UBC Faculty Research and Publications

Cultural adaptations to augment health and mental health services: a systematic review Healey, Priscilla; Stager, Megan L; Woodmass, Kyler; Dettlaff, Alan J; Vergara, Andrew; Janke, Robert; Wells, Susan J Jan 5, 2017

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12913_2016_Article_1953.pdf [ 857.92kB ]
JSON: 52383-1.0361995.json
JSON-LD: 52383-1.0361995-ld.json
RDF/XML (Pretty): 52383-1.0361995-rdf.xml
RDF/JSON: 52383-1.0361995-rdf.json
Turtle: 52383-1.0361995-turtle.txt
N-Triples: 52383-1.0361995-rdf-ntriples.txt
Original Record: 52383-1.0361995-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessCultural adaptations to augment healthand mental health services: a systematicreviewPriscilla Healey1, Megan L. Stager1, Kyler Woodmass1, Alan J. Dettlaff2, Andrew Vergara1, Robert Janke3and Susan J. Wells1*AbstractBackground: Membership in diverse racial, ethnic, and cultural groups is often associated with inequitable healthand mental health outcomes for diverse populations. Yet, little is known about how cultural adaptations of standardservices affect health and mental health outcomes for service recipients. This systematic review identified extantthemes in the research regarding cultural adaptations across a broad range of health and mental health servicesand synthesized the most rigorous experimental research available to isolate and evaluate potential efficacy gainsof cultural adaptations to service delivery.Methods: MEDLINE, PsycINFO, CINAHL, EMBASE, and grey literature sources were searched for English-languagestudies published between January 1955 and January 2015. Cultural adaptations to any aspect of a service deliverywere considered. Outcomes of interest included changes in service provider behavior or changes in the behavioral,medical, or self-reported experience of recipients.Results: Thirty-one studies met the inclusion criteria. The most frequently tested adaptation occurred in preventiveservices and consisted of modifying the content of materials or services delivered. None of the included studiesfocused on making changes in the provider’s behavior. Many different populations were studied but most researchwas concerned with the experiences and outcomes of African Americans. Seventeen of the 31 retained studiesobserved at least one significant effect in favor of a culturally adapted service. However there were also findingsthat favored the control group or showed no difference. Researchers did not find consistent evidence supportingimplementation of any specific type of adaptation nor increased efficacy with any particular cultural group.Conclusions: Conceptual frameworks to classify cultural adaptations and their resultant health/mental health outcomeswere developed and applied in a variety of ways. This review synthesizes the most rigorous research in the fieldand identifies implications for policy, practice, and research, including individualization, cost considerations, and patientor client satisfaction, among others.Keywords: Cultural safety, Cultural appropriateness, Cultural competence, Health, Mental health, Racial disparities,Ethnicity* Correspondence: susan.wells@ubc.ca1Centre for the Study of Services to Children and Families, University ofBritish Columbia, ASC 453, 3187 University Way, Kelowna, BC V1V 1V7,CanadaFull list of author information is available at the end of the article© The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Healey et al. BMC Health Services Research  (2017) 17:8 DOI 10.1186/s12913-016-1953-xBackgroundMany people of diverse racial, cultural, and ethnic groupsin the United States and Canada experience differences inthe quality of health and mental health services associatedwith their group identity [1, 2]. Differences in treatmentoccur in access to care [3, 4], quality of provider-patientinteractions [5, 6], and engagement in care [7]. Forexample, in the U.S., a nationally-representative sample ofLatinos was less likely than non-Latin whites to seekhealth information or refer to it in conversation with theirphysicians [8]. In California and Chicago studies, AfricanAmericans were less likely than whites to feel that theyhad received high-quality care or had their medical needsmet [5, 9]. In Canada, Van Herk, Smith, and Andrewfound Aboriginal mothers in an urban center felt disre-spected by mainstream care providers [10]. Such differ-ences have been associated with negative health andmental health outcomes including, for example, underuseof services [11, 12] and failure to comply with physicianadvice [13].Health and mental health professionals are increasinglyconcerned with delivering more linguistically appropriate,culturally competent, and culturally safe services [14–16].Systematic reviews imply that culturally-adapted interven-tions can be successful but the nature and process of theadaptations are often lost during the reporting of results[14, 17, 18] and many reviews are narrowly focused.Examples of the latter include diabetes care [19, 20],asthma [21], HIV [22, 23], obesity [24], and psychotherapy[25, 26]. Critical assessment of the literature is compli-cated by the nature of the research conducted. Forexample, Lie, Lee-Rey, Gomez, Bereknyei, and Braddockreviewed the efficacy of cultural adaptations for serviceproviders but found no studies with equivalent controlgroups [27]. To demonstrate efficacy, cultural adaptationsmust be compared to the same intervention, minus theadaptations in question. Unless the experimental andcomparison groups are identical (but for the adaptation),it is impossible to determine whether any observed effectresulted from the adaptation itself, or some other aspectof the intervention. No systematic reviews have yet aggre-gated studies from the health and mental health literaturewhich isolate cultural adaptations from other aspects ofthe intervention and/or research design (see Additionalfile 1 for a list of reviews that are related to this study’s re-search question).Research questions were developed in consultation withan advisory panel of experts, knowledge users, and com-munity representatives. The researchers sought evidence ofcultural adaptations to any aspect of service delivery whichimpact: (a) the behavior of the service provider, (b) the re-cipient’s self-reported experience, or (c) outcomes for theservice recipient. Distinctions among the terms, race, eth-nicity, and culture are essential to understanding theseissues. Markus and Moya describe race as group member-ship assigned to people based on “perceived physical andbehavioral human characteristics” and used as a basis forthe conferral of “differential … power, and privilege” [28].Ethnicity is described as “ideas and practices” throughwhich people identify with a group based, for example, on“commonalities including … language, history, nation, …customs, …and/or ancestry”, and culture as “ideas andpractices attached to all the important social distinctions inour lives” [28]. Cultural competence, appropriateness, andsafety each have specific implications for improved servicedelivery. See Additional file 2 for more detailed definitions.For the purpose of this article, the term cultural adaptationis used to represent all modifications made to standardservice methods in order to make services more ac-ceptable, relevant, useful, and/or effective for diversepopulations. The terms patient, client, and consumer areused interchangeably depending on the context of thereferences cited.MethodsThe project methodology is consistent with theCochrane Collaboration guidelines and supplementalsources [29–34]. Due to the vast quantity of informa-tion available on this topic and a burgeoning interestin the field, it was necessary to adopt stringentcriteria with regard to inclusion in this review. Thescope of this review was progressively narrowed toinclude only randomized controlled trials (RCTs) andquasi-experimental research with parallel cohorts.Only studies which isolated the cultural adaptationfrom control interventions were considered. This lim-ited the kinds of adaptations which could be includedin the study. For example, changes in organizationalpolicy were of interest to the project, but no studiesmet the strict criteria for comparison groups. Thesearch included adaptations in any aspect of servicedelivery. Outcomes of interest included: (1) healthoutcomes of the recipient, (2) behavioral outcomes ofthe recipient, (3) self-reported outcomes of the recipi-ent, including service satisfaction, and (4) behavioraloutcomes of the service provider. A complete sum-mary of reviewers’ inclusion and exclusion criteria isprovided in Table 1.Search strategyThe final strategy was iterative; the search results guidedrefinement of the search terms. The original databasesearch of MEDLINE, PsycINFO, CINAHL, and EMBASEwas performed in August and September 2011. A com-bination of keywords and database-specific subjectheadings were used to search the following concepts andsynonyms: “cultural competency” or “culturally tailored”or “racial disparities” or intercultural or “communicationHealey et al. BMC Health Services Research  (2017) 17:8 Page 2 of 26barriers” related to race or ethnicity. See Additional file 3for a complete list of search terms, dates the searches wereconducted, as well as full database search histories. Theresults were updated in 2012 and again in 2015. Inaddition to the database search, key reports and literaturereviews were identified and hand-searched. These docu-ments were selected based on the degree to