A scoping review of Dialectical Behavior Therapy with patients with minoritized identity constructs

Abstract

Objective.

Dialectical Behavior Therapy (DBT) is an evidence-based psychological intervention (EBP) for reducing suicide behavior. While DBT, like many EBPs, was developed and first evaluated in largely white samples in Western contexts, it is now broadly applied. This review assessed how DBT has been adopted or adapted when used with patient populations with minoritized identity constructs.

Methods.

We conducted a PRISMA-guided scoping review¹ of articles using DBT with a minoritized patient population. We used an implementation science framework, Framework for Reporting Adaptations and Modifications-Enhanced,² to characterize adaptations to the content and context of DBT application. Through PubMed, Embase and PsycINFO, we identified 273 articles where the title or abstract mentioned DBT and described a minoritized adult patient population.

Principle Findings.

We found 83.5% (228/273) articles reviewed reported adaptations to DBT content. Specifically, 66.3% (181/273), 44.6% (122/273), and 39.6% (108/273) of articles described reducing, expanding, or changing DBT content, respectively. We found that 42.1% of articles (115/273) reported adaptations to the DBT context. Moreover, 57.5% of content adaptations and 65.3% of context adaptations did not specify a reason for adaptation.

Conclusions.

This scoping review indicates that the majority of articles of studies using DBT with patients with minoritized identity constructs made adaptations to DBT, but few provided justification for these adaptations. We encourage clinicians and clinical administrators, researchers, and policymakers to document adaptations and justifications thereof, and evaluate clinical and implementation outcomes.

Introduction

While global suicide rates have decreased between 2000 and 2019,³ many minoritized identity groups have experienced an increase in suicide behaviors,^4–7 suggesting the need to examine health inequities. For example, in the United States, between 2018 and 2021, age-adjusted suicide rates increased substantially among non-Hispanic American Indian or Alaska Native persons (26% increase), non-Hispanic Black or African American persons (19.2% increase), and for Hispanic persons (6.8% increase), while the age-adjusted suicide rate among non-Hispanic White persons declined (a 3.9% decrease⁶). From 2018 to 2022, suicide attempts among transgender and non-binary young people increased 7–72% in states that enacted anti-transgender laws.⁸ While evidence-based psychotherapies (EBPs) have been demonstrated to reduce suicide behavior,^9–11 many of these treatments were developed and evaluated in predominantly white, Western, high socioeconomic contexts where sexual orientation and gender identity were not systematically assessed. As pointed out in a Health Equity Summit convened by the Agency for Healthcare Research and Quality,¹² the limited representation of historically minoritized individuals in the development of EBPs may minimize visibility of relevant differences and perpetuate inequities. As a result, EBPs found effective in one population may not work as well in another population.^13–15 For example, a review of well-established or efficacious EBPs found that in eight studies where ethnicity was a moderator of treatment effect, in all but one study, fewer benefits were found for ethnic minority than for European American participants.¹⁶ Some stated that EBP effectiveness cannot be assumed to apply to populations different than the original study population¹⁷, and some studies show client characteristics like race or racial composition of a therapy group moderate treatment effects.^18,19

A discrepancy between the sample with which a treatment was originally studied or validated and the identity constructs of the patient needing treatment (e.g., race, ethnicity, gender, language) or the treatment context (e.g., setting, treatment deliverers, etc.) can pose a challenge for clinicians and administrators who wish to provide effective, evidence-based care. A patient’s personal identity constructs can influence their approach to and experience of psychotherapy,^20–22 consistent with data showing race or ethnicity may be associated with treatment outcome^23,24 or retention in treatment,^25,26 suggesting an opportunity to consider tailoring EBP content (what treatment is provided, what examples are used, what strategies or values are emphasized) to the individual patient.¹⁴ Delivering suicide prevention EBPs in clinical practice settings different than the settings in which the EBP was initially developed and tested also presents an opportunity to consider the impact of treatment context: where and how the treatment is delivered, including contextual variables such as resource availability, and social, economic, historical, and political forces that shape these interventions’ delivery in their settings.^17,27–29

