Introduction

Racial and ethnic minority youth (REM) within the United States (US) are a large and growing population. In 2019, for the first time in US history, more than half of all youth identified as a member of a REM group (Frey, 2019). This number is expected to grow such that, by 2060, two in three Americans is projected to be a race other than White (Vespa et al., 2018). Moreover, by 2030, immigration is projected to serve as the primary source of population growth within the U.S. (Vespa et al., 2018). By 2060, further, nearly one in five Americans is expected to be foreign-born (Colby & Ortman, 2015). These demographic shifts underscore the importance of addressing the needs of REM youth within the US.

REM youth have been noted to be at heightened risk for developing mental health problems as compared to their White peers (Alegria et al., 2015). Numerous factors have been thought to contribute to this risk, including challenges associated with socio–economic status, exposure to adverse childhood experiences, family structure across development, and neighborhood-level factors (Alegria et al., 2015). Despite this heightened risk, REM youth are more likely to experience barriers to mental health service use, including higher rates of mental health stigma (Nadeem et al., 2007) and insufficient availability of same language behavioral health providers in the community (Flores, 2009). Further, among those who do gain access to mental health services, REM youth are more likely to prematurely drop out of treatment than their racial/ethnic majority counterparts (de Haan et al., 2018). Without adequate receipt of mental health treatment, these disparities can lead to serious negative, long-term consequences on the well-being of youth (Thapar et al., 2012). Thus, racial and ethnic disparities in mental health service utilization and provision of quality mental health treatment is a serious public mental health concern (Alegria et al., 2011).

In an effort to address these disparities in mental health problems, service utilization, and retention in treatment, cultural adaptations to evidence-based interventions (EBIs) have been highlighted as a solution to meeting cultural competence standards in the delivery of mental health services to diverse populations (Alegria et al., 2011). More specifically, cultural adaptations may be particularly useful with REM populations to address difficulties with treatment engagement, unique risk or resilience factors, and differing presentations of symptoms (Castro et al., 2010). In the following review, literature related to the use of EBIs with REM youth, as well as approaches to cultural adaptations of EBIs, is presented. Finally, empirical research examining the use cultural adaptations is reviewed.

EBIs with REM Youth

A wide range of EBIs for the treatment of psychological concerns among youth currently exist (Chorpita et al., 2011). However, their relevance and generalizability to REM youth has been questioned (Whaley & Davis, 2007). In particular, despite the benefits presented by the greater structure and uniformity offered by these interventions, whether they sufficiently consider culture and context has been called into question (Bernal et al. 2009). A wealth of research has underscored the need to consider treatment preferences and cultural beliefs about mental health (Bernal & Scharró-del-Río, 2001; Cauce et al., 2002). Further, considering that most research supporting EBIs has been conducted with White, middle class individuals, the underrepresentation of REM youth in clinical trials presents concerns regarding the potential generalizability of EBIs (Bernal et al., 2009). Moreover, despite evidence supporting the use of certain EBIs with some REM groups (see Pina et al., 2019), research with certain REM subgroups remains understudied; further, they often fail to include particularly disenfranchised or historically marginalized individuals with the greatest need (Arora et al., 2017). Accordingly, proponents of the ethnic disparity perspective purport that EBIs will be less effective for REM groups given that many treatments were developed by and for European-origin populations and generally incorporate Western principles (De Anda, 1997; Huey & Polo, 2010), which underscores the need for culturally-informed interventions to address the mental health needs of REM youth.

The need for adaptations to EBIs are thus believed to be needed under a number of conditions, including when (1) client engagement is below what is expected, (2) nonsignificant outcomes exist among a particular subgroup, (3) specific culturally-related risk or protective factors should be incorporated into an intervention, or (4) unique symptoms of a common disorder are evident (Barrera & Castro, 2006; Lau, 2006). Others, however, have noted that, the development of interventions designed for and specifically tested with certain target populations may be a more practical and cost-effective strategy for accounting for differential cultural perspectives (Holleran et al., 2008). Thus, systematic and intentional adaptations may help to minimize spontaneous factors for which it may be difficult to account (Ferrer-Wreder et al., 2012).

Cultural Adaptation Frameworks

Cultural adaptations, defined as “the systematic modification of an evidence-based treatment or intervention protocol to consider language, culture, and context in such a way that it is compatible with the client’s cultural patterns, meanings, and values” (Bernal et al., 2009; p. 361–362), have been put forth as a way to support the effectiveness and dissemination of EBIs to diverse cultural groups. Specifically, cultural adaptations are purported to integrate both “top-down,” or universal, and “bottom-up,” or culture-specific, approaches such that the integrity of an original intervention is maintained while simultaneously considering culture and context (Barrera et al., 2013). Several models guiding the cultural adaptation of EBIs for use with REM populations have been proposed. In the following review, frameworks guiding the cultural adaptation of intervention content, outlining the conditions under which cultural adaptations to interventions should be made, and the process of making cultural adaptations to interventions are presented.

First, models describing the content (or the “what”) of adaptations have been put forth. Among the first developed of such models was Bernal’s (1995) Ecological Validity Model (EVM) which proposed that EBIs can be adapted across eight dimensions (i.e., language, persons, metaphors, content, concepts, goals, methods, and context) to better fit the needs of the target population. Originally developed to support the needs of Latinx clients, the model has been generalized to other REM populations (Bernal & Saez-Santiago, 2006). Another prominent cultural adaptation framework focusing on the content of cultural adaptations is the Psychotherapy Adaptation and Modification Framework (PAMF; Hwang, 2006), initially created for adapting interventions for recently immigrated Asian American clients. This framework outlines six domains that researchers and practitioners should consider when adapting EBIs for racially and ethnically diverse populations: (a) dynamic issues and cultural complexities; (b) orienting clients to the process of psychotherapy; (c) understanding cultural beliefs and providing psychoeducation; (d) improving the client-therapist relationship; (e) understanding cultural differences in the expression and communication of distress; and (f) addressing salient cultural issues specific to the population.