which theyfocused on the project’s research questions. A forward cit-ation title search was conducted using Google Scholar andWeb of Knowledge; items found were screened by titleand abstract. Backward citation searches involved title-screening the reference lists of key reports and literaturereviews to identify any relevant literature cited within thisstudy’s search findings. Twenty percent of these resultswere double-screened by a second reviewer to ensureconsistency. Authors of retained reports were alsocontacted to identify research that may have been missedin the search. A grey literature search was conducted toidentify unpublished or omitted material (see Additionalfile 4). Inclusion of a database was guided by relevance tothe study focus and relevance of returns from initialsearches.ScreeningTitles and abstracts were used to eliminate documentsthat were deemed irrelevant or outside the scope of theresearch questions. Reviewers then evaluated the full-text of documents and applied the inclusion criteria toidentify the strongest research in the health and mentalhealth literature. At the onset of the study, four pairs ofreviewers conducted title and abstract screening. Eachpair independently double-screened a sample of theTable 1 Reviewers’ inclusion and exclusion criteriaInclusion Exclusion1. English language from any country 1. Study findings not in English2. Published 1950 or after 2. Prior to 1950 or abstracts not available3. RCTs and quasi-experimental designs with parallel cohorts of controlor comparison groups3. Studies which were not RCTs or quasi-experimental designs, e.g.,observational studies, moderator analyses4. Services included health or mental health 4. Other human services5. Described adaptation(s) intended to make services more responsiveto or effective for diverse racial and ethnic populations; adaptationsmay target:• individual service provider OR• service system5. Did not contain a description of the specific activities undertaken toimprove cultural competence, appropriateness, or safety, and/orthe study did not justify the inclusion of an adaptation withculturally-grounded rationale and/or existing research6. Explicitly tested the effectiveness of the cultural adaptation separatefrom any other health or mental health intervention studied. Thismust result in intervention and control groups that differ only onthe included cultural adaptation6. Studies in which the cultural component and the health or mentalhealth intervention were not evaluated separately from the otherservice provided. Also excluded studies that tested a generally usedintervention to study its impact on a cultural, minority, ethnic, ordisadvantaged population without adapting it to specifically suit theneeds of the target population7. Focus of study was on provision of a service 7. Studies that: only tested the translation of psychometricinstruments, questionnaires, and diagnostic tools, focused onengaging visible minorities in research, or involved service deliveryat some unspecified future time, such as genetic registries8. Studies pertained to people and organizations in the mainstreamculture making adjustments to include and serve those who aresubject to inequity in service delivery or service outcomes8. Service recipients did not represent a group subject to disparities inservice delivery or outcomes, or target subjects’ data wereconfounded with those of another group that is not subject tohealth disparities and/or is not the target of the cultural adaptationunder study9. Reported outcomes that included:• change in service provider behavior OR• change in self-reported experience or outcomes of servicerecipient OR• change in observed outcomes for service recipient9. Did not contain evidence of having measured outcomes of theadaptation to enhance cultural competence, appropriateness, orsafety with specific reference to:• change in service provider behavior OR• change in self-reported experience or outcomes of servicerecipient OR• change in observed outcomes for service recipient10. Outcomes and data were provided and analyzed in a way thatallowed an evaluation of the direct results of the culturaladaptation10. Outcomes and conclusions were not substantiated in the reportwith sufficient data11. There were no flaws in the study methodology and/or deliverydeemed likely to threaten the internal validity and interpretabilityof the study’s results11. The research design, intervention delivery, or assessment ofoutcomes involved a confounding variable which threatens theinternal validity of results, e.g., clinically meaningful differencesbetween groups at baseline, lack of experimental control,inadequate statistical reporting, etc.Healey et al. BMC Health Services Research  (2017) 17:8 Page 3 of 26same documents and established inter-rater reliabilityusing Cohen’s Kappa. Reviewers discussed any disagree-ments to resolve them, and consulted the Principal Inves-tigator (PI) where an agreement could not be reached.Consistency among reviewers was maintained through useof the same decision rules, constant communication,meetings, and oversight from the PI. Once reviewersachieved a Kappa of .90, each reviewer screened itemsindependently. Reliability was periodically checked bydouble-rating a random 10% sample of the screened arti-cles for each set of at least 100 reviewed documents, thenproducing a new Kappa. When the Kappa slipped below.90, reviewers returned to double-screening each docu-ment until an agreement of .90 or greater was achieved. Inthe 2015 update, sufficient resources were available todouble-screen all database documents.In full-text screening, each pair of raters double-screeneduntil they achieved 100% agreement on a random sampleof documents, at which point they worked independently,double-screening a random 20% to assess reliability. Be-cause inter-rater agreement remained at or near 100%,double-screening was reduced to a random 10% of every100 documents. In addition, any article included at thisstage was cross-screened by the second reviewer to con-firm the validity of the inclusion decision. In the event of adisagreement, the PI was consulted to settle the discrep-ancy. During the 2015 update, resources were again avail-able for double-screening of all documents. Authors werecontacted for additional information when necessary.Data extractionThe data extraction form was based on the work ofHasnain et al. [18], Littell et al. [32], and the CochraneCollaborative GRADE approach [35]. The form included,but was not limited to: details of the study population,baseline characteristics, details of the setting, studymethodology, study outcomes, and bias/quality informa-tion. Inter-rater reliability was assessed by comparingthe content of extraction forms until 100% agreementwas attained. In 2012, a random third of articles wascompared for consistency. This was reduced to a ran-dom 10% because reviewers maintained consistent inter-rater reliability.The 2015 data extraction began with a trial period inwhich three reports were extracted and evaluated inconsultation with the PI to ensure raters’ accuracy andcomprehension of the process. The remaining reportswere double-extracted independently by each member ofa single pair of reviewers. Reports were discussed in-depth with the PI when: 1) the reviewers disagreed withone another, or 2) the reviewers’ decisions changed as aresult of discussion and consideration of the researchdesign. Some reports presented issues which necessi-tated further specification of the inclusion/exclusioncriteria. For example, Breitkopf et al. studied culturally-framed messages for African American, Latina, andWhite women, but collapsed data across these threeethnicities, confounding the populations of interest withWhite data [36]. This resulted in a need to specify thatdata for the population of interest must be evaluable inisolation from the general population. The final inclu-sion process was iterative. Articles from all three stagesof the project were revisited and discussed until thecurrent pool of items was identified.ResultsSearch resultsThe electronic databases returned the following results:2011 (n = 5141), 2012 update (n = 529), and 2015 update(n = 1954) after de-duplication, for a total number ofelectronic database documents of 7624. The flow ofdocuments retained at each step in the review process ischarted in Fig. 1.For the reports reviewed during title and abstractscreening, the most common reason for eliminationwas that they did not test a culturally-adapted inter-vention to improve health or mental health outcomes forservice recipients. The remaining eliminations were studiesof translated instruments or studies lacking comparison/control groups with parallel cohorts. In full-text screening,the vast majority were omitted because the interventions inthe control or comparison groups were not identical, savefor the cultural adaptation in question. One study wasomitted at the point of data extraction because the resultsreported in narrative could not be substantiated by the dataprovided in the publication. A handful of reports wereomitted because they could not be obtained from theauthors. Requests made to key authors produced no add-itional studies or findings.Study characteristicsThe search identified 38 retained documents published inthe United States from 1993 to 2015. Three documentsrepresented completed pilots (two of which were subse-quently folded into the analysis of the final report), and fiverepresented interim findings that could be combined withlater final