Dialectical behavior therapy is a multi-modal psychotherapy that targets severe emotion dysregulation and behavioral dyscontrol³⁰ and is well supported for reducing suicide attempt and self-injurious behavior.¹⁰ DBT is a principle-driven approach designed to be flexibly applied to clients from different backgrounds and various target behaviors impeding clients’ goals.³⁰ As a behavior therapy, DBT prioritizes individualized assessment, case conceptualization, and intervention plans. Comprehensive DBT includes four core modes of treatment: skills training, individual psychotherapy, between-session skills coaching, and regular clinician consultation team meetings for clinicians. The skills training is typically delivered in a group format and covers four modules: Mindfulness, Emotion Regulation, Interpersonal Effectiveness, and Distress Tolerance. Although many randomized controlled studies have evaluated comprehensive DBT, a dismantling study found DBT skills group plus clinician consultation team with rigorous risk management and a manualized case management protocol was equivalent to comprehensive DBT in reducing suicide attempts and ideation among participants who had a history of multiple suicide attempts in the last 5 years, including one in the year prior, and a history of non-suicidal self-injury in the 8 weeks prior.³¹

Adaptations to full model DBT or DBT skills groups plus clinician consultation team with diverse patient populations have been undertaken by researchers and clinicians in the field,^17,32–34 as the samples in the original RCTs primarily included white, female patients of undocumented sexual orientation from above-average socioeconomic status in Western contexts.^15,35 Early RCTs evaluated DBT in a highly specific population of women in one area of the United States (US), who met criteria for borderline personality disorder, and were aged 18–45.^15,35 A systematic review of DBT RCTs showed that some sociodemographic variables (e.g., age, sex, race/ethnicity) were more commonly reported than others (e.g., sexual orientation, gender identity), and treatment outcomes were rarely reported by participant identity group, limiting the ability to discern whether treatment outcomes vary by identity group.³⁴

When considering whether to adopt or adapt DBT when working with patients with identity constructs that are different than the sample with whom the EBP was developed or in contexts different than that in which an EBP was initially evaluated, clinicians and administrators may experience what has been called the fidelity-adaptation dilemma: arguments in favor of adopting an EBP with fidelity to the way it was originally designed and tested versus arguments in favor of adapting the EBP to the current patient population or context.^17,36 Adopting EBPs as developed prevents inadvertently compromising the treatment components driving clinical impact, but could mean that the treatment provided is not effective for the current patient population or is not as feasible in the context.

DBT treatment developer, Dr. Marsha Linehan, argued for adoption before adaptation, to prevent clinicians from altering active components of DBT driving treatment effect,³⁰ and more contemporary work has demonstrated that treatment adherence is associated with more effective outcomes.³⁷ Additionally, a recent systematic review of randomized controlled trials of DBT in the US suggests that the literature base appears to more adequately represent ethnoracial and sexual orientation minority groups, and improvements are needed to more thoroughly understand representation and differences among patients with other minoritized identity constructs, including minorized gender identity groups.³⁴ A recent analysis of five US-based randomized controlled trials of DBT found that sexual orientation and race/ethnicity did not independently predict treatment outcomes, suggesting that adoption of DBT may be indicated when using DBT with ethnoracial and sexual orientation minority groups in the US.³⁸ These findings suggest that DBT may already be effective across patient populations of different identity constructs, however comparative effectiveness has only been assessed for some identity constructs. To date, published literature has historically included rather limited systematic assessment and documentation of patients and study participants identity constructs, thus limiting the field’s ability to accurately assess and compare treatment outcomes.

Conversely, adapting, or deliberately modifying an EBP with the goal of improving its effectiveness with a given group or in a given context, is another option.³⁹ A meta-analysis evaluating 76 mental health EBPs found a moderately strong benefit when the EBPs made cultural adaptations (d = .45).⁴⁰ Interventions conducted in patients’ native languages were twice as effective as those conducted in English, and interventions tailored to one specific cultural group were four times as effective as those provided to patients from a variety of cultural backgrounds.⁴⁰ These differences suggest substantial value in adapting EBPs to a specific patient population or context.^28,41,42 For DBT specifically, adaptations for various client characteristics, have been developed, including for specific identity characteristics (e.g., adolescent and university student populations), as well as specific clinical presentations (e.g., eating disorders, substance use disorders, and PTSD), and treatment settings (e.g., inpatient and forensic settings).⁴³ Nonetheless, developing and evaluating EBP adaptations can be time- and resource-intensive. Moreover, it would be infeasible to extensively research a culturally adapted version of EBPs for every possible combination of personal identity constructs. Dedicating the time and financial resources to studying adaptations of each EBP could inadvertently delay the delivery of EBPs to individuals who need them,³⁶ and reduce accessibility to EBPs for individuals seeking services at facilities with limited resources, which could exacerbate health inequities by delaying treatment for patients with elevated suicide risk.

