format, and construct categories of mental health did not influence effectiveness of adapted interventions. However, moderators were noted in overall health outcomes (i.e., mean age of participants in mixed age groups), aggregated mental health outcomes (i.e., South American participant ethnicity), and aggregated physical health outcomes (i.e., mean age of both younger and older participants in mixed age groups and Cuban participant ethnicity)
van Loon et al^Citation21	N=9 studies Focus = Depression, anxiety (ethnic minority in western world)	Age = U Gender = Women overrepresented Background = Main Groups: African American, Asian American, Latino American	Effect size = Pooled random standardized difference in means on clinical outcome was 1.06 (95% CI =0.51–1.62, p=0.00)	Cultural adaptation works if combined with a focus on patients
	Type = Effectiveness Design of studies = RCT Intervention types = CBT for anxiety and depression Settings = Not reported Outcomes = Anxiety or depression Economic evaluation = Experimental, quasi-experimental, no comparison group	Target problems = Anxiety or depressive problems were described as phobic, panic disorder, posttraumatic stress disorder, major depressive disorder, risk for stress and depression, risk for perinatal depression, and probable depression Acculturation level = Experimental, quasi-experimental, no comparison group Region = USA	Heterogeneity analysis = Considerable heterogeneity among studies Risk of bias analysis = + Full details of risk bias analysis process not provided Search strategy = Adequately described Inclusion/exclusion criteria = Described	cultural values, beliefs, and symptom presentation
Hall et al^Citation15	N=78 studies Focus = Effect Type = Effect Design of studies = Randomization (randomized versus quasi-experimental), treatment versus prevention studies Intervention types = CBT adapted in one third studies (30%) Settings = Experimental, quasi-experimental, no comparison group Outcomes = Mood/anxiety; psychotic; externalizing; general measures of psychopathology Economic evaluation = Experimental, quasi-experimental, no comparison group	N=13,998 Age = U Gender = U Background =95% non-European American Target problems = Mood/Anxiety symptoms, psychotic symptoms, externalizing symptoms (eg, substance use, impulse control), general measures of psychopathology Acculturation level = Experimental, quasi-experimental, no comparison group Region = USA	Effect size = Random-effects multilevel regression model, overall effect size (g=0.67, p<0.001) and medium effect size favoring the effectiveness of culturally adapted interventions over unadapted versions of the same interventions (g=0.52). Adapted interventions had 4.68 times greater odds than other conditions to produce remission from psychopathology (p<0.001) Heterogeneity analysis Risk of bias analysis = Yes Search strategy = Adequately described Inclusion/exclusion criteria = Described	Language of intervention; therapist ethnicity; whether therapy adaptation was top–down or bottom–up moderate effect
Degnan et al^Citation23 (unpublished)	N=46 studies (43 trials) Focus = Schizophrenia Types = Nature of adaptations and efficacy Intervention types = Cognitive behavior therapy interventions Design of studies = RCTs, cluster RCTs, one block RCT and nonrandomized pilot trials Outcomes = Total, positive, negative, and general symptoms Intervention types = FI, PE, CBT, MCT, IMR, SST Settings = Not reported Outcomes = Psychosis Economic evaluation = Experimental, quasi-experimental, no comparison group	N=7,828 Age: 24–57 years Gender: 42% female Background = Various Target problems =100% diagnosis schizophrenia spectrum disorders (F20–29) (including schizophreniform, schizoaffective disorders, delusional disorder or psychosis [NOS]) Acculturation level = Experimental, quasi-experimental, no comparison group Region = (13 different countries, including South/Southeast Asian, Middle Eastern, African, European, Latino/South American) (majority studies 39 (85%) from non-western world; 15% from western world)	Effect size = Pooled weighted SMDs using random-effects models for total symptoms g =−0.23, CI =−0.36 to −0.09; positive symptoms g =−0.56, CI =−0.86 to −0.26; negative symptoms g =−0.39, CI =−0.63 to −0.15; and general psychopathology g =0.75, CI =−1.21 to −0.29 Heterogeneity analysis Risk of bias analysis = Yes Search strategy = Adequately described Inclusion/exclusion criteria = Described Analysis of bias/risk: Cochrane risk of bias (poor reporting in papers). Attrition, intention-to-treat, and pretreatment differences included as moderators in exploring heterogeneity. Funnel plots showed slight publication bias for positive symptoms	Cultural adaptations studied using qualitative research. Adaptations were grouped into nine themes: language, concepts, family, communication, content, cultural norms and practices, context and delivery, therapeutic alliance, and treatment goals