results, yielding 31 studies. The research repre-sented a total of 9,831 participants. The majority of studiestargeted a specific racial, ethnic, or cultural group: AfricanAmerican (n = 17), Latino (n = 7), Asian American (n = 4),and Gay/Bisexual (n = 1). Two studies targeted ethnically-diverse populations [37, 38]. All 31 studies tested aculturally-adapted intervention for the service recipients.Topics were diverse and included enhancement of dietthrough increased fruit and vegetable intake, cessation ofsmoking, uptake of preventive services, and improvementin psychological functioning. In total, 17 of the 31 studiesHealey et al. BMC Health Services Research  (2017) 17:8 Page 4 of 26observed significant effects in favor of a culturally adaptedintervention.Risk of biasEach study was assessed by two reviewers for bias inseven areas outlined by the GRADE Criteria: selec-tion, allocation concealment, performance, detection,attrition, reporting, and other. Authors were con-tacted for further design details where possible. Ofthe retained studies, only one reported enough infor-mation to fully assess bias in each domain. Thirteenstudies were rated high risk for bias in one or moredomains, and 30 studies had an unknown risk of biasin one or more area. Table 2 provides a summary ofreviewers’ bias ratings.Analysis of cultural adaptationsThe included studies used a range of cultural adaptationsdesigned to promote cultural competence, appropriate-ness, or safety. A variety of frameworks for classificationof the results were reviewed, but none were directlyapplicable to the assortment of adaptations found. Forexample, Chowdhary et al. [39] elaborated on the targetsof adaptation in psychotherapy as well as on how todevelop adaptations. They described adaptations of lan-guage, therapist adaptations such as matching or training,use of metaphors to increase cultural relevance, adaptingthe content of therapy, adapting communication to in-corporate cultural constructs such as somatization, client-derived treatment goals, changes in therapeutic methods,and addressing clients’ socio-economic contexts to reduceFig. 1 Breakdown of results during the search process. *Documents have been de-duplicated where possible given database and software restrictionsHealey et al. BMC Health Services Research  (2017) 17:8 Page 5 of 26barriers to treatment. While helpful in characterizing thepsychological treatments reviewed, the framework doesnot incorporate the breadth of health and mental healthadaptations found in this review. Castro, Barrera, and col-leagues’ [40, 41] reviews report similar findings. In 2004,they outlined the sources of mismatch between theprovider and consumer (group characteristics, programdelivery staff, and administrative/community factors) thatcould be used as targets of intervention [41]. In 2006 theypresented a Heuristic Framework [42] that may be used toguide the development of adaptations. In 2012, they iden-tified several frameworks for characterizing adaptations,most notably, Bernal et al.’s eight dimensions of therapythat could be culturally adapted [41]. The elements in-cluded aspects of treatment such as goals, methods, andcontext of treatment, as well as characteristics of the clientthat could be incorporated such as language and familiarcultural expressions [43].Although these frameworks include many aspects ofhealth and mental health services, the conceptualizationTable 2 Reviewers’ judgments regarding sources of biasStudy SelectionbiasAllocationconcealmentPerformancebiasDetectionbiasAttritionbiasReportingbiasOtherbiasArd et al. 2008 [53] X X ? ? √ ? ?Burrow-Sanchez et al. 2015 [77] √ ? X ? √ ? XBurrow-Sanchez & Wrona, 2012 [76] √ ? X ? √ ? ?Chiang & Sun, 2009 [79] X ? ? ? √ ? ?Fitzgibbon et al. 2005 [50] X X √ √ √ √ ?Gondolf, 2008 [47] √ ? ? ? √ √ √Halbert et al. 2010 [68] ? ? ? ? √ √ √Havranek et al. 2012 [58] √ √ √ √ ? ? ?Holt et al. 2009 [71] ? ? ? ? √ ? ?Holt et al. 2012a [66]; 2012b [67] √ ? √ √ √ ? XHuey & Pan, 2006 [64]; Pan et al. 2011 [65] ? ? ? √ √ √ ?Hwang et al. 2015 [80] √ ? ? √ √ ? XJandorf et al. 2013a [51]; 2013b [52] √ ? ? ? √ ? √Johnson et al. 2005 [37] ? ? ? ? √ √ √Kalichman et al. 1993 [46] ? ? ? ? ? ? ?Kreuter et al. 2004 [60]; 2005 [62] √ √ ? √ √ √ √La Roche et al. 2006 [38] √ ? X ? √ ? XLee et al. 2013 [54] √ ? ? ? ? ? ?McCabe et al. 2009 [55]; 2012 [56] √ √ X √ √ ? √Mohan et al. 2014 [59] √ √ X √ √ ? √Newton & Perri, 2004 [45] ? ? ? ? √ √ √Nollen et al. 2007 [73] √ √ √ √ √ √ √Orleans et al. 1998 [74] X ? ? X √ X √Resnicow et al. 2009 [63] ? ? √ ? √ √ √Sanders Thompson et al., 2010 [70] √ ? ? ? ? ? ?Shoptaw et al. 2003 [78] √ ? ? ? √ √ √Skaer et al. 1996 [57] X ? ? √ √ ? ?Unger et al. 2013 [72] √ ? ? √ √ ? XWang et al. 2012a [48]; 2012b [49] √ ? ? ? √ ? √Webb, 2009 [75] ? ? √ √ √ √ √Webb et al. 2010 [69] ? ? ? ? √ ? ?Total Low Risk:Total Unclear Risk:Total High Risk:√ = 17? = 9X = 5√ = 5? = 24X = 2√ = 6? = 20X = 5√ = 18? = 12X = 1√ = 27? = 4X = 0√ = 11? = 19X = 1√ = 14? = 12X = 5Healey et al. BMC Health Services Research  (2017) 17:8 Page 6 of 26is somewhat abstract and the categories have not yetbeen developed to the extent that they could be used asa more specific classification scheme. For instance, alocal healer could be classified as therapist matching,inclusion of cultural knowledge, or modification of treat-ment methods. The work done to date is helpful, butnot yet sufficient for classification of wide-ranging adap-tive cultural arrangements and activities.To establish a framework better suited for the pur-poses of this review, the two senior researchers in thestudy conducted a content analysis of the research in-cluded to identify three primary domains of adaptation:1) community outreach and involvement, 2) changes inthe structure and process of service delivery, and 3)adaptation of content. Once the framework had beenidentified, it was expanded to encompass other adapta-tions encountered during the literature review that weredeemed to be hypothetically testable in comparative re-search designs (Table 3). The adaptations described inthe framework focus on changes that affect, in someway, the interface with the client. For example, it may bepossible to make beneficial changes in organizationalmanagement or legislation, but it is the resultingchanges to the nature of service delivery that are catego-rized, rather than the way in which the adaptations aregenerated.This framework is designed to serve the full range ofhealth and mental health domains and organize adapta-tions in such a way as to be as exhaustive and mutuallyexclusive as possible. Culturally sensitive changes to atreatment may have more than one adaptation, but theyall can be categorized using the framework in Table 3,allowing practitioners and researchers to begin to describeand compare adaptions using a common approach. Thefollowing description and the accompanying analysis ofthe adaptations found are representative of this classifica-tion system. Resulting categorizations and descriptions ofimplemented adaptations are provided in Additional file 5.For future analyses of adaptation, the authors have alsoprovided an example of how the classification could beadapted to reflect the depth and/or detail of adaptationaddressed by Resnicow et al. [44] in Additional file 6.Overview of adaptationsWithin the retained studies, the most popular method ofadapting an intervention was to modify the content ofmaterials or dialogue to include racial, ethnic, or culturalfacts, values, imagery, or other cultural components.The next most common was to change the manner inwhich a service was delivered, including increases intime and attention paid to recipients, cultural matchingof providers to clients, and provision of additional re-sources. Consultation with community was implementedto inform, and in concert with, changes to structure,process, and/or content of service delivery. The majorityof adaptations were tested in one or more of theretained studies.There were no discernible differences in adaptationselection or impact when it came to specific health prob-lems; adaptations from each category identified in theframework were fairly evenly featured for each uniquehealth concern, and cultural adaptation was not particu-larly effective for any one health concern. There werelikewise no discernible differences in adaptationTable 3 Conceptual framework for cultural adaptations1. Community outreach and involvementa. Community needs assessment (e.g., outlining the issue from theirperspective)b. Involvement in development of the adaptationc. Participation in the implementation/management/delivery ofservices2. Changes in structure and process of service deliverya. Change in geography/location (e.g., location of center, home vs.office visits, etc.)b. Change made to the physical space (e.g., pictures, room orbuilding design, etc.)c. Change in mechanism of service delivery (e.g., face-to-face,electronic, mailed, etc.)d. Changes to service provider/presenter (e.g., selection and training)i. Language matching to clientii. Race, gender, or cultural matching to cliente. Change in manner of service delivery (e.g., interaction style,proximity to client, active or passive speech, intonation, rapportbuilding, self-presentation, group composition, etc.)f. Provision of supplemental services, resources, or supporti. Supplemental providers (e.g., traditional healer, patient navigators)ii. Funds for a specific service or resourceiii. Supplemental services (e.g., child care, transportation, paid leavefrom work)iv. Translated materials (e.g., documents, signs, etc.)iv. Other3. Adaptation of contenta. Level of personal specificityi. Individualizedii. Targeted to subgroupb. Inclusion of cultural contenti. Graphicsii. Cultural allusions (affect-free content with which the recipientmay personally identify)iii. Culturally-relevant factual informationiv. Targets or references negative-valence beliefs, values, orexperiences (e.g., fatalism, stigmatization)v. Targets or references neutral or positive-valence beliefs, values, orexperiences (e.g., familial involvement, time-orientation)Healey et al. BMC Health Services Research  (2017) 17:8 Page 7 of 26selection or impact when it came to targeting specificcultural groups.Community outreach and involvementFourteen studies included consultation with members of,and experts from, the community of interest. All fourteenstudies engaged community members in the developmentof the adaptation’s content via focus groups, but commu-nity outreach and involvement was not the primary adap-tive goal of any study. Focus groups typically elicitedcultural themes, values, and preferences from participants.Changes in structure and process of service deliveryTwenty-one studies featured changes in the structureand/or process of service delivery. Only Newton andPerri implemented and isolated changes to the physicalintervention space between conditions, with the adaptedprogram being held at a site located in the AfricanAmerican community [45].Changes to service provider Six studies tested providermatching. These studies matched the providers’ or pre-senters’ race [45–49], gender [46], or language [48, 49]to those of the recipients in an effort to facilitate changesin behavior, such as screening uptake. Beyond racialmatching, three studies selected providers/presenters whowere a cultural or community match to service recipientsin an effort to enhance their identification with theirprovider [47, 50–52].Change in manner of service delivery Sixteen studieschanged the manner in which an intervention was deliv-ered or portrayed. Studies employed a wide variety oftechniques, and typically included more than one adap-tation. For example, Jandorf and colleagues had peernavigators relate their personal experiences with colon-oscopy and model effective coping skills [51, 52]. Alter-natively, Ard et al. organized group interventions suchthat all participants were of the same race [53], and Leeet al. had therapists spend time building rapport withclients and emphasized collaboration [54].Provision of supplemental services, resources, orsupport Five studies provided recipients with supple-mentary resources to facilitate uptake or retention in aservice. Resources included provision of translated andsimplified materials [55, 56], a voucher to be redeemedfor a free mammogram in an effort to increase breastcancer screening [57], provision of access to child-careand transportation, and accommodation for recipients’work schedules [54]. Havranek et al. provided a supple-mental values-affirmation exercise to clients prior tomeeting with their general practitioner to enhance self-efficacy [58], and Mohan, Riley, Schmotzer, Boyington,and Kripalani provided clients with a simplified andillustrated medication management tool to facilitateunderstanding of pharmaceutical regimens [59].Adaptation of contentTwenty-six studies adapted the content of the interven-tion to reflect the norms, values, and culture of thetarget population. Studies referred to culturally-salientstatistics and historical events as motivating factors forchange. Adaptations often included positive culturalvalues, beliefs, and norms to facilitate, enhance, or mo-tivate change during the intervention. Negative culturalvalues and experiences were frequently referred to as motiv-ating factors (e.g., African Americans’ history of oppression),or were targeted by the adaptation as a barrier to change(e.g., belief in fatalism). Changes were implemented in manymediums of program content. Four studies included indi-vidualized content for each recipient. Three tailored inter-vention content and provider delivery based on recipients’level of acculturation [60–65]. The study undertaken byMcCabe and colleagues personalized parent–child inter-action therapy after a family-based needs assessment;changes addressed cultural beliefs about the causes of be-havioral problems, familial roles, discipline, etc. [55, 56].Packages of adaptationsOf 31 studies, only five tested a lone adaptation. Twenty-seven1 studies tested multiple cultural adaptations in con-cert, such that it is impossible to isolate the effect of anyone change to the service. It is possible that any observedeffect may have resulted from: 1) a single adaptation fromamongst the many, 2) the sum of adaptations made to-gether, or 3) an interaction among adaptations made. Twostudies employed multiple comparison groups, allowingfor the isolation of numerous adapted components as wellas the complete package of adaptations [46, 47]. The num-ber of adapted elements ranged from one to five, based onan adaptation of Hasnain and colleagues’ framework foridentifying the number of adaptations present in a service[18]. Additional file 7 includes a Forest Plot that illustratesthe effects of number of adaptations tested.Analysis of outcomesThe retained studies tested health outcomes from five ofthe six domains identified as potentially meeting this re-view’s criteria regarding health outcomes (see Table 4 forthe categorization chart).Table 4 Conceptual framework for health outcomes1. Service Provider Behavioral Outcomes2. Service Uptake: Completion/Participation3. Service Recipient Awareness, Beliefs, Knowledge, and Attitudes4. Service Recipient Behavioral Outcomes5. Indicators of Health/Mental Health StatusHealey et al. BMC Health Services Research  (2017) 17:8 Page 8 of 26No retained studies assessed service provider behav-ioral outcomes. Havranek et al. assessed client-providercommunication from the client and provider perspec-tives, but clients were the intended target of the inter-vention [58]. Overall, 19 studies assessed the awareness,knowledge, and/or attitudes of recipients. Twenty-threeof the studies assessed service uptake in some form. Six-teen studies assessed service recipient behavioral out-comes. Ten studies measured indicators of health ormental health status. Seventeen studies observed healthor mental health outcomes which significantly favoredthe culturally adapted group, but there was no clear pat-tern as to which outcomes were affected, or which adap-tations were implemented. However, three studiesobserved results which significantly favored the standardgroup.Interventions implemented within the retained studieswere focused either on targeting ongoing health con-cerns (treatment services) or on prevention of futurehealth problems (preventive services). The interventionsimplemented by Mohan et al. and Havranek et al. werethe only retained studies to adapt a physiological treat-ment service [58, 59]. In contrast, preventive servicestargeted physical conditions, such as cancer, HIV, smok-ing, and asthma. Decisions to obtain medical screeningwere categorized as preventive service uptake outcomes.No retained studies assessed the medical decisions madebased on screening results.In addition to testing the health outcomes of interestto this review, retained studies also tested some out-comes not directly related to health experience, such asrecipient shared material with a friend. These outcomesdid not fit the target definition for health outcomes, sointervention impact on such outcomes is not discussedherein. Studies are tallied and categorized based on theirprimary research goal. See Table 5 for a detailed list ofresults, accompanying data, and comparison groups foreach study. Additional file 5 categorizes outcomes bytype of adaptation and outcome.Uptake of preventive servicesSeven studies attempted to increase uptake of preventivescreening services. Jandorf et al. and Holt et al. sought toincrease colorectal cancer (CRC) screening in AfricanAmericans [51, 52, 66, 67]. No statistically significant im-provements were observed in preventive screening uptake.Kalichman et al. sought to increase HIV screening ratesand awareness among African American women [46].Adaptations of their promotional video resulted in an in-crease in HIV screening and more favorable responses topresenters. Halbert et al. also targeted African Americanwomen in an effort to increase genetic screening followinga counseling session about breast cancer genes, but foundno significant gains for the culturally adapted group [68].Kreuter et al., Skaer et al., and Wang et al. eachattempted to increase mammography rates in AfricanAmericans, Latinas with low-incomes, and ChineseAmericans, respectively [48, 49, 57, 60–62]. Kreuter et al.found tailoring on both cultural and behavioral vari-ables showed the highest increase in self-reportedmammography, but this was not significant relative tobehaviorally-tailored materials alone at the 18 monthfollow-up [60–62]. Wang et al. found that a culturallyadapted promotional video was not more effectivein increasing mammography compared to control,although mammography uptake was moderated byacculturation status [48, 49]. Skaer et al. observed thelargest effect of the included studies: Latina womenwith low-incomes receiving vouchers for free mam-mography were over 47 times more likely to receive amammogram than controls [57].Awareness, knowledge, and attitudesSix studies aimed to modify recipients’ awareness,knowledge, and/or