The Model of Essential Elements (Podorefsky et al., 2001) offers another guide to considering the content of cultural adaptations to EBIs. Based on the results of an extended sequence of alliance-building efforts, it proposes the following essential elements to consider in adapting content of EBIs: (a) an expansion of the definition of key terms (e.g., depression, resilience) to consider culture and context; (b) increased clinician flexibility; (c) an ecological approach to build partnerships with community, caregiver, and familial systems; (d) an awareness and emphasis of cultural issues. The Surface and Deep Structure Intervention framework proposed by Resnicow et al., 1999 also provides guidance on the content to be considered in the adaptation of EBIs. While surface structure is defined as increasing the acceptability of an intervention by including familiar elements of a target cultural group (e.g., people, places, language, food, etc.), deep structure includes a consideration of how cultural, social, historical, environmental, and psychological factors impact health behavior.

Finally, Kreuter’s (2003) Targeted and Tailored Approaches model underscores five content areas to address within the context of cultural adaptations: (a) peripheral adaptations, which includes edits to colors, imagines, fonts, pictures or titles to better represent the target population; (b) evidential adaptations, which seek to enhance the relevance of the concern by including evidence of its impact on the target population; (c) linguistic adaptations, or adaptations to the language of content; (d) constituent-involving adaptations, or those that seek to involve individuals indigenous to the target population; and (e) sociocultural adaptations, or those that integrate relevant cultural beliefs, behaviors, and values.

Additional frameworks have sought to clarify the conditions under which adaptations to EBIs should be made. For instance, Lau’s (2006) Selective and Directive approach proposed a data-driven approach to determining specific areas where there may be a poor fit between the EBIs and the target population as well as directing the design of the culturally adapted treatment. Building on Lau’s model, Barrera and Castro (2006) introduced the Heuristic Framework for the Cultural Adaptation of Interventions to propose three forms of cultural equivalence to determine whether adaptations may be needed: (a) engagement, (b) action theory, (c) conceptual theory.

Finally, frameworks have also provided guidance in the process of making cultural adaptations. The Cultural Adaptation Process Model (CAP; Domenech-Rodriguez & Wieling, 2004), drawn from Rogers’ (1995) framework of diffusion of innovations, consists of three phases. In the first phase, relevant stakeholders (e.g., community members, intervention developer, and cultural adaptation specialist) collaborate to assess community needs and gather information to inform adaptations to the intervention. The second phase involves the initial adaptation of the intervention followed by pilot work to evaluate the cultural appropriateness of the adapted intervention. The goal of the third and final phase is to refine the adaptations in order to disseminate the intervention. Another such model is Whitbeck’s (2006) Theoretical Model of Culturally Specific Prevention Research. Originally developed for the purpose of developing culturally specific prevention research for Native American communities, this theoretical model proposes the following five stages: (a) identifying key risk and resilience factors in existing research regarding cultural majority groups; (b) identifying key risk and resilience factors from existing research specific to the cultural group; (c) working collaboratively with cultural experts to adapt key risk and resilience factors to fit the target cultural group; (d) identifying or developing measures of unique cultural risk and resilience factors; (e) conducting trials and assessments of the new culturally adapted intervention. Additional frameworks proposed by Leong and Lee (2006) and Burrow-Sanchez et al. (2011) present further guidance in the process of making cultural adaptations to EBIs, with a common focus across all being an emphasis on integrating a participatory approach to intervention adaptation.

Examining Cultural Adaptations to EBIs

Thus, a wide range of frameworks guiding the content and process of cultural adaptations to EBIs have been proposed to provide guidance in addressing the specific needs of REM populations. The efficacy of the resulting culturally adapted EBIs has been examined in a variety of studies; further, a number of systematic literature reviews and meta-analyses have summarized the results of these studies. A review of the literature on the efficacy of cultural adaptations is provided below.

Numerous meta-analyses have sought to examine the effectiveness of culturally adapted psychological interventions on mental health outcomes, with most suggesting that culturally adapted interventions are more effective for REM populations than unadapted or no treatments (Benish et al., 2011; Griner & Smith, 2006; Nagayama Hall et al., 2016; Smith et al., 2011; Smith & Trimble, 2016; Soto et al., 2018). Effect sizes in these studies have ranged from near zero (Huey & Polo, 2008) to large (Chowdhary et al., 2014; van Loon et al., 2013), with effect sizes being moderated by variables such as patient acculturation level, mental health outcome, and study design (i.e., culturally adapted vs. no intervention, culturally adapted vs. another intervention). For instance, studies that adapted an EBI for a specific population reported larger effects than treatments that were culturally adapted for a mix of REM populations (Griner & Smith, 2006; Smith et al., 2011).

Studies have also sought to examine moderation by content of cultural adaptations, with most examining content as defined by Bernal’s (1995) EVM model (e.g., Smith & Trimble, 2016). For instance, Soto et al. (2018) found that cultural adaptations to the spoken or written language of the intervention produced larger effect sizes than those that did not; similar results were also found for inclusion of goals based on cultural values and cultural metaphors in this and other studies (e.g., Smith & Trimble, 2016). Generally, studies that had more cultural adaptations (as defined by Bernal’s (1995) EVM model), had greater impact as measured by larger effect sizes (Griner & Smith, 2006; Smith & Trimble, 2016; Soto et al., 2018). Fewer studies have examined the content of cultural adaptations defined in alternative ways. For instance, in line with Resnicow’s model of cultural adaptations, Escobar and Gorey (2018) found that “deep structure” cultural adaptations were more effective than “surface structure” for nonadapted psychological interventions for depressed Hispanic adults. Thus, while initial attempts examining content of cultural adaptations have been made, they have primarily been restricted in their definitions of content. Moreover, despite the importance of and models guiding the process of making cultural adaptations, few studies have sought to examine the use of participatory approaches to intervention adaptation in their systematic reviews of existing studies. In one such study, a systematic review of the literature on adaptations of treatments for depressive and anxiety disorders for REM adult populations within Western countries, van Loon et al., 2013 reported on whether nine studies incorporated collaboration with or were advised by members of the target group within the context of focus groups, finding that two of the nine studies involved some aspects of participatory-based approaches. In a meta-analysis of culturally adapted interventions for youth and adults, Smith & Trimble (2016) found that 46% of 79 studies “indicated that they had developed the cultural adaptations through consultation with individuals from the culture” (p. 136).

With regards to age, many of the above systematic reviews have limited their scope to examinations of cultural adaptations among adult samples (e.g., Chowdhary et al., 2014; Escobar & Gorey, 2018; van Loon et al., 2013), with fewer having exclusively examined the use of culturally adapted interventions among youth (i.e., Hodge et al., 2010; Hodge et al., 2012; Huey & Polo, 2008; Jackson et al., 2010; Pina et al., 2019). While general analyses have demonstrated small to medium effect sizes of culturally adapted interventions for youth, most have not have systematically examined either the content or process of cultural adaptations of these studies (Hodge et al., 2010; Hodge et al., 2012; Jackson et al., 2010). Most recently, Pina et al. (2019) reported the models that researchers used to adapt their interventions but did not expand on the content of adaptations in individual studies nor review the process by which the adaptations were made.

Considering the aforementioned critiques that cultural adaptations to interventions are insufficiently described in the youth mental health literature (Huey & Polo, 2008), additional detail regarding the content and process of cultural adaptations made to youth psychological intervention in the US is generally needed. Increased consistency in descriptions of content of cultural adaptations to psychological interventions for REM youth would allow for improved comparison across studies (Huey & Polo, 2008). Additionally, an improved understanding of the content and process of cultural adaptations to psychological interventions for REM youth populations would permit an enhanced examination of the impact of these adaptations on relevant outcomes. Thus, additional information on the content and process of culturally adapted interventions for REM youth populations is needed.

Current Study

There is a high prevalence of mental health problems among youth, with REM youth in particular experiencing unique risk factors and barriers to mental health treatment. To date, few reviews exclusively examining cultural adaptations of EBIs for youth mental health have been conducted; moreover, despite the importance in understanding such variables, none explicitly review the content of the cultural adaptations made nor describe the process by which cultural adaptations are made to EBIs. The goal of the current study was to engage in a comprehensive systematic review of existing culturally-adapted psychological interventions for REM youth within the U.S. seeking to systematically summarize information missing from past reviews.

Method

Initial Search

A systematic three-stage review was conducted by the research team. (See Fig. 1.) The initial stage included a comprehensive search of two scholarly databases, PsychInfo and PubMed. Search terms representing mental health (i.e., psych*, mental health), intervention (i.e., treat*, interven*, therap*, prevent*), REM status (i.e., ethn*, minorit*, Asian, Afr*, Black, Latin*, Native American, Hispanic, rac*, Pacific), and cultural adaptation (i.e., modif*, adapt*, cultur* specific, cultur* sensitive, cultur* appropriate, cultur* informed) were developed collaboratively by the research team based on their expertize and an informal review of relevant literature to best represent concepts of interest in the study. Keywords were used to search titles and abstracts within the two databases. As noted above, the Boolean search modifier “*” was used with certain keywords to identify studies with the same root word and different word endings. The Boolean search operators “AND” and “OR” were used in order to conduct a single search with all selected keywords. The search was restricted to articles published within peer-reviewed journals, written in English, and published between 1980 and 2018. This initial step yielded a total of 6487 articles.

Fig. 1
figure 1

Flow Chart of the Search Strategy and Inclusion/Exclusion Process

Full size image

Titles and abstracts of these articles were manually reviewed to determine whether articles were relevant to the previously described keywords. Additionally, duplicate articles were removed. As a result of these steps, 6223 articles were removed; 264 articles remained. Further, all meta-analyses, systematic reviews, and literature reviews (i.e., 44 articles) were removed. These 44 articles were supplemented by additional relevant meta-analyses, systematic reviews, and literature reviews (i.e., 10 articles; e.g., Kumpfer et al., 2012; Peterson et al., 2017; Smith & Trimble, 2016). References within these meta-analyses, systematic reviews, and literature reviews (i.e., 54 articles) were then manually reviewed for relevance to the keywords and potential inclusion in the systematic review. In combination with the results from the previous searches, a total of 298 articles were identified.

Inclusion Criteria

In the next stage, three research team members screened the eligibility of the 298 potentially relevant studies through an in-depth examination of the full manuscripts. Specifically, for each article, two reviewers independently examined its eligibility based on the inclusion criteria: (1) the article described or evaluated a cultural adaptation to a psychological intervention, (2) the cultural adaptation was made to address the needs of REM populations, (3) and the intervention had the goal of addressing at least one mental health outcome. Studies that outlined a theoretical approach to the cultural adaptation of interventions were also included; studies that described or evaluated interventions that were originally developed for a REM group were excluded. Additionally, only studies that explicitly stated adaptations were based on culture, ethnicity, or race were included; studies that were primarily adapted for other characteristics (i.e., gender, socio–economic status, sexual orientation, etc.) were excluded.

Interrater reliabilities for the inclusion and exclusion criteria were calculated using Cohen’s kappa (Cohen, 1960). A reliability coefficient of 0.590 was obtained, representing moderate agreement between coders (Landis & Koch, 1977). When disagreement arose in terms of eligibility, the reviewers as well as the principal investigator discussed each discrepancy until agreement was achieved. After this review, 156 articles remained.

Article Coding

The next stage involved coding to extract information from the final pool of 156 articles. The codes were collaboratively developed by the research team based on a review of the literature and existing systematic reviews of the cultural adaptations to psychological interventions. Article type (e.g., theoretical, descriptive, or evaluative) was coded. Theoretical articles were defined as those that proposed a theory of cultural adaptation research, including defining the content or types of cultural adaptations, clarifying the conditions under which cultural adaptations to EBIs should be made, or detailing the process by which to make cultural adaptations. Descriptive articles were defined as those that described the adaptation of a specific intervention without formally evaluating the effectiveness, feasibility, or acceptability of the adapted intervention. Evaluative articles were defined as those which formally evaluated the effectiveness, feasibility, or acceptability of the adapted intervention. For theoretical articles, the name of the theory, the content or types of cultural adaptations, and/or the number and description of stages of the cultural adaptation process were coded. Descriptive and evaluative articles were coded for participant characteristics (e.g., population age, population risk, race/ethnicity of target audience, country), intervention characteristics (e.g., name of the original program, name of the adapted program, target mental health outcome(s), intervention format, delivery setting, intervention orientation, interventionist), and cultural adaptation characteristics (e.g., name and description of theory (if used) to guide cultural adaptation process, content of cultural adaptations, individuals involved in adaptation process, whether or not participatory/partnership-based research methods were utilized). For most categories, information that was not explicitly stated was coded as “unknown” and information that was not captured by a code was coded as “other” with a qualitative description provided. The majority of codes required coders to “select all that apply.”

While the initial search did not restrict the parameters to those studies conducted only with certain age groups (e.g., youth) or within certain counties, only studies coded for participant age as children (i.e., ages 0–18), parents of children, or families as the sample population were included in the current study; further, only descriptive or evaluation studies conducted within the US were included. Theoretical articles or studies conducted with adults (i.e., >18 years of age) as the population age group and outside of the US were excluded. Thus, a total of 52 studies were included in the current study.

Coding of Cultural Adaptation Content

Codes for the content of cultural adaptations were created by the research team based on a review of common cultural adaptations cited in the literature. Specifically, meta-analyses and systematic reviews noted above (e.g., Chowdhary et al., 2014; Nagayama Hall et al., 2016; Smith et al., 2011) were reviewed to determine how, if at all, cultural adaptation content was coded. All types of cultural adaptations coded by these articles were included in a list of content of cultural adaptations. As the goal was to create a broad and comprehensive list, the research team engaged in an informal discussion to expand on findings and create a more detailed list. For example, many of the previous meta-analyses and systematic reviews utilized Bernal’s (1995) eight dimensions of cultural adaptations. These (as well as content specified by other frameworks) were incorporated and deconstructed into the smallest components to increase specificity of content of cultural adaptations that were coded.

As a result of the aforementioned literature review and informal discussions among the research team, a final list of 35 cultural adaptation content types was used. Articles were coded to denote whether the original intervention was culturally adapted in terms of language (e.g., translating the spoken language of the intervention, translating the written intervention materials), literacy (e.g., adjusting literacy levels, reducing tasks requiring literacy), interpersonal variables (e.g., racial-matching of clients and interventionists, involving family or community supports), training (e.g., adding specific training for interventionalists), intervention characteristics (e.g., adapting the format of interventions, changing delivery setting), intervention content (e.g., using culturally relevant symbols or images, incorporating lifestyle elements such as food or music), mental health literacy (e.g., adding psychoeducation about mental health treatment, addressing shame or stigma related to mental health treatment), assessment methods (e.g., adapting the format of the evaluation, translating assessment measures to the target language), and treatment outcomes (e.g., addressed culturally relevant treatment goals). Finally, raters were also allowed to write in additional types of cultural adaptation content identified within a study that were not included in our coding manual.

Training and Reliability

Five coders were trained in the use of the codes. In particular, codes were thoroughly reviewed as a team and operationalized. Practice coding of eight articles was then conducted among coders. Coders met weekly to discuss discrepancies in coding, clarify any inconsistencies in interpretation of codes, and add or adjust codes as needed. In particular, cultural adaptation content codes were consistently reviewed by all coders to further define codes and to add codes for observed types of cultural adaptation content that had not previously been included. When practice coding was complete and the codes were finalized, each article was assigned to a coder using a random number generator. For reliability purposes, a portion of the articles were coded twice (n = 48; 30.7%). Interrater reliability was calculated using Cohen’s kappa (Cohen, 1960). A reliability coefficient of .757 was obtained, representing substantial agreement between coders (Landis & Koch, 1977). Disagreements in coding were discussed and manually resolved through consensus coding to ensure accuracy of codes. Post-discussion interrater reliability was calculated, with a reliability coefficient of 1.00 obtained, representing nearly perfect agreement (Landis & Koch, 1977).

Results

Article Type and Characteristics

A total of 52 studies were included in the current review. Full study characteristics are reported in Table 1. Sixteen studies (30.8%) were descriptive studies, in that they described the development of a culturally adapted intervention program and a formal evaluation was not conducted; the remainder (n = 36 studies; 69.2%) were considered to be evaluative studies, defined as those that formally evaluated the effectiveness of a culturally adapted intervention on improving target outcomes, on feasibility of implementation, and/or on the participants’ acceptability of the intervention. Forty-eight studies (92.3%) indicated which original programs were adapted, while six studies (11.5%) did not indicate the name of the original program that was adapted. Across studies, there were 34 unique original programs that were culturally adapted. The most common programs that were adapted were Parent Child Interaction Therapy (PCIT; n = 4; 7.7%); Parent Management Training – Oregon Model (PMTO; n = 4; 7.7%); Strong Teens (n = 3; 5.8%); and Trauma-Focused Cognitive Behavioral Therapy (TF-CBT; i.e., n = 3; 5.8%). For three studies (5.8%), components of more than one original program were used in the cultural adaptation process (i.e., Chavira et al., 2018; Parra-Cardona et al., 2017; Weiner et al., 2009). The most common culturally adapted programs were Criando conn Amor: Promoviendo Armoní y Superación (CAPAS; n = 5; 9%), Guiando a Ninos Activos/Guiding Active Children (GANA; n = 3; 5.7%), Accommodated Cognitive Behavioral Therapy (A-CBT; n = 2; 3.8%), and Jóvenes Fuertes (n = 2; 3.8%).

Table 1 Study characteristics
Full size table

Participant Characteristics

Population Age

Twenty studies (38.5%) included exclusively youth populations (i.e., individuals 18-years-old or younger), while 15 (28.9%) included exclusively parents/caregivers of youth populations. Seventeen studies (32.7%) included both youth and parent/caregiver populations.

Population Risk

Half of the studies (n = 26; 50.0%) were implemented with clinical populations (i.e., individuals with a mental health diagnosis or moderate to severe symptoms). Studies were also implemented with at-risk populations (n = 10; 19.2%), defined as individuals at-risk for developing a mental health disorder, and general populations (n = 11; 21.15%), defined as any individuals in the population. A few studies (n = 3; 5.8%) were implemented with a combination of at-risk and clinical populations. Two studies (3.8%) did not indicate the population risk of a clinical diagnosis (i.e., Domenech-Rodríguez et al. 2011; Ngo et al. 2008).

Race/Ethnicity of Target Audience

Culturally adapted interventions were targeted toward the following groups: Hispanic or Latinx youth (n = 33; 63.5%); African American or Black youth (n = 7; 13.5%); Indigenous American youth (n = 5; 9.6%); and Asian and Pacific Islander youth (n = 3; 5.8%). Two studies (3.9%) targeted both African American and Latinx youth (i.e., Misurell & Springer, 2013; Saulsberry et al. 2013). For the remaining six studies (11.5%), the target audience for the culturally adapted interventions were described broadly, for instance by noting, “racial and ethnic minority,” “culturally diverse,” or “culturally and linguistically diverse” youth.

Culturally-adapted Intervention Characteristics

Target Mental Health Outcome

The most frequently addressed mental health outcomes of the included studies were behavioral or externalizing problems (n = 15; 28.9%), post-traumatic stress disorder (PTSD) or trauma symptoms (n = 10; 19.2%), and mood disorders (e.g., bipolar, depression, etc.; n = 10; 19.2%). Several studies targeted individuals experiencing substance use (n = 5; 9.6%) or anxiety (n = 4; 7.7%), while social-emotional skill development was the target of intervention in four studies (7.7%). Further, there were a few studies that targeted individuals experiencing stress (n = 3; 5.8%), interpersonal problems (n = 2; 3.9%), suicidality (n = 1; 1.9%), psychotic disorders (n = 1; 1.9%), or parenting skills (n = 1; 1.9%). Overall, the studies generally identified one primary mental health outcome; however, three studies (5.8%) indicated two or three primary mental health outcomes (i.e., Ballard et al., 2017; Hurwich-Reiss et al., 2014; McDonald et al., 2012).

Intervention Format

The most frequently utilized format by which the intervention was delivered was via a parent/caregiver group (n = 19; 36.5%). Groups of youth participants (n = 16; 30.8%) were also commonly used formats. Studies also described interventions delivered to individual families (n = 11; 21.2%) or individual children or parents (n = 10; 19.2%). Other intervention formats incorporated included classroom-based interventions (n = 5; 9.6%), parent–child dyads (n = 3; 5.8%), and groups of families (n = 2; 3.9%). Other intervention formats (e.g., parent/caregiver–child group, teacher group, and media-based campaign) were included in four studies (n = 4; 7.7%). Eleven studies (21.2%) incorporated more than one intervention format (e.g., interventions which included child group, parent group, and teacher group sessions).

Delivery Setting

The most common delivery settings were primary and secondary schools (n = 18; 34.6%). Community settings whose primary function was not the provision of mental health services (e.g., local community centers) were also well represented in the selected studies (n = 12; 23.1%). A number of interventions were implemented in outpatient mental health clinics (n = 6; 11.5%), hospitals (n = 5; 9.6%), religious settings (n = 5; 9.6%), and clients’ homes (n = 4; 7.7%). Four (7.7%) were conducted over the phone while one (1.9%) was conducted via a digital platform. University clinics served as the delivery setting in two studies (3.9%). Eight interventions (15.4%) were conducted in more than one delivery setting (e.g., intervention includes some school-based sessions and some home-based sessions). The delivery setting was not reported in six studies (11.5%).

Theoretical Orientation

The theoretical orientation of the adapted intervention as described by authors was also coded. The most commonly described orientation of the adapted interventions was cognitive-behavioral therapy (CBT; n = 19; 36.5%). Several studies described the adapted intervention orientation as family therapy (n = 8; 15.4%) or behavioral therapy (n = 6; 11.5%). Play therapy (n = 3; 5.8%) and dialectical behavior therapy (n = 2; 3.9%) were identified as the intervention orientation in a small number of studies. Additional intervention orientations (e.g., interpersonal psychotherapy, mindfulness, problem solving therapy, psychoeducation) were endorsed in a small number of studies each (n = 1 each; 1.9%). In seventeen studies (32.7%), the authors did not explicitly state the intervention orientation of the adapted intervention. Five studies (9.6%) indicated that the intervention incorporated more than one theoretical orientation.

Interventionist

Interventions were most frequently implemented by mental health professionals (e.g., psychologists, psychiatrists, guidance counselors, social workers, psychiatric nurses; n = 31, 59.6%). Common interventionalists also included graduate/undergraduate students (n = 20; 38.5%) and lay professionals, defined as those not in a mental health or health professions (n = 10; 19.2%). Health professionals, defined as professionals providing physical health treatment (e.g., physicians, nurse practitioners; n = 4; 7.7%), parents or caregivers (n = 2; 3.9%), or peers (n = 2; 3.9%) served as interventionalists in a small number of studies. Three studies (5.8%) did not explicitly state the level of training or background of the interventionalist. Seventeen studies (32.7%) reported having more than one type of interventionalist, with the most common combination of co-interventionalists being mental health professionals and lay professionals (n = 5; 9.6%) or mental health professionals and graduate students (n = 4; 7.7%).

Cultural Adaptation Characteristics

Theory

Half of the studies (n = 26; 50.0%) identified a cultural adaptation framework used to guide the process of culturally adapting the intervention, with 14 unique cultural adaptation frameworks identified among the studies. The most common cultural adaptations theories used were Bernal’s (1995) EVM (n = 14; 26.9%) and Domenech-Rodriguez and Weiling’s (2004) CAP (n = 4; 7.7%). Other theories highlighted in more than one study included Resnicow et al. 1999 surface structure changes and deep structure changes framework (n = 3; 5.7%), Kreuter’s (2003) Targeted and Tailored Approaches (n = 2; 3.8%), Hwang’s (2006) PAMF (n = 2; 3.8%), and Burrow-Sanchez et al. (2011) Cultural Accommodation Model for Substance Abuse Treatment (n = 2; 3.8%). Eight studies (15.4%) used two cultural adaptations theories to guide the cultural adaptation process, with the EVM and the CAP frameworks as the most frequent combination.

Cultural Adaptation Content

The most common content of cultural adaptations was incorporating culturally relevant risk factors (e.g., discrimination, acculturation, intergenerational family conflict, migration) (n = 35; 67.3%). Translating the spoken language of the intervention (n = 30; 57.7%), incorporating cultural values (e.g., familism) and traditions (n = 30; 57.7%), having therapist-client match on a variable other than race/ethnicity (n = 28; 53.9%), or incorporating culturally relevant examples, scenarios, and stories (n = 27; 51.9%) were the content of other frequently implemented cultural adaptations. Several studies implemented culturally-informed assessment measures in the evaluation (n = 24; 46.2%), translated the written language of the intervention (n = 23; 44.2%), or incorporated culturally appropriate and syntonic language (n = 22; 42.3%). Additionally, twenty-one studies (40.4%) translated the assessment measure to the target language of the participants. Overall, thirty-six (67.9%) studies reported the use of a language other than English. Of these studies, interventions were most commonly delivered in Spanish (n = 31; 86.1%). Six studies (16.7%) used other languages, such as “African American language expression,” Chinese, Dakota, Hmong, Karen, Korean, Omaha, and Polish. Culturally-informed therapeutic techniques (n = 19; 36.5%) or culturally relevant strengths or protective factors (n = 17; 32.7%) were also incorporated in several studies. Sixteen studies (30.8%) included therapist-client race/ethnicity match as the content of the cultural adaptation, while, in fifteen studies (28.9%), the intervention was culturally adapted by directly involving the client’s family. A number of studies added culturally-relevant training for staff (n = 14; 26.9%) or provided extra services (n = 14; 26.9%) (e.g., child care, caseworker, transportation) as the content of the cultural adaptations. Thirteen studies (25.0%) were culturally adapted by changing the length of the intervention. Several studies culturally adapted the intervention by reducing the focus of tasks requiring literacy (n = 12; 23.1%). Further, several studies addressed shame or stigma related to the mental health disorder and/or treatment (n = 12; 23.1%) by explicitly discussing shame or stigma with participants or changing the content or language used in the intervention for the purposes of reducing shame or stigma. Culturally relevant symbols or images were incorporated as cultural adaptations in eleven studies (21.2%). Cultural adaptations included in ten studies each included addressing culturally relevant goals for the intervention, changing the setting of the intervention, or changing the targets of the evaluation (19.2% each). The remaining cultural adaptations were mentioned in fewer than 10 studies. See Table 2 for a list of most commonly reported cultural adaptations and their corresponding frequencies. See Supplemental Table 1 for further descriptions and examples of cultural adaptations. Of note, one pair of cultural adaptation types was frequently (i.e., greater than 75% of the time) implemented together (i.e., translating the spoken language of the intervention and having therapist-client match on a variable other than race/ethnicity [n = 26]).

Table 2 Content of Cultural Adaptations
Full size table

Individuals Involved

Researchers (n = 42; 80.8%) and practitioners (n = 24; 46.2%) represented the primary individuals responsible for culturally adapting the original interventions. Stakeholders (n = 11; 21.2%) were also directly involved in making cultural adaptations in several studies. Youth were involved in making cultural adaptations in one study (1.9%). Five studies (9.6%) did not provide sufficient information to be able to determine who made the cultural adaptations to the intervention. Twenty-five studies (48.1%) indicated that a combination of individuals with different areas of expertize worked together to make the adaptations, with the most common collaborations being between researchers and practitioners (n = 13; 25.0%) or among researchers, practitioners, and stakeholders (n = 6; 11.5%).

Participatory-Based Approaches

A majority of studies (n = 35; 67.3%) noted incorporating aspects of participatory-based approaches in the process of culturally adapting interventions. Some examples of such approaches included: (a) collaborating with stakeholders to identify community needs, barriers, and program goals; (b) conducting focus groups with community stakeholders to inform intervention adaptations; and (c) involving stakeholders in the cultural adaptation process (e.g., planning the adaptations, creating the adapted materials, co-implementing the adapted intervention, co-evaluating the culturally-adapted intervention, and further refining the intervention). Two studies (3.8%) explicitly reported using community-based participatory research (i.e., Ijadi-Maghsoodi et al., 2017, Parra-Cardona et al., 2012). Seventeen studies (32.7%) did not indicate using any participatory-based approaches.

Discussion

In response to the limited information on the content (i.e., “what”) and processes (i.e., “how”) of cultural adaptations to psychological interventions for REM youth, we engaged in a comprehensive systematic review as part of which data from 52 existing studies on culturally-adapted EBIs with REM youth within the U.S. were systematically extracted and summarized. Results of this review provide a detailed narrative of the aspects common among cultural adaptations to psychological interventions for REM youth populations. Further, results of the study expand on the scope of previous systematic reviews through in-depth examination of the “what” and “how” of cultural adaptations to better inform replication and improvements to future cultural adaptation efforts. Given the heterogeneity across study designs and intervention characteristics, aggregating effect sizes to summarize the efficacy of culturally adapted interventions has been put into question in previous meta-analyses (Huey & Polo, 2008). Thus, this study extends previous findings by providing a narrative of the cultural adaptation content and process to capture this heterogeneity. Additionally, as adults (e.g., parents/guardians) serve as important gatekeepers to youth’s mental health service use (Arora & Khoo, 2020; Rickwood et al., 2007), this study adds to existing research by including studies not only with youth but also parents and other caregivers as intervention participants.

When reported, the most common cultural adaptation framework used to guide adaptations among studies was Bernal’s (1995) EVM. This is consistent with previous reviews of cultural adaptations literature with adults (e.g., Chowdhary et al., 2014; Nagayama Hall et al., 2016; Smith et al., 2011). Further, this is in line with current reviews, though limited, of youth mental health cultural adaptations literature (Pina et al., 2019). This study extends current knowledge by incorporating existing frameworks while increasing specificity of cultural adaptation content, resulting in a comprehensive list of 43 unique cultural adaptations. These 43 unique cultural adaptations are aligned with existing cultural adaptation frameworks, including the EVM Bernal’s (1995), the PAMF (Hwang, 2006), the Model of Essential Elements (Podorefsky et al., 2001), the Surface and Deep Structure Intervention framework (Resnicow et al., 1999) and the Targeted and Tailored Approaches model (Kreuter’s (2003)), though provide a more detailed approach to defining cultural adaptation content thus facilitating consideration of and communication about cultural adaptations being made to EBIs across studies. Moreover, the 43 cultural adaptation types go beyond each individual framework to provide a more inclusive list of cultural adaptation content.

As noted, this study extends previous findings by adding to literature on the content of cultural adaptations to interventions for youth mental health in particular. Specifically, this study demonstrates how researchers and practitioners have translated existing frameworks into practical application via the description or evaluation of culturally adapted EBIs. The most common content of cultural adaptations types included incorporating culturally relevant risk factors, translating the spoken language of the intervention, incorporating cultural values and traditions, having therapist-client match on a variable other than race/ethnicity, or incorporating culturally relevant examples, scenarios, and stories. These cultural adaptation types are not surprising given the focus on these areas within frameworks (e.g., EVM; Bernal’s (1995)), with a balance between both surface and deep (Resnicow et al., 1999) or targeted and tailored approaches (Kreuter’s (2003)).

Results also demonstrate the frequent use of aspects of participatory-based approaches in the process of culturally adapting interventions (n = 35; 67.3%). This finding underscores the weight placed on engaging the community directly to meet the needs of REM youth within the context of youth mental health cultural adaptations research. Despite the strong emphasis on these approaches in the studies presently reviewed, previous reviews of cultural adaptations to psychological interventions for REM youth have not reported on the use of participatory-based practices (Hodge et al., 2010; Hodge et al., 2012; Huey & Polo, 2008; Jackson et al., 2010; Pina et al., 2019).

Active involvement of community members and organizational representatives throughout all aspects of the research process remains a crucial aspect of participatory based approach to intervention development (Israel et al., 2001). Results indicated that researchers and practitioners were the primary individuals responsible for culturally adapting interventions, with fewer studies involving stakeholders in this process and only one study involving youth directly. This finding expands on those from existing reviews and underscores the need for further involvement of relevant stakeholders, including youth, in the process of making cultural adaptations to psychological interventions for REM youth.

Notably, 63% of studies (n = 33) addressed culturally-adapted interventions targeting Hispanic/Latinx youth populations, with much smaller percentages targeting other REM youth. Many (e.g., Pina et al., 2019, Soto et al., 2018) have found similar results regarding the underrepresentation of African American/Black, Asian, and Indigenous Americans in the cultural adaptations literature. Additional examinations of culturally-adapting interventions targeting African American/Black, Asian and Pacific Islander, and Indigenous American youth populations are thus needed.

Limitations

Despite the study’s contributions, several limitations exist. The decision to conduct a systematic literature review instead of a meta-analysis is a key limitation. While providing additional evidence on the effectiveness of cultural adapted interventions to psychological interventions for REM youth may have proved beneficial, the primary purpose of this systematic literature review was to contribute to gaps in the current literature by accruing a comprehensive list of the content of cultural adaptation psychological interventions, as well as the data regarding the process by which these adaptations are being made. Additionally, we chose to limit studies to those that described or examined the use of culturally adapted interventions with youth populations within the US in particular. This decision was made in order to best capture the uniqueness of this population, both in terms of development and location. Future research may seek to examine both the content and the process of cultural adaptations to psychological interventions with adult samples or with samples outside of the U.S.

Further, it is important to underscore concerns with cultural adaptations to EBIs generally. Even though cultural adaptations seek to solve the one-size-fits-all issue of EBIs, the creation of adaptations for a specific cultural subgroup may also inadequately respond to the heterogeneity within a cultural subgroup (Castro et al., 2010). One solution to this issue is the use of population segmentation to identify smaller, homogenous subcultural groups to effectively focus cultural adaptations for each of these subcultural groups (Castro et al., 2010). Another solution proposed is the use of “adaptive” intervention designs (Collins et al., 2004), which mirrors individualized approaches in regular clinical practice by providing explicit guidelines for decision rules with regard to cultural adaptations rather than providing a menu of options. However, the cultural competence of the intervention developers and interventionists requires further attention in determining these decision rules. Such concerns, while relevant to the study of cultural adaptations, are not addressed within the context of this study.

Implications

The results of this study have implications for future research on the cultural adaptation of mental health interventions for youth. First, results of this study indicated that some of the studies reviewed were lacking in detail regarding the content of and process by which cultural adaptations were created. Further, in some cases, limited information about the cultural adaptations process resulted in a lack of clarity regarding why particular adaptations were chosen for specific REM youth populations. In addition to resulting in “unknown” codes in the current study, the underreporting of this information may lead to difficulty for others in replicating and improving upon culturally-adapted interventions. Future studies should take care to thoroughly report on this information. Further, journals that publish these types of studies should also seek to ensure that such essential information is included in published studies.

Findings from this study also underscore the need to further investigate and delineate best practices in cultural adaptations of EBIs with youth. In particular, this study identified 43 unique types of cultural adaptations presented in the literature with each study reporting up to 19 different types of cultural adaptations. Given the range of implemented adaptations and our current understanding, it is difficult to determine the threshold for a culturally-adapted intervention to be deemed as effective. Further, while the current study presents the most frequently implemented types of cultural adaptations for REM youth generally, data on the most salient types of cultural adaptations for specific REM groups is not yet available. As such, questions such as, “What is the optimum combination of cultural adaptations that would lead to an effective intervention above and beyond an unadapted intervention?” remain unanswered. Previous research has noted that the quantity and quality of cultural adaptations is positively related to the adapted intervention’s effectiveness (Escobar & Gorey, 2018; Soto et al., 2018; Smith & Trimble, 2016; Smith et al., 2011). Future research should thus aim to provide guidance to practitioners and researchers to determine when an EBI has received adequate cultural adaptations to be effective and acceptable.

Additionally, despite the current literature outlining several theories to guide the content and process of making cultural adaptations, there is no standardized method for either determining the content of cultural adaptations an intervention requires or assessing whether a program has been adequately adapted to a target population. Future research helping to support such efforts via the development of, for instance, standardized tools for rating cultural adaptations made to EBIs may prove beneficial. Such tools could, for instance, bridge the gap between the conceptual frameworks for cultural adaptation and the practical implementation of those frameworks by researchers and their community partners during the process of culturally adapting evidence-based interventions for REM populations.

Implications for practitioners will be informed by establishing greater consensus among researchers and practitioners regarding the content of and process for adequately culturally adapting EBIs, as well as creating standardized tools for the development and evaluation of culturally-adapted interventions. This process may best be accomplished by integrating the input from both researchers and practitioners (i.e., stakeholders) with diverse viewpoints and relevant experience. Such a process could more systematically seek to obtain consensus via, for instance, a Delphi method (Arora et al., in preparation). Once this has occurred, practitioners will be able to more systematically apply evidence-based cultural adaptations to EBIs. Since the majority of the culturally-adapted interventions target Latinx/Hispanic populations (e.g., Pina et al. 2019, Soto et al. 2018), it will be especially important for practitioners to apply cultural adaptation procedures to interventions targeting African American/Black, Asian and Pacific Islander, and Indigenous American youth populations.