Abbreviations: TAU, treatment as usual; IPT, interpersonal therapy; PST, problem solving therapy; DOT, dynamically oriented art group therapy; MRC, Medical Research Council framework; CSI, culturally sensitive interventions; SMD, serious mental disorders; PT, psychological therapies; FI, family intervention; PE, psychoeducation; CBT, cognitive behavioral therapy; MCT, metacognitive therapy; IMR, illness management and recovery; SST, social skills training; U, uncertain; N/A, not available.

Table 2 Quality assessment of included reviews using AMSTAR* criteria

Review (author)	1. Was a priori design provided?	2. Was there duplicate study selection and data extraction?	3. Was a comprehensive literature search performed?	4. Was the status of publication (ie, gray literature) used as an inclusion criterion?	5. Was a list of studies (included and excluded) provided?	6. Were the characteristics of the included studies provided?	7. Was the scientific quality of the included studies assessed and documented?	8. Was the scientific quality of the included studies used appropriately in formulating conclusions?	9. Were the methods used to combine the findings of studies appropriate?	10. Was the likelihood of publication bias assessed?	11. Was the conflict of interest included?	Global quality of meta-analysis*
Griner and Smith^Citation14	Y	U	Y	U	N	Y	Y	Y	Y	Y	N	7
Huey and Polo^Citation17	Y	N	N	N	Y	Y	Y	Y	Y	Y	N	7
Hodge et al^Citation16	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	N	8
Smith et al^Citation20	Y	Y	Y	Y	N	Y	Y	U	U	Y	N	7
Benish et al^Citation12	Y	N	Y	Y	N	Y	N	U	Y	U	N	8
Chowdhary et al^Citation13	Y	Y	Y	U	Y	Y	Y	Y	Y	Y	N	8
Rojas-García et al^Citation19	Y	U	Y	N	Y	Y	Y	U	Y	Y	N	7
Rojas-García et al^Citation18	Y	U	Y	Y	Y	Y	Y	U	Y	Y	N	8
Sutton^Citation22 (dissertation)	Y	U	Y	Y	N	Y	Y	U	U	Y	N	6
van Loon et al^Citation21	Y	U	Y	N	Y	Y	Y	U	U	Y	N	6
Hall et al^Citation15	Y	N	Y	Y	N	Y	Y	U	U	Y	N	6
Degnan et al^Citation23 (unpublished)	Y	Y	Y	Y	N	Y	Y	Y	Y	Y	N	9

Notes: Indicate whether each criterion is met: Yes (Y), No (N), Uncertain (U). Quality of meta-analyses assessed using a scale of 0–10, where 0 is poor and 10 is excellent.

* Assessment of Multiple SysTemAtic Reviews (AMSTAR) criteria (^*Shea BJ, Hamel C, Wells GA, et al. AMSTAR is a reliable and valid measurement tool to assess the methodological quality of systematic reviews. J Clin Epidem. 2009;62(10):1013–1020).

1. Was a priori design provided? The research question and inclusion criteria should be established before the conduct of the review.

2. Was there duplicate study selection and data extraction? There should be at least two independent data extractors and a consensus procedure for disagreements should be in place.

3. Was a comprehensive literature search performed? At least two electronic sources should be searched. The report must include years and databases used (eg, Central, EMBASE, and MEDLINE). Key words and/or MESH terms must be stated and where feasible the search strategy should be provided. All searches should be supplemented by consulting current contents, reviews, textbooks, specialized registers, or experts in the particular field of study, and by reviewing the references in the studies found.

4. Was the status of publication (ie, gray literature) used as an inclusion criterion? The authors should state that they searched for reports regardless of their publication type. The authors should state whether or not they excluded any reports (from the systematic review), based on their publication status, language, and so on.

5. Was a list of studies (included and excluded) provided? A list of included and excluded studies should be provided.

6. Were the characteristics of the included studies provided? In an aggregated form such as a table, data from the original studies should be provided on the participants, interventions, and outcomes. The ranges of characteristics in all the studies analyzed, for example, age, race, sex, relevant socioeconomic data, disease status, duration, severity, or other diseases, should be reported.

7. Was the scientific quality of the included studies assessed and documented? A priori methods of assessment should be provided (eg, for effectiveness studies if the author(s) chose to include only randomized, double-blind, placebo-controlled studies, or allocation concealment as inclusion criteria); for other types of studies alternative items will be relevant.

8. Was the scientific quality of the included studies used appropriately in formulating conclusions? The results of the methodological rigor and scientific quality should be considered in the analysis and the conclusions of the review, and explicitly stated in formulating recommendations.

9. Were the methods used to combine the findings of studies appropriate? For the pooled results, a test should be done to ensure the studies were combinable, to assess their homogeneity (ie, Chi-squared test for homogeneity, I2). If heterogeneity exists, a random-effects model should be used and/or the clinical appropriateness of combining should be taken into consideration (ie, is it sensible to combine?).

10. Was the likelihood of publication bias assessed? An assessment of publication bias should include a combination of graphical aids (eg, funnel plot, other available tests) and/or statistical tests (eg, Egger regression test).

11. Was the conflict of interest included? Potential sources of support should be clearly acknowledged in both the systematic review and the included studies.

MoherDLiberatiATetzlaffJAltmanDGThe PRISMAGroup2009Preferred reporting items for systematic reviews and meta-analyses: The PRISMAStatementPLoS Med67e100009719621072

 PubMed Web of Science ®Google Scholar

GrinerDSmithTBCulturally adapted mental health intervention: a meta-analytic reviewPsychotherapy (Chic)200643453154822122142

 PubMed Web of Science ®Google Scholar

HueySJrPoloAJEvidence-based psychosocial treatments for ethnic minority youthJ Clin Child Adolesc Psychol200837126230118444061

 PubMed Web of Science ®Google Scholar

HodgeDRJacksonKFVaughnMGCulturally sensitive interventions and health and behavioral health youth outcomes: a meta-analytic reviewSoc Work Health Care201049540142320521205

 PubMed Web of Science ®Google Scholar

SmithTBRodríguezMDBernalGCultureJ Clin Psychol201167216617521105069

 PubMed Web of Science ®Google Scholar

DuvalSTweedieRTrim and fill: A simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysisBiometrics200056245546310877304

 PubMed Web of Science ®Google Scholar

BenishSGQuintanaSWampoldBECulturally adapted psychotherapy and the legitimacy of myth: a direct-comparison meta-analysisJ Couns Psychol201158327928921604860

 PubMed Web of Science ®Google Scholar

ChowdharyNJotheeswaranATNadkarniAThe methods and outcomes of cultural adaptations of psychological treatments for depressive disorders: a systematic reviewPsychol Med20144461131114623866176

 PubMed Web of Science ®Google Scholar

Rojas-GarcíaARuíz-PérezIGonçalvesDCRodríguez-BarrancoMRicci-CabelloIHealthcare interventions for perinatal depression in socially disadvantaged women: a systematic review and meta-analysisClin Psychol Sci Pract2014214363384

 Web of Science ®Google Scholar

Rojas-GarcíaARuiz-PerezIRodríguez-BarrancoMGonçalves BradleyDCPastor-MorenoGRicci-CabelloIHealthcare interventions for depression in low socioeconomic status populations: a systematic review and meta-analysisClin Psychol Rev201538657825797906

 PubMed Web of Science ®Google Scholar

SuttonCCulturally adapted interventions for Latinos: a meta-analysisSchool of Graduate Psychology [Internet] Available from: http://commons.pacificu.edu/spp/1134Accessed May 28, 2015

 Google Scholar

van LoonAvan SchaikADekkerJBeekmanABridging the gap for ethnic minority adult outpatients with depression and anxiety disorders by culturally adapted treatmentsJ Affect Disord20131471–391623351566

 PubMed Web of Science ®Google Scholar

HallGCNIbarakiAYHuangERMartiCNSticeEA meta-analysis of cultural adaptations of psychological interventionsBehav Ther2016476993101427993346

 PubMed Web of Science ®Google Scholar

DegnanABakerSEdgeDThe nature and efficacy of culturally-adapted psychosocial interventions for schizophrenia: a systematic review and meta-analysisPsychol Med2017114

 Web of Science ®Google Scholar