Implementation science provides an approach to understand and reduce gaps in health inequities by studying the factors, processes, and strategies that influence the uptake, use, and sustainability of EBPs among vulnerable populations.^44,45 EBP adaptations can be considered an implementation strategy that addresses the fundamental interplay between the content of the EBP and the context in which it is implemented.⁴⁶ When adaptation is used, documenting the adaptation process and adapted intervention content is critical for the progression of EBP evaluation in different contexts. Thorough documentation allows other clinicians, researchers, and administrators to understand what content was delivered (including any adaptations) and to interpret results.^46–48 The Framework for Reporting Adaptations and Modifications-Enhanced (FRAME) is a framework for systematically documenting intervention adaptations content and process that can be used to characterize EBP implementation in different contexts.^2,11,33

Given that DBT is being widely applied outside of the population in which it was originally tested, it is important to understand more about how DBT is being applied when used with clients with identity constructs different than the patient population with which DBT was originally tested, and whether clinicians and researchers were adopting DBT as written or adapting DBT to fit the patient population. To investigate these questions, we conducted a scoping review, to identify and map available evidence and gaps in evidence,¹ and applied an implementation science framework, FRAME², to characterize adaptations to the content and context and the adaptation process. We intend this information to be useful in identifying what types of adaptations are most common among what populations, which could offer critical insights for clinicians, researchers, and policy-makers alike.

Methods

Search strategy and literature sources

Following the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) checklist for scoping reviews,¹ we searched PubMed, Embase and PsycINFO for articles published until 13 December 2024. For inclusion at the title/abstract level, we searched for articles mentioning variations of “dialectical behavior therapy.” Please see the supplementary table for a complete list of search terms (Appendix Table 1). The search strategy was informed by an expert library scientist (JGS) and the scoping review was organized through Covidence and registered (https://osf.io/65pyj; https://doi.org/10.17605/OSF.IO/65PYJ).

Inclusion criteria

Basic eligibility criteria were defined a priori (Table 1). The abstract stage inclusion criteria were that 1) the title/abstract described DBT, DBT components, or a DBT-based treatment, 2) the title/abstract described a patient population defined by identity constructs, operationalized using an expanded ADDRESSING model:⁴⁹ Age, Disability (developmental and acquired), Religion, Ethnicity and Race, Socioeconomic status, Sexual orientation, National language and origin, Gender; we expanded this by including veteran status, rurality, and indigenous populations), 3) the title/abstract was available in English. To test reviewer agreement of eligibility criteria, all reviewers reviewed the same random selection of (n=20) sources from the literature; level of agreement was 82%. The reviewers screened each title and abstract for relevancy, and copies of the full text were obtained and uploaded into Covidence for those that appeared to fit eligibility criteria.

Table 1.

Full-text review inclusion criteria using the SPIDER framework⁵⁵

Category	Description
Sample	Describes a patient population defined by cultural aspect(s) including (urban/rural, race, ethnicity, sex (not female), gender (not woman), gender identity, religion, national origin, language, income level, incarcerated, community treatment setting, geographic location, military culture, ability level, medical condition, study in another country, other),
Phenomenon of Interest	Describes DBT, at least one DBT component (individual, skills group, phone coaching, consultation team), or DBT-based treatment.
Design	Design includes case study. Pilot. Open trial. RCT. Protocol paper. QI eval. Single subject design. Implementation of DBT/DBT component. (NOT: commentary, review, introduction, etc.)
Evaluation	Evaluates outcomes (e.g. patient outcomes, intervention feasibility, acceptability, or uptake/adoption. Provider outcomes. Implementation experience/process.)
Research Type	Uses quantitative, qualitative, mixed-methods, or a protocol paper

For this analysis, we excluded articles focusing on children and adolescents (i.e., anyone below the age of 18) as there are empirically validated and replicated adaptations of DBT (DBT for Adolescents^50–52 and DBT for Children^53,54) justifying a different point of comparison.

After first stage of review, we refined the inclusion criteria for the full-text stage using SPIDER (Sample, Phenomenon of Interest, Design, Evaluation, Research type⁵⁵) outlined in Table 1. Again, all reviewers evaluated the same random sample of (n=10) articles and discussed any discrepancies, which were then resolved through consensus.

All abstracts and full-texts from the original search were reviewed by two authors: the study first author (FMA) and one other author from the review team (ES). Disagreements between the two reviewing authors were discussed and consensus was reached.

Data extraction, synthesis, and analysis

A standardized data extraction template was designed by the first author (FMA) and reviewed by the study team for relevance and appropriateness. The data extraction template included predefined data items for extraction: demographic data about the study (e.g. treatment setting, location, number of patients involved), the type of DBT offered (e.g. full model DBT, DBT Skills Group with consultation team, etc.), and the patient population served, as defined by the same expanded ADDRESSING model.⁴⁹ Finally, we applied an implementation science framework (FRAME²) to characterize types of adaptations to DBT, comparing each article’s protocol to the evidence-based DBT protocol³⁰ to assess whether adaptations were made to treatment content or context. The evidence-based DBT protocol includes weekly individual psychotherapy, a DBT skills group of two to two-and-a-half hours’ duration following a six-month curriculum and teaching skills in four distinct modules, clinicians attending a weekly clinician consultation team, and patient access to clinicians for between-sessions telephone coaching for skills generalization.¹ Content adaptations included whether the treatment content was reduced (e.g., treatment elements skipped, treatment sessions were condensed, or treatment duration was shortened), expanded (e.g., treatment elements added, sessions lengthened, or treatment duration extended), or modified (e.g., changes to written content, tailoring of language or examples). Context adaptations included adaptations to the context in which DBT was implemented, including changes to personnel, treatment setting, and treatment format (e.g. virtual session, group session, individual session). The data extraction form was pilot tested by two reviewers (FMA and SED) with 5 articles, and then all reviewers independently completed data extraction for a random 10% of the sample. Following, two reviewers independently extracted data; the first author (FMA) reviewed data extraction for all articles and reviewed any discrepancies. Having charted characteristics of the articles, a numerical and descriptive summary of the data extraction results was used to present findings.

Results

The search yielded a total of 3,198 references. After 119 duplicate articles were removed, we screened the remaining 3,079 abstracts and 746 references which met the inclusion criteria at the title/abstract level. After reviewing the full text articles with the more stringent review criteria, we identified 436 articles that were eligible for inclusion in the review. We then excluded the 163 articles that focused on children and adolescent populations, as DBT for adolescents^50–52 and children^53,54 have been evaluated in multiple studies. In total, we included 273 articles (Figure 1) describing the use of DBT or DBT elements in adult patient populations that were defined by being culturally different in some way from the largely white, female, United States sample in which DBT was originally developed (See Appendix Table 2 for a list of all manuscripts included in the scoping review).

Figure 1.

PRISMA Flow Diagram

Of the 273 articles that were identified, 73.6% (201/273) described outpatient treatment settings, 8.4% (23/273) described correctional or forensic settings, 7.3% (20/273) described inpatient or residential settings, and 10.6% (29/273) described other settings. The most common DBT implementation was DBT Skills Group plus clinician consultation team (61.7%, 129/273) followed by full model DBT (47.8%, 100/273). The 16.7% (35/273) articles included elements of DBT but were neither full model DBT nor DBT Skills Group (plus clinician consultation team). The remaining 4.3% (9/273), did not provide enough detail in the article to understand the DBT component. Nearly three quarters of the articles included were either from North America (40.3%, 110/273) or Europe (30.8%, 84/273). The remaining articles were based in Australia and Oceania (10.6%, 29/273), the Middle East (8.1%, 22/273), Asia (7.3%, 20/273), and South America (1.8%, 5/273; Table 2). There was one article from Africa (0.4%).

Table 2.

Country of DBT implementation

	N	%
North America	110	40.3%
United States	93	34.1%
Canada	15	5.5%
Mexico	2	0.7%
South America	5	1.8%
Brazil	3	1.1%
Columbia	1	0.4%
Argentina	1	0.4%
Europe	84	30.8%
UK	31	11.4%
Ireland	12	4.4%
Sweden	10	3.7%
The Netherlands	9	3.3%
Germany	6	2.2%
Italy	4	1.5%
Denmark	3	1.1%
France	4	1.5%
Spain	3	1.1%
Switzerland	1	0.4%
Norway	1	0.4%
Asia	20	7.3%
India	3	1.1%
Nepal	3	1.1%
China	5	1.8%
Taiwan	6	2.2%
Malaysia	2	0.7%
Singapore	1	0.4%
Middle East	22	8.1%
Iran	18	6.6%
Turkey	2	0.7%
Pakistan	1	0.4%
Saudi Arabia	1	0.4%
Australia and Oceania	29	10.6%
Australia	24	8.8%
New Zealand	5	1.8%
Africa	1	0.4%
South Africa	1	0.4%
Not specified	2	0.7%

We examined the content and context adaptations of the articles included in the scoping review (Table 3). Of the 273 articles, 83.5% (228/273) made adaptations to the content of the intervention protocol. Overall, 66.3% (181/273) of the articles reported reducing the time of the intervention in some capacity. We found that 33.0% (90/273) shortened session time and 53.8% (147/273) condensed the number of sessions (mean = 11.2, SD = 6.5, range = 1–23). Over half of the articles skipped content elements (54.2%, 148/273), which included skipping generally taught skills, homework review, and treatment elements like phone coaching for comprehensive DBT or clinician consultation team.

Table 3.

Types of content and context adaptations by reason stated for adaptation

	Articles reporting content adaptation		Patient identity constructs		Patient psychiatric diagnosis		Organization/setting requirements		Not specified
Content adaptations	N	% of all studies	n	%	n	%	n	%	n	%
Skipped content elements	148	54.2%	12	8.1%	8	5.4%	20	13.5%	108	73.0%
Shortened session time	90	33.0%	5	5.6%	6	6.7%	6	6.7%	73	81.1%
Condensed weeks of treatment	147	53.8%	13	8.8%	7	4.8%	18	12.2%	109	74.1%
Added elements	99	36.3%	20	20.2%	7	7.1%	28	28.3%	44	44.4%
Lengthened session time	12	4.4%	0	0.0%	0	0.0%	1	8.3%	11	91.7%
Extended number of weeks	31	11.4%	3	9.7%	8	25.8%	3	9.7%	17	54.8%
Changes to written content	63	23.1%	33	52.4%	15	23.8%	6	9.5%	9	14.3%
Tailored language and examples	87	31.9%	49	56.3%	16	18.4%	4	4.6%	18	20.7%
Context adaptations	N	%	n	%	n	%	n	%	n	%
Format (e.g. in person, virtual)	53	19.4%	8	15.1%	5	9.4%	13	24.5%	27	50.9%
Setting (e.g. where delivered)	49	17.9%	6	12.2%	1	2.0%	0	0.0%	42	85.7%
Personnel (e.g. facilitators)	19	70%	1	5.3%	2	10.5%	6	31.6%	10	52.6%

Note. One article could include more than one adaptation.

Conversely, 44.7% (122/273) of articles reported expanding some aspect of the protocol’s content. Just over one third of articles (36.3%, 99/273) reported adding treatment elements for specific patient populations, including adding culturally relevant metaphors and examples (e.g., Ramaiya introduced Nepali corollaries for DBT-specific terminology with adequate comprehensibility, acceptability, and colloquial equivalence to standard DBT terms and concepts⁵⁶), adding additional skills or skills modules (e.g., Pistorello et al. added a module on validation to a DBT adaptation for college students based on pilot work⁵⁷; Wagner et al. added skills to use in the methadone dosing line for ethnic minority individuals diagnosed with HIV and substance use⁵⁸), and incorporating aspects from other psychotherapies (e.g., Linton & Fruzzetti added exposure to feared movements and emotionally sensitive stimuli for patients with chronic pain.⁵⁹ Very few articles (4.4%, 12/273) lengthened session time or extended treatment duration (11.4%, 31/273; mean = 47.8 weeks, SD = 17.0, range 26–104).

Finally, 39.6% (108/273) of articles reported making changes to the protocol’s content. Approximately a quarter of the articles made changes to written content (23.1%, 63/273), for example, by simplifying handouts and worksheets to facilitate comprehension and incorporating images in text (e.g., simplifying acronyms and adding colorful visual aids to illustrate concepts in a prison setting;⁶⁰ modifying diary cards to use images to represent targets and skills for adults with developmental disabilities.⁶¹ Over one-third (31.9%, 87/273) tailored language and examples to fit specific populations or settings (e.g, male forensic population.⁶² These adaptations included use of translations⁶³ (e.g., use of a Swedish translation of the DBT Skills System;⁶⁴ use of Auslan to deliver DBT skills training;⁶⁵ note that some articles did not specify whether translation was used) and adding culturally tailored examples, like adding culturally tailored examples⁵⁶ (e.g., for Deaf individuals in Australia;⁶⁵ using military examples and values for deployed soldiers.⁶⁶

Across the 273 articles, a total of 677 distinct content adaptations were reported. For content adaptations, reasons for adaptions included patient identity constructs (19.9%, 135/677), patient psychiatric diagnosis (9.9%, 67/677) articles, and organization or setting requirements (12.7%, 86/677) articles. No reason was provided for 57.5% (389/677) of the adaptations.

We examined specifically the subset of articles (n=135) that identified patient identity as a reason for content adaptation. See Appendix Tables 3 and 4, which summarize adaptations in articles where patient identity was identified as a reason for adaptation, for both full model DBT and DBT Skills Group, respectively. When examined by identity construct, articles providing DBT adaptations to adults of a specific age range included some addition of skills for depressed older adults,⁶⁷ and reduction in number of distress tolerance skills provided to college students.⁵⁷ For adults with developmental and/or intellectual disabilities and autistic adults, multiple articles included modifications to written materials, including addition of models and images to illustrate skills,⁶⁸ or to facilitate participation by individuals who did not read.⁶⁹ For adults with chronic different abilities, tailoring language or examples was common: authors noted that some DBT skills were re-named or re-arranged to translate better into American Sign Language (e.g., the skill DEAR MAN became a culturally-affirming DEAF CAN.⁷⁰

We also examined context adaptations, or the adaptations to the settings in which the EBP was implemented (Table 3). Of the articles, 42.2% reported a context adaptation (115/273). Only 19.4% (53/273) articles indicated an adaptation to the protocol format, for example, offering DBT virtually, pre-recording sessions, or doing sessions via phone. We found that 17.9% (49/273) of the articles implemented the protocol in a setting that was different than the original environment (e.g. corrections, medical inpatient unit), and less than 10% of articles (19/273, 7.0%) had different personnel facilitating the protocol (e.g. peers, having inpatient milieu participate on the clinician consultation team).

Across all types of context adaptations, patient identity constructs accounted for 12.4% (15/121) of reasons for adaptation, while patient psychiatric diagnosis was cited as the reason for adaptation for 6.6% (8/121) articles, and organization or setting requirements was cited as the reason for adaptation for 15.7% (19/121) articles. No reason was provided for 65.3% (79/121) of the adaptations.

Discussion

This scoping review examined all published articles that described the utilization of a well-established and evidence-based suicide prevention intervention, DBT,³⁰ conducted in samples with identity constructs different than the sample in whom DBT was initially evaluated. We found that 83.5% of the articles included made content adaptations to DBT and 42.2% made adaptations to intervention context. Overall, these findings suggest that when DBT has been evaluated in patient populations different than the original evaluation samples, there are frequent adaptations to intervention content (e.g., eliminating skills, modifying handouts) and some adaptation to intervention context (e.g., changing the format of skills delivery, using non-healthcare-professionals as interventionists). Notably, when DBT content or context adaptations were made, only 42.5% and 36.5% of those articles, respectively, provided a rationale for the adaptation.

We found that the majority of studies evaluated (83.5%) adapted the content of DBT in some way, and over half (66.3%) reported condensing some form of the treatment, either omitting content, reducing time, or duration of treatment. This content adaptation is consistent with barriers to DBT frequently identified by clinicians working in community or public health settings, namely the time commitment to deliver DBT ^71,72 and the workload demand on clinicians.^73,74 When reasons for content adaptations were provided, patient identity constructs accounted for one-fifth of adaptations.

When specifically examining adaptations for patient identity constructs (see Appendix Tables 2 and 3), it is evident that these adaptations are often closely linked to the unique challenges faced by individuals with those identities. For instance, for racial and ethnic minorities, culturally specific tailoring of language and examples fostered increased relatability and engagement. A common theme across the adaptations for patient identity constructs was simplification (e.g., simplifying language and materials to increase comprehension and access), and reduction (e.g., shortening time in individual and group sessions or reduced numbers of sessions overall). This approach not only respects the differing capacities of diverse patient populations but also ensures that interventions remain effective and accessible. DBT is often perceived as complex and burdensome, which can hinder engagement and progress, especially among those with differing abilities. Thus, these adaptations make sense as they respond directly to the diverse needs of individuals who may find traditional approaches overwhelming or difficult. By implementing these thoughtful modifications, treatment can be made more relevant and supportive, ultimately leading to improved outcomes for all patients. Consequently, future research is needed to examine the treatment outcomes associated with these adaptations and further explore their impact on patient care.

Context adaptations occurred less frequently and included changing the setting (e.g., providing DBT in the burn unit of a hospital⁷⁵) or format (e.g., offering DBT virtually.⁷⁶) Adaptations included expanding who received DBT, such as including spouses in treatment,⁷⁷ or expanding who was providing DBT, such as having vocational therapists⁷⁸ or peers⁷⁹ facilitate groups or participate on the consultation team. When reasons for adaptation were given, patient characteristics were the more common reasons for adding or changing treatment components, while organizational or setting demands were the most commonly cited reasons for adaptations that reduced treatment components.

However, quite often reasons for adaptions were unspecified (57.5% of content adaptations and 65.3% of context adaptations), and had no detail about the process of modification. According to FRAME, thorough documentation outlining why the adaptations were made, who was involved in the adaptation process, and the impact of the adaptation are critical in order to interpret clinical outcomes and replicate clinically meaningful adaptations.² The high rate of unspecified adaptations suggests a meaningful gap in the literature. Without a clear rationale, other clinicians may struggle to replicate or utilize these adaptations effectively. To support this documentation, implementation of EBPs among minoritized populations may benefit from stakeholder-engaged processes like community-based participatory research^80,81,^82,83 and evidence-based quality improvement⁸⁴ to work together to develop and document adaptations. Some notable examples of this from the articles reviewed include Nyamathi et al., who used input from a community advisory board and semi-structured interviews with the target population to develop a DBT adaptation;^85,86 Huntjens et al. who collaborated with autistic adults to simplify the explanations of DBT skills to improve accessibility;⁸⁷ and Ramaiya et al. who synthesized feedback from clinicians and client interviews, as well as joint workshops, to make culturally-sensitive adaptations to DBT content and format to align with Nepali patient population.⁵⁶ Clinicians, researchers, and policy-makers all have a role to play in shifting the status quo, and in considering the systems or contexts that shape an individual patient’s access and response to suicide prevention care.⁴⁴

Research agenda.

Based on our findings, we present a research agenda for improving understanding of cultural adaptations for suicide prevention EBPs like DBT with specific activities for various stakeholders, including clinicians and clinical administrators, researchers, policymakers. Adaptation may be a critical component to improve patient outcomes or improve the feasibility of implementation,^43,46 and the rationale for adaptation should carefully considered and documented.

Clinicians and clinical administrators.

Clinicians and clinical administrators who find themselves wishing to adapt EBPs for a given patient’s identity constructs or a given cultural context are encouraged to systematically identify and document potential reasons for adaptations. Determining whether adaptations are needed may be based on input from individuals of that identity (e.g., therapist Elena Aranda, MA identified ways that DBT skills training excluded her community of Latinx immigrants, and sought to adapt the skills to resonate with Latinx parents⁸⁶), or based on evidence of lower attendance, program completion, or outcomes effectiveness for a particular cultural identity construct (e.g., while meta-analysis of clinical trials data showed DBT retention and outcome did not vary by racial/ethnic or sexual orientation minority status, an interaction was found between race/ethnicity and sexual orientation³⁴, suggesting an opportunity to assess barriers or acceptability for a particular subgroup). As noted by the authors, available data did not permit analysis of all client identities, suggesting the need for further resarch.³⁴ In the absence of compelling data suggesting a need for adaptation, given the relationship between intervention fidelity and outcomes in DBT clinical trials,³⁷ clinicians and clinical administrators might first attempt to implement DBT with fidelity and evaluate outcomes, balancing intervention rigor with cultural humility and awareness of potential power dynamics.^88,89

If it is determined that adaptation is needed, several sets of adaptation clinical guidelines are available, such as the ADAPT-ITT model⁴⁷ or three step process.⁴⁸ Clinicians are encouraged to evaluate any adaptation using routine outcome monitoring or standardized measures administered at more than one time point,⁹⁰ and to evaluate the adaptation’s acceptability to patients. Clinicians are further encouraged to share their findings via publications that include the adaptation’s effects on both implementation outcomes (e.g., acceptability, adoption, feasibility, etc.⁴⁶) and service and client outcomes (e.g., effectiveness, client satisfaction.) Finally, clinicians and clinical administrators are encouraged to build partnerships with community members and patients¹² to inform their adaptation decisions. For example, Ritschel et al. thoroughly assessed the feasibility, acceptability, and patient satisfaction of DBT skills training among autistic adults without intellectual disability which provided critical data to inform further modifications,⁹¹ and Edwards et al., evaluated the feasibility and acceptability of DBT protocol tailored for justice-involved Veterans.⁹² Building partnerships with community partners, and evaluating implementation efforts can increase the likelihood of care that meets the needs of patients and can be implemented in the setting.

Researchers.

Researchers can promote equity of access to effective suicide prevention by systematically studying the need for or use of EBP adaptations. Researchers can examine feasibility, acceptability, and clinical outcomes of adopted and adapted EBPs in different cultural groups so there is a more accurate characterization of what types of adaptations are necessary to support patient participation and to achieve clinical impact. When adaptations are made, researchers are encouraged to document the reasons for adaptations, types of adaptations made (content, context), and outcome of each adaptation, using a framework such as FRAME.²

Researchers must also take deliberate steps to include people from minoritized identity groups in psychotherapy research. The historical exclusion of people from minoritized identity groups in psychotherapy research has resulted in a lack of culturally sensitive treatment options for these groups.^93–95 Suicide prevention researchers are encouraged to include people from minoritized identity groups at every research stage, including identifying whether a suicide prevention intervention requires adaptation,⁹⁶ preparing to adapt a suicide prevention intervention,⁹⁷ and actually conducting an adapted intervention’s clinical trial. Building trust to include marginalized populations in research can be challenging given the history of exploitation and oppression experienced by them, including under the guise of research.⁹⁸ Researchers must work to build trust with participants through transparency, communication, and culturally responsive practices.^99,100 Prioritizing inclusion and cultural sensitivity in psychotherapy research is essential to address disparities and mental health outcomes for all individuals.

Policymakers.

Funders, professional organizations, and publishers are well positioned to set research agendas and establish reporting requirements that could normalize the description of intervention components, including adaptations. For example, the US National Institutes of Health have released a memorandum requiring inclusion of individuals from minoritized racial/ethnic groups in research.¹⁰¹ The Patient-Centered Outcomes Research Institute requires researchers to include people affected by the proposed research in the research procedures.¹⁰² The American Psychological Association Journal Article Reporting Standards for studies with experimental manipulations, including psychotherapy clinical trials, currently require only that studies report on how participants met inclusion or exclusion criteria.¹⁰³ We recommend that these standards be updated to include greater description of study samples (e.g., sample race, ethnicity, sexual orientation, etc.), and that journals and publishers require adherence to these more detailed reporting requirements. Requiring researchers to assess and describe additional patient identity characteristics can facilitate the evaluation of clinical outcomes by patient identity groups. Additionally, more specific descriptions of study samples will facilitate more rigorous systematic reviews or meta-analyses, aiding the field in further understanding whether a given intervention is effective among people of different identity groups.³⁴ Encouraging diversity in research and its reporting may be an effective policy-oriented approach to promote more representation in both areas.¹⁰⁴

Limitations of this research include the potential that we may have missed literature on implementations of DBT. This was mitigated by extremely broad search terms and rigorous review of articles. For practicality reasons and to maintain a high level of data quality, we chose not to include non-peer reviewed articles, and we recognize that important adaptations may exist outside of peer-reviewed sources. We also excluded articles that were not available in English, though this occurred in only 20 cases. Additionally, we may have overrepresented articles emanating from the same dataset, which could influence the conclusion drawn from this review. Finally, this dataset reflects only published articles; it is probable that clinical teams are implementing culturally adapted versions of DBT and not publishing data on their protocols.

This scoping review synthesizes existing cultural adaptations of DBT, sets an agenda for research to reduce inequities in suicide prevention research, and presents recommendations for clinicians and clinical administrators, researchers, and policymakers to reduce inequities at multiple system levels. While this scoping review focuses on DBT, findings are more broadly applicable to clinicians, administrators, and researchers interested in adapting any treatment for a cultural context, developing a new intervention, or implementing an evidence-based practice with cultural humility.