attitudes as a primary goal. Webbet al. attempted to increase smoking-related diseaseawareness and perceptions of risk in African Ameri-cans who smoke [69]. Culturally adapted materialsdid not lead to greater knowledge than controls, butdid increase risk perceptions and result in strongerintentions to quit. Sanders Thompson, Kalesan, Wells,Williams, and Caito, as well as Holt et al. targetedcancer screening beliefs among African Americans[70, 71]. Sanders Thompson et al. did not observe astatistically significant difference between adapted andcontrol groups [70]. Holt et al. found that recipientsin the adapted group reported higher usage of mate-rials, but observed mixed results with regard to self-efficacy [71]. Mohan et al., La Roche et al., andUnger, Cabassa, Molina, Contreras, and Baron eachattempted to increase knowledge of medications [59],asthma [38], and depression [72], respectively, andwere successful in at least one measure related toknowledge.Smoking behaviorsFour studies attempted to modify smoking behaviors astheir primary research goal. Nollen et al., Orleans et al.,and Webb et al. targeted materials to African Americansand observed mixed results [73–75]. Nollen et al. foundthat despite significantly greater usage of adapted mate-rials, there were no statistically significant differences insmoking outcomes [73]. Webb et al. observed greaterreadiness-to-quit and more quit attempts in the standardgroup and no difference in abstinence rates betweengroups [75]. Adapted materials, however, were ratedmore favorably in several areas. In contrast, Orleans etal. found a significant increase in quitting behaviorsHealey et al. BMC Health Services Research  (2017) 17:8 Page 9 of 26Table5Characteristicsofincludedstudiesandreportedbetween-groupsoutcomesStudySampleInterventionOutcomeArdetal.2008[53]AfricanAmericansN=377CulturallyAdapted(CA):Raciallymatchedparticipantsingroupweight-lossprogram.Standard(STD):Multiculturalparticipantgroup.○aNosignificantdifferenceinattendance(p=.09),changeinweight(p=.97),fruit/vegetableintake(p=.60),fiberintake(p=.94)orfatintakeatfollow-up(p=.46).○Nosignificantdifferenceinpercentageofrecipientsgetting>180min.physicalactivityperweekatfollow-up(p=.18).Burrow-Sanchezetal.2015[77]Latinos(Adolescents)N=70bCulturallyAdapted(CA):Culturallytailoredcognitivebehavioraltherapy(CBT).Standard(STD):StandardCBT.○Nosignificantdifferenceinreductionofpast-90-daydruguse(p=.66).Burrow-Sanchez&Wrona,2012[76]Latinos(Adolescents)N=35CulturallyAdapted(CA):Culturallytailoredcognitivebehavioraltherapy(CBT).Standard(STD):StandardCBT.○Nosignificantdifferenceinreductionofpast-90-daydruguseorprogramretention†.●ParentsintheCAconditionweremoresatisfiedwiththeprogram(p=.02).○Nosignificantdifferenceinadolescentsatisfaction,(p=.09).Chiang&Sun,2009[79]AsianAmericans(Chinese)N=128CulturallyAdapted(CA):8-weekculturallytailoredwalkingprogram.Standard(STD):Non-tailoredprogram.○Nosignificantdifferenceinpost-testbloodpressureorwalkingendurance†.Fitzgibbonetal.2005[50]AfricanAmericans(Obese/over-weight,women)N=59CulturallyAdapted(CA):Faith-based12-weekweight-lossprogram.Standard(STD):Weight-lossinterventionwithnoactivefaithcomponent.○Nosignificantdifferenceinprogramretention(>75%attendance)†.○Nosignificantdifferenceinenergyexpenditureat12weeks(p=.08).○Nosignificantdifferenceindietaryfatconsumptionat12weeks(p=.91).○Nosignificantdifferencein12-weekweightchange:Kg(p=.34),%(p=.41).○NosignificantdifferenceinBMIcchangeat12weeks(p=.37,d=0.27).○Nosignificantdifferenceineithervigorousphysicalactivity(p=.36)ormoderatephysicalactivity(p=.06)at12weeks.Gondolf,2008[47]AfricanAmericans(Men)N=372CulturallyAdapted(CA1):16-weekracially-matchedgroupcounselingprogramwithstandardcurriculumfordomestic-violenceoffenders.CulturallyAdapted(CA2):Racially-matchedcounsellorandculturally-targetedprogramcurriculum.Standard(STD):MulticulturalgroupwithCaucasiancounsellorsandstandardcurriculum.○Programcompletionwascomparableacrossgroups†.Halbertetal.2010[68]AfricanAmericans(Women)N=176CulturallyAdapted(CA):Culturallytailoredgeneticcounseling.Standard(STD):Standardgeneticcounseling.○Nosignificantdifferenceinriskperceptionatfollow-up(LRT=0.07,p=.79).○Nosignificantdifferenceincounselingcompletion(p=.70).○Geneticscreeninguptakewascomparablebetweengroups†.Healey et al. BMC Health Services Research  (2017) 17:8 Page 10 of 26Table5Characteristicsofincludedstudiesandreportedbetween-groupsoutcomes(Continued)Havraneketal.2012[58]AfricanAmericansN=99CulturallyAdapted(CA):Avalues-affirmationexercisetoreducestereotype-threatandboostself-efficacyofcli-entsduringrace-discordantclient-providercommunications.Standard(STD):Neutralcomparisonexercise.●CAgroupprovidedandrequestedsignificantlymoreinformationaboutmedicalcondition(p=.03),butnottherapeuticregimen(p=.56),lifestyle(p=.42),orservices(p=.70).○Nosignificantdifferenceintrustinprovider(p=.55)orpatientvisitsatisfaction(p=.32).Holtetal.2009[71]AfricanAmericans(Men)N=49CulturallyAdapted(CA):Spiritually-based“Sunday-school”prostatecancereducationsession.Standard(STD):Non-spiritualprostatecancereducationalsession.○Groupswerecomparableinratingtheacceptability/appropriatenessoftheinterventionandinfindingithelpfulformakinginformeddecisions†.●CAgroupreadsignificantlymoreofthematerials(p<.01).○Differenceinchangeinself-efficacywasnotsignificantbetweengroupsforscreening,decisionmakingregardingprostatespecificantigen,ordecisionmakingregardingdigitalrectalexamination†.○Groupschangedcomparablyinscreeningbeliefs,knowledge(prostatecancer,screeningcontroversy,relationshipbetweenscreeningandmortality),andbarrierstoscreeningdecisions†.Holtetal.,2012a[66],2012b[67]AfricanAmericansN=285CulturallyAdapted(CA):Spiritually-themedcolorectalcancereducationsession.Standard(STD):Non-spiritualcolorectalcancereducationsession.○NosignificantdifferenceinCRCdknowledgeatfollow-up(p=.65[2012a]).●STDgroupself-reportedsignificantlymoreFOBTewithinprevious12months(p=.03[2012b]).○Nosignificantdifferenceinfollow-upself-reportoflifetimeFOBT(p=.55),flexiblesigmoidoscopy(p=.52),colonoscopy(p=.55),orbariumenemas(p=.32[2012b]).○Nosignificantdifferenceinfollow-upperceivedCRCscreeningbenefits(p=.16),FOBTbenefits(p=.20),FOBTbarriers(p=.33),colonoscopybenefits(p=.80),orcolonoscopybarriers(p=.54[2012b]).Huey&Pan,2006[64];Panetal.2011[65]AsianAmericansN=30CulturallyAdapted(CA):Culturallytailoredsingle-sessionexposuretreatmentforphobias.Standard(STD):Standardone-sessionexposuretreatmentforphobias.○Nosignificantdifferencesinavoidance/anxiety,catastrophicthinking,generalfear,orDSM-IVTRfphobicsymptomsatfollow-up(2011)†.○CAgrouphadsignificantlylowersubjectivedistressratingsatoneweek,butnotat6months(2011)† .●Nosignificantdifferenceofclinicianratingoffearatoneweek,buttheCAgroupwasratedashavingsignificantlylowerfearresponseatsixmonths(2011)†.Hwangetal.2015[80]AsianAmericans(Chinese)N=50CulturallyAdapted(CA):CulturallytailoredCBTfordepression.Standard(STD):StandardCBT.○Nosignificantdifferenceinprogramretention†.○Nosignificantdifferenceinseverityofdepressionbysession12†.●Log-lineargrowthmodelrevealedCAgroupobservedsignificantlygreaterdecreaseindepressionscoresfrombaselinetosession12despitebaselinedifferences(p=.047).Healey et al. BMC Health Services Research  (2017) 17:8 Page 11 of 26Table5Characteristicsofincludedstudiesandreportedbetween-groupsoutcomes(Continued)Jandorfetal.,2013a[51],2013b[52]AfricanAmericansN=304gCulturallyAdapted(CA):Peer-ledpatientnavigationforAfricanAmericansreferredforcolonoscopy.Standard(STD):Physician-ledpatientnavigation.○Groupsweresimilarinratesofcolonoscopyscreeningatfollow-up(2013b)†.○Nosignificantdifferenceintrustinprovideratfollow-up(p=.56[2013a]).○Nosignificantdifferenceinperceivedmessageandsourcecredibility(p=.97[2013a]).○Groupswerecomparableinsatisfaction(p=.07[2013a])†.Johnsonetal.2005[37]Multicultural(Children)N=3157CulturallyAdapted(CA):8-session,50min.multiculturalanti-smokingcurriculum.Standard(STD):Standardanti-smokingcurriculum.○Nosignificantdifferencesinpast-monthsmokingorlifetimeever-having-smokedby8thgrade†.Kalichmanetal.1993[46]AfricanAmericans(Women)N=106CulturallyAdapted(CA1):CulturallytailoredcontentandbehaviorofpresentersinanAIDS/HIVeducationalvideo.CulturallyAdapted(CA2):RacialandgendermatchingofpresentertoaudienceinanHIV/AIDSeducationalvideo.Standard(STD):StandardHIV/AIDSeducationalvideowithmixed-gender/racepresenters.●CA1obtainedsignificantlymoreHIVtests(p<.01).●CA1andCA2togetherweresignificantlymorelikelytorequestcondomsatpost-test,(p<.001).○NosignificantdifferencesinHIV/AIDSinformationseekingatpost-test,condompurchasing,orattemptingtousemorecondoms†.○NosignificantdifferencesinHIV/AIDSknowledgeandattitudesatpost-test†.●CA1presentersweresignificantlymoreperceivedasexpressingconcern(p<0.01)thantheothergroupscombined.○Nosignificantdifferencesinratingsofpresenterexpertise†.Kreuteretal.2003[61],2004[60],2005[62]AfricanAmericans(Women)N=599hCulturallyAdapted(CA):Culturally&behaviorallytailoredcancereducationmagazinestoincreasemammography/fruit&vegetableintake.Standard(STD):Magazinestailoredonbehavioralcontentalone.○CAgroupwasnotsignificantlymorelikelytohaveobtainedamammogramby18monthsthantheSTDgroup(2005)†.○Groupsincreasedcomparablyinmedianfruit/vegetableservings(2005)†.○Nosignificantdifferenceinhavingreceivedandreadmaterialsat6months(2004)† .LaRocheetal.2006[38]AfricanAmericans,LatinosN=22iCulturallyAdapted(CA):Allocentricfamilyasthma-managementprogram.Standard(STD):Standardfamilyasthma-managementprogram.●CAgroupreducedthenumberofemergencydepartmentvisitsinthe12monthfollowupperiodby50%† .●CAgroupwassignificantlygreaterinparentalasthmaknowledgeat12months(p<.05).○Nosignificantdifferencesinparentalskills,childskills,orchildknowledgeat12months†.Leeetal.2013[54]LatinosN=53jCulturallyAdapted(CA):Culturallytailoredsingle-sessionmotivationalinterviewingtoreducealcohol-inducedbehavioralproblems.Standard(STD):Standardmotivationalinterviewing.○Nosignificantdifferenceintreatmentengagement†.○Nosignificantdifferenceinprogramsatisfaction†.○Groupsdecreasedcomparablyfrombaselineinpast-monthheavydrinking.TheCAgroupobservedanon-significant,butgreatereffect(p=.08,η2=0.10).●CAgrouphadgreaterdecreasesinalcohol-inducedproblembehaviorscoresontheDrInCkImplusivitysubscale,(p=.009,η2=0.14).TheotherDrInCsubscalesdidnotsignificantlydifferbetweengroups† .Healey et al. BMC Health Services Research  (2017) 17:8 Page 12 of 26Table5Characteristicsofincludedstudiesandreportedbetween-groupsoutcomes(Continued)McCabe&Yeh,2009[55];McCabeetal.2012[56]Latinos(MexicanAmerican)N=58CulturallyAdapted(CA):CulturallytailoredParent–childInteractionTherapy(PCIT)forfamilieswithchildrenwhohavebehaviorproblems.Standard(STD):StandardPCIT.○CAgroupshowedgreaterimprovementforallhealthoutcomes,butdifferenceswereallnon-significantbetweengroups:ECBIlIntensitySubscale(p=.77,d=.09),ECBIProblemSubscale(p=.34,d=.28),CBCLm(p=.10,d=.36),ECInODDosymptoms(p=.13,d=.07),ECICDpsymptoms(p=.12,d=.26),ECIADHDqsymptoms(p=.18,d=.08),PSIr(p=.53,d=0.09),andPLOCs(p=.10,d=.35[2012])†.○CAgroupshowedsignificantlygreaterimprovementontheCBCLInternalizingsubscale(p=.049),butthiswasnolongersignificantafteraBonferronicorrection(2012).○Groupswerecomparableintreatmentsatisfactionanddropout(2009)†.○Nosignificantdifferencesinparent–childpositive/negativeinteractionstyles(doanddon’tskills[2009])†.○Nosignificantdifferenceinpositiveparentingbehaviorscoresatpost-test(2009)†.Mohanetal.2014[59]LatinosN=200CulturallyAdapted(CA):TAUtplusasupplementarysimplifiedandillustratedmedicationmanagementtool.Standard(STD):TAU.●CAgrouphadsignificantlygreaterknowledgeandunderstandingofmedicationregimensatfollow-up(p<.001).○Nosignificantdifferenceinself-reportedmedicationadherenceatfollow-up†.Newton&Perri,2004[45]AfricanAmericansN=42uCulturallyAdapted(CA):10-sessionculturallytailoredgroup-exerciseprogramandwrittenmaterials.Standard(STD):Standardprogramandmaterials.○Nosignificantdifferenceincompletionofprescribedexercise(p=.39).●CAgroupratedgroupleadersasshowingsignificantlymoreappreciation(p=.03).○Nosignificantdifferenceinself-reportedphysicalactivityatpost-test†.○Groupsincreasedcomparablyinmaximumoxygencapacity†.○Therewasnosignificantdifferenceinself-efficacyatpost-test†.Nollenetal.2007[73]AfricanAmericansN=500CulturallyAdapted(CA):Culturally-tailoredanti-smokingvideoandprintguide.Standard(STD):Standardvideoandprintguide.●CAgroupusedtheguidesignificantlymore(p=.03).○Nosignificantdifferenceinvideousage(p=.37),perceivedbenefitsoftheguideinattemptingtoquit(p=.07),orofthevideoinattemptingtoquit(p=.32).○NosignificantdifferenceinprogressionalongtheStagesofChangecontinuumintermsofreadinesstoquitby6months†.○Nosignificantdifferencein7-dayabstinenceat6months(p=.27).○Nosignificantdifferenceinchangefrombaselineinthenumberofcigarettessmokedperdayat6months(p=.61)orself-reportednicotinepatchuse(p=.75).Healey et al. BMC Health Services Research  (2017) 17:8 Page 13 of 26Table5Characteristicsofincludedstudiesandreportedbetween-groupsoutcomes(Continued)Orleansetal.1998[74]AfricanAmericansN=1422CulturallyAdapted(CA):Culturallytargetedstop-smokingcounselingsessionandwrittenmaterials.Standard(STD):Standardcounselingandmaterials.○Nosignificantdifferenceinself-reportedreadingofmaterialorproportionofrecipientswhofoundtheguidehelpfulat6months†.●STDgroupratedtheguideassignificantlymoresuitableforotherfamilymembersat6months(p=.01).●CAgroupsignificantlyreducedthenumberofcigarettessmoked(p=.002),wasmorelikelytosetaquitdate(p=.001),andwasmorelikelytoswitchtoalower-nicotinebrandofcigarettesby6months(p=.001).●CAgroupmadesignificantlymorequitattempts(p=.007),andusedmorepre-quittingstrategies(p=.05)by6months.○Nosignificantdifferenceinself-reportedweek-longabstinence,progressionalongtheStagesofChangecontinuum,orinsmokingabstinenceby6months†.●CAgrouphadahigherquitrate(p=.034),andweremoreadvancedalongtheStagesofChangecontinuum(p=.035)at12months.○Nosignificantdifferenceinnicotinepatchorgumuse,ormediannumberofquitattemptsat12months†.Resnicowetal.2009[63]AfricanAmericansN=560CulturallyAdapted(CA):Culturallytailoredfruit&vegetablepromotionalmaterials.Standard(STD):Standardmaterials.○Nosignificantdifferenceinmeandailyfruit/vegetableintakeby3months(p=.13).○Groupswerecomparableinself-reportedreadingofmost/allnewslettersat3months†.SandersThompsonetal.2010[70]AfricanAmericansN=771CulturallyAdapted(CA):Culturallytailoredcolorectalcancerrisk-reductionmaterials.Standard(STD):Standardmaterials.○Nosignificantdifferenceinaffect,engagement,easeofunderstanding,cognitiveprocessing,orintenttoscreenat22weeks† .Shoptawetal.2005[78]Gay/Bisexuals(Men)N=80vCulturallyAdapted(CA):Culturallytailoredcognitivebehavioraltherapy.Standard(STD):Standardcognitivebehavioraltherapy.○Nosignificantdifferenceinprogramretention†.○CAgroupsignificantlyreducedself-reportedunsafereceptiveanalintercourseduringfirst4weeksoftreatment.Differencesbetweenconditionswerenon-significantat12months†.●CAgrouphadsignificantlyhigherTreatmentEffectivenessScoresformethabstinenceatendoftreatment(p<.05).○Nosignificantdifferenceinpercentofnegativeurinesamplesorreporteddaysofpast-monthmethuse†.Healey et al. BMC Health Services Research  (2017) 17:8 Page 14 of 26Table5Characteristicsofincludedstudiesandreportedbetween-groupsoutcomes(Continued)Skaeretal.1996[57]Latinas(Low-income,Women)N=80CulturallyAdapted(CA):Provisionofvouchertoredeemforonefreemammogram.Standard(STD):Novoucherprovided.●CAgroupwas47timesmorelikelytoobtainamammogramatfollow-up,usinglogisticregressionanalysis(p=.0001).Ungeretal.2013[72]LatinosN=139CulturallyAdapted(CA):Illustratedfotonovelatoincreasedepressionknowledgeandreducestigma.Standard(STD):Standarddepressionpamphlet.●CAgroupwassignificantlylowerinantidepressantstigma(p<.05)andmentalhealthcarestigma(p=<.05)atpost-testw.●CAgroupwassignificantlyhigherindepressionknowledgeatpost-test(p<.05).○Nosignificantdifferencesinself-efficacytoidentifydepressionorwillingnesstoseekhelp(p>.05)atpost-test.Wangetal.2012a[48];2012b[49]AsianAmericans(Chinese)N=442xCulturallyAdapted(CA):Culturallytailored,mailedmammographypromotionalvideo.Standard(STD):Standardmailedmammographypromotionalvideo.○Groupswerecomparableinincreasesinmammographyfrombaseline(2012b)†.○Nosignificantdifferencesinintenttoobtainmammogramatpost-test(2012a)†.○Nosignificantdifferenceinculturalviewsofhealthcare,knowledge,perceivedrisk,perceivedbenefits,orperceivedbarriersatpost-test(2012a)†.Webb,2009[75]AfricanAmericansN=261CulturallyAdapted(CA):Culturallytargetedwrittenmaterialsforsmokingcessation.Standard(STD):Standardmaterials.●CAmaterialwassignificantlymorelikelytocaptureattention,provideencouragement,andhelpinquitting†.●STDmaterialwasseenassignificantlymorecredible(p<.05).○Nosignificantdifferenceinbookletutilization(p=.09).●CAgroupwassignificantlymoresatisfiedwithcontent(p=.03).●STDgroupwas1.97(95%CI[1.09,3.55])timesmorelikelytomakeaquitattemptbyfollow-up(p=.03).●STDgroupscoredsignificantlyhigherontheContemplationLaddermeasureatfollow-up(p=.01).○Nosignificantdifferenceinpointprevalentabstinenceorsmokingreduction†.Healey et al. BMC Health Services Research  (2017) 17:8 Page 15 of 26Table5Characteristicsofincludedstudiesandreportedbetween-groupsoutcomes(Continued)Webbetal.2010[69]AfricanAmericansN=243CulturallyAdapted(CA):Culturallytargetedwrittenmaterialsforsmokingcessationandexercise.Standard(STD):Standardsmokingandexercisematerials.●CAgroupwassignificantlyhigherinperceptionofpersonalrisksofsmokingatpost-test(p=.02,η2=0.02).●CAgroupwassignificantlyhigherinperceptionofculturally-specificrisksofsmokingatpost-test(p=.04,η2=0.02).●CAgroupwassignificantlyhigherinintentionstoquitatpost-test(p=.04,η2=0.02).○NosignificantdifferenceinContemplationLadderscoresatpost-test†.○Nosignificantdifferenceinsmokingknowledgeatpost-test†.a ○Denotesanon-significantoutcome.●Denotesasignificantoutcomeasdefinedbytheoriginalauthors’criteria.†Denotesanoutcomewhichisreportedintheoriginaldocument,butforwhichprobabilityvalueswerenotprovidedbN’srepresentthesamplesizeanalyzedinthefinalreport.Notethatinterimreportsmayhaveanalyzeddatarepresentingadifferentsamplesizefromthatofthefinalreport,e.g.,duetoattritionc BodyMassIndexdColorectalCancer(CRC)e FecalOccultBloodTest(FOBT)f DiagnosticandStatisticalManualofMentalDisorders,4thEditionTextRevision(DSM-IVTR)gNote:ThisnumberrepresentsthesamplesizeoftheCAandSTDgroupsonly,omittingtheTAUsample,whichwasnotofcentralinteresttothisreviewhNote:ThisnumberrepresentsthesamplesizeoftheCAandSTDgroupsonly,omittingthe“culturallyrelevanttailoring”group,asneitherBRTnotCRT+BRTcanserveasanadequatecontroltotestthisgroupi Thisnumberrepresentsthenumberoffamiliesparticipating,notthenumberofindividualsj Thisnumberrepresentsthenumberofparticipantsthatweresaidtoberandomizedk Drinkers’InventoryofConsequences(DrInC)l EybergChildBehaviorInventory(ECBI)mChildBehaviorChecklist(CBCL)nEarlyChildhoodInventory(ECI)oOppositionalDefiantDisorder(ODD)pConductDisorder(CD)qAttentionDeficitHyperactivityDisorder(ADHD)r ParentingStressIndex(PSI)s ParentalLocusofControl(PLOC)t TreatmentAsUsual(TAU)uThisnumberrepresentsthesamplesizeoftheCAandSTDgroupsonly,omittingtheTAUsample,whichwasnotofcentralinteresttothisreviewv Note:ThisnumberrepresentsthesamplesizeoftheCAandSTDgroupsonly,omittingthecontingencymanagement(CM)andCBT+CMgroups,becauseneithergroupcouldserveasanadequatecontrolfortheCAgroupwOutcomesreportedarefrompost-test,asthefollow-updatawasconfoundedwhenparticipantsineithergroupexchangedreadingmaterialsafterthepost-testmeasurex Note:ThisnumberrepresentsthesamplesizeoftheCAandSTDgroupsonly,omittingthefact-sheetsample,becausethisgroupcannotserveasanadequatecontrolfortheCAgroupHealey et al. BMC Health Services Research  (2017) 17:8 Page 16 of 26among the adapted group, as well as higher rates ofsmoking abstinence at 12 months [74]. Johnson et al.targeted materials to multicultural schoolchildrenfor the purpose of smoking prevention [37]. Theiradapted education program reduced the odds ofsmoking by eighth grade when compared to a non-intervention control, whereas the standard anti-smokingcurriculum did not. Johnson et al. also observed that theirmulticultural curriculum was significantly more effectiveonly among Latino students in Latino-dominant schools.Similarly, their standard program was most effective onlyamong Asian-American students within Asian-American/multicultural schools [37].Substance use behaviorsFour studies focused on other substance-use. Burrow-Sanchez and Wrona, and Burrow-Sanchez, Minami,and Hops found no significant group differences indrinking outcomes or treatment satisfaction amongLatino adolescents, however, treatment outcome wasmoderated by recipients’ ethnic identity and measuresof familism [76, 77]. Lee et al. observed greaterreductions in alcohol-induced problem behavior for Lati-nos in the culturally-adapted motivational-interviewinggroup [54]. Shoptaw et al. had mixed results with meth-amphetamine use and HIV-related sexual risk behaviorsamong gay and bisexual men. The adapted cognitivebehavioral therapy (CBT) group achieved higher averageTreatment Effectiveness Scores, but also had higher meth-use than the standard group [78].Other health behaviorsSix studies focused on non-substance use health be-haviors. Four addressed physical activity in ChineseAmericans [79] and African Americans [45, 50, 53].Activity outcomes did not significantly differ betweenexperimental and control groups in the three studiesthat assessed activity [45, 50, 79]. Participants inNewton and Perri’s cultural group rated their groupleaders as more appreciative than those in the stand-ard group [45]. Three studies targeted fruit and vege-table intake among African Americans: Ard et al.[53], Kreuter et al. [60–62], and Resnicow et al. [63],with Kreuter et al. specifically targeting women.Kreuter et al. reported that cultural and behavioraltailoring of materials resulted in greater increases inrecipients’ daily fruit and vegetable intake, but notsignificantly more so than behavioral tailoring alone[60–62].Mental healthThree studies focused on mental health outcomes.McCabe and colleagues modified Parent Child InteractionTherapy (PCIT) for Mexican American children withexternalizing behavioral problems [55, 56], Pan, Huey, andcolleagues tailored exposure therapy for Asian Americanswith phobias [64, 65], and Hwang et al. targeted depres-sive symptoms among Asian Americans [80]. Pan et al.found a significantly greater reduction in phobic outcomesin the adapted group at time two compared to the stand-ard exposure treatment, but both groups were comparableat the long term follow-up [64, 65]. Moderator analysesindicate reductions in catastrophic thinking and generalfear were greatest for Asian Americans who were lessacculturated to American society.Uptake of treatment servicesTwo studies focused on treatment participation. Gondolfattempted to increase participation in domestic violencecounseling. Gondolf found that neither the all-African-American standard counseling nor culturally-focusedcounseling resulted in increased treatment completioncompared to a multicultural, standard-curriculum coun-seling group. However, for men with high racial identifi-cation, the completion rate was between 63% and 65%when data from both adapted conditions were pooled,compared to a 40% completion for men with high racialidentification in the multicultural condition [47]. Havraneket al. also targeted African Americans in an attempt toboost self-efficacy and reduce stereotype threat via a values-affirmation exercise. They found that patients receiving theexercise requested and provided more information abouttheir medical condition, and that patient-provider commu-nication was characterized as significantly more positive[58].Excluded studiesThere were numerous adaptations observed in theliterature which were not tested under the stringentdesign requirements set forth by this review’s inclu-sion and exclusion criteria. A number of adaptationspresent in the literature were difficult to isolate in aresearch design with direct, equivalent group compari-son. The stringent criteria also proscribed inclusion ofstudy designs using retrospective or moderator ana-lyses as their sole method of evaluation, as these didnot meet the criteria of an intervention being imple-mented with the clear intention of targeting specificcultural groups. Studies identified as having non-cultural confounds that could be thought to plausiblyaffect the health outcomes above and beyond theimpacts of cultural adaptation limit the ability toeffectively analyze the internal validity of culturaladaptations, and are therefore not included within thedescriptive portion of this review. However, implica-tions of findings discussed below were compiled uponreview of the body of literature encountered through-out the process of this review as a whole.Healey et al. BMC Health Services Research  (2017) 17:8 Page 17 of 26DiscussionThe included studies differed in number of adaptations,type of adaptations, and the extent of modification, butall sought to improve the experiences and health out-comes of underserved populations through modificationof health and mental health services. This review isunique in that it goes beyond a synthesis of culturallytailored interventions and seeks to identify and analyzeonly studies in which the research design and data analysissupport some confidence regarding the validity of thestudy conclusions. By limiting the studies to those withdirect comparisons between culturally adapted interven-tions and the same interventions in their un-adaptedform, the adaptation is truly tested for effectiveness. Bylimiting the outcomes to those which are experienced bythe service recipient, one is not left to guess whetherincreased sensitivity of the provider actually results in im-proved experience for the recipient. While other researchdesigns and by extension other reviews may haveaddressed similar questions, they are constrained by theinability to separate the effects of the adaptation from theeffects of the medical or mental health service.Of course, the very thing that helps isolate the effect ofan adaptation requires a highly structured interventionwhich will not always be reflective of the patients, con-texts, and processes found in other settings. The sectionon limitations details these issues. However the ability tomore fully determine the effectiveness of the adaptationand the existence of other reviews of less rigorous ap-proaches (see Additional file 1) weighed heavily in favor ofthis approach. The breadth of this review also led to theidentification of core cultural adaptations that occuracross health and mental health services and an examin-ation of their efficacy in various settings.Casting a wide net resulted in the development of twoframeworks with which practitioners, policy makers, andresearchers may conceptualize adaptations and out-comes in future work (Tables 3 and 4). The frameworksdescribe the extent to which cultural modification ispossible and will foster more consistent measurement ofhealth experience throughout the identified categories ofhealth and mental health service adaptations and out-comes. Although conceptual frameworks of culturaladaptations are already present within the research lit-erature [18, 41], the framework within Table 3 is distinctin that it goes beyond summaries of stages in the adapta-tion process and instead offers a concrete list of all con-ceivable adaptations at different levels of serviceimplementation. Future research can therefore be in-formed beyond how to adapt to what can be adapted,and which adaptations can thereafter be evaluated in iso-lation. See Castro and colleagues’ review of issues andchallenges in the design of culturally adapted interven-tions for more information and guidance [41]. Lastly,Additional file 6 depicts one way in which the adaptationframework can be applied, illustrating the “level of en-gagement” of recipients with the cultural adaptation.Included studiesAs has been previously indicated, there appears to be nouniversally accepted standard for creating or testing cul-tural adaptations. The majority of studies tested pack-ages of adaptations that included multiple components.Though some studies implemented adaptations in a spe-cific category (e.g., adapting the content), only five testedsingular adaptations (e.g., adding graphics into the con-tent). As a result, researchers in such studies could notassign resulting effects to any specific adaptation, butrather evaluated the package of adaptations as a whole.Of the 31 retained studies, 9 were identified by the re-search team as having one or more foreseeable, practicalimpacts on the health experience of their service recipi-ents. Two of these interventions were categorized asprovision of supplemental resources, services, or support,and each effectively addressed a separate barrier to serviceuptake. Skaer et al. noted that previous research had indi-cated that financial concerns were identified as the greatestbarrier to mammography screening uptake for Latinas[57]. Their approach was to provide vouchers for freemammography screening to low-income Latina women,which was effective in achieving significantly greater pre-ventive screening. Havranek et al. similarly address barriersto treatment by targeting communication as a barrier to ef-fective treatment uptake for African Americans [58]. Clientsreceived a values-affirmation exercise prior to meetingwith their general practitioner that guided them to identifytheir own values and strengths in an effort to reduce per-ceived stereotype threat and improve the quality ofprovider-client interactions. Those who participated in theexercise were more likely to request information regardingtheir medical condition which in turn enhanced theprovider-patient interaction.Five other studies implemented effective packages of in-terventions. Orleans et al. implemented a package inter-vention consisting of both culturally relevant materialsand culturally sensitive counselling, which was effective atincreasing participants’ smoking quit-rate at 12 months[74]. Lee et al. adapted a motivational interviewing sessionthat resulted in a significant decrease in scores on a scalethat measures serious legal and physical harms related toalcohol use (DRInC Impulse scale). The authors alsofound a difference in the reduction of number of heavydrinking days per month that approached significance(p = .082 = .10, f = .33) [54]. A study by Kalichman et al.found that cultural adaptations to their AIDS video re-sulted in more participants requesting condoms and talk-ing about AIDS with their friends [46]. In addition, onlythe participants in the group with both adapted contentHealey et al. BMC Health Services Research  (2017) 17:8 Page 18 of 26and ethnically matched providers went for HIV testing inthe 2 weeks after the intervention. The MultifamilyAsthma Group Treatment (MFAGT) implemented byLaRoche et al. was likewise effective; in their study,MFAGT was significantly better at increasing parentalasthma knowledge and reducing visits to the emergencydepartment [38]. Lastly, Hwang et al. observed a signifi-cant interaction of treatment by time with regard todecreased depression through the use of their culturallyadapted CBT program [80].Two other studies of note did not find significant re-sults regarding health indicators or health behavior out-comes, but did observe differences that could be seen asmeaningful to the health experience of service recipients.For instance, although Unger et al. found no statisticaldifferences in willingness to seek help between groups,participants who read the culturally adapted fotonovelareported significantly less stigma regarding antidepres-sants and mental health care, in addition to increaseddepression knowledge [72]. Similarly, McCabe et al.found that their culturally adapted Parent Child Inter-action Therapy outperformed the standard treatment onall outcomes, with between-groups Cohen d’s rangingfrom .09 to .36, though no differences reached statisticalsignificance [56] (Table 5).Excluded studiesThe exclusion of studies that used single-group, wait-list,or other non-equivalent designs resulted in the exclusionof some popular and interesting methods of culturalmodification from this review. For example, many studiestested the introduction of patient navigators in the serviceuptake process, but none controlled for both attentionaland informational confounding factors inherent in patientnavigation. The same was true of educational interven-tions aimed at improving providers’ cultural competence.The inclusion criteria of this review required that a pro-vider intervention: (a) isolate the cultural adaptation fromother interventions and (b) measure providers’ behavioraloutcomes or the recipients’ outcomes. None of the identi-fied studies met both of these criteria, as many studies didnot utilize adequate controls or assessed only changes inproviders’ knowledge, attitudes, and awareness.Another common cause of omission from this reviewwas a lack of cultural justification for the adaptation, thatis, if there was no reason given to indicate that the chosenadaptation would benefit the target population any morethan it would benefit the general population. For example,standard psychotherapies were omitted if they were notindicated to be especially efficacious in, or adapted for, agiven population. Only studies that supplied cultural justi-fication for supplemental services, such as Skaer et al.,were retained. Skaer and colleagues justified their financialintervention with survey research previously undertakenin the community of interest regarding barriers to mam-mography screening [57]. This process conformed withthe intent of this review to identify adaptations that weretruly cultural in nature and that were implemented tomeet the needs of specific cultural groups.Adaptations and efficacy for different health problemsThe main method of addressing the health concernstargeted within the retained studies appeared to bereformatting intervention content to contain culture-specific information and themes or convey this content ina manner congruent with the target culture. Despite widevariation in health targets, this basic principle appeared toremain. It was only once studies were classified into pre-ventive- versus treatment-based interventions that anypattern could be discerned. For example, if the goal wascancer prevention, the focus often rested on increasingthe knowledge of target populations and a consequentchange in behavior such as obtaining screening services.Two studies specifically targeted medical treatment:Havranek implemented changes to affect the medicaltreatment dialogue [58] and Mohan implemented changesto medical instructions [59]. No retained studies assessedpreventive strategies targeting mental health outcomes,other than Unger et al.’s use of a fotonovela to increasewillingness to seek help for depression [72].Adaptations and efficacy for different populationsThere was no one clear method of adaptation for anyspecific population or culture but there were a few ob-served similarities in the values targeted for distinctgroups. For example, familism was frequently presentedas a value central to Latino culture. At the same time,other studies identified collectivist beliefs as present inmany Asian cultures. These orientations may representsimilar characteristics with respect to familial relation-ships and may result in similar adaptations. Observa-tions of this type highlight the fact that race or commonancestry alone is not useful for selecting types of culturaladaptations. Variations within cultures and the degree towhich people practice or support values commonly asso-ciated with their culture are important in designingeffective adaptations. To address this, some researchershave focused specifically on degree of acculturation,while others have attempted to devise more individual-ized adaptations.In addition to specific cultural orientations, someproviders have also attempted to be more aware of, andresponsive to, past traumatization of whole populationssuch as African Americans in the U.S. [13] and indigen-ous peoples in Canada and the U.S. [10]. In some cases,the most appropriate goal may be to build trust from themoment of the service user’s introduction to service pro-viders or their organizations, including for example,Healey et al. BMC Health Services Research  (2017) 17:8 Page 19 of 26developing more welcoming physical surroundings.While these changes are not uncommon, no studies thatwere found isolated and tested such changes.Moderator effectsOccasionally, researchers observed null findings overall, butreported significant interaction effects related to culture. Anumber of studies found recipients’ level of acculturationto the predominant society moderated the effects ofcultural adaptations. For example, Wang et al. found that aculturally adapted video was more effective for women whowere lower in acculturation [48]. Pan et al., Gondolf, andBurrow-Sanchez et al. likewise found that adapted pro-grams were more effective amongst Asian Americans less-acculturated to U.S. culture, African Americans with highracial identification, and Latino adolescents with highethnic identity commitment, respectively [47, 65, 77].These findings provide some support for tailoring basedon level of acculturation, which was featured in threereviewed studies: Huey and Pan and Pan et al., Kreuter etal., and Resnicow et al. [60–65]. Two studies that went onto use cultural tailoring based on acculturation did notfind it significantly more effective than non-tailoredtreatments. Resnicow et al. found that culturally tai-lored materials based on ethnic identity were signifi-cantly more effective than standard materials fortheir Afro-centric subgroup [63].A complex issueThe observed interaction effects suggest that treatmentefficacy is contingent on multiple variables. One methodof specified tailoring observed in the literature is toindividualize content based on numerous demographicvariables, which could range from ethnicity and agegroup to preferred communication style and interest inmaterials [81]. While this review did not focus onindividualization, the studies found indicate it may showpromise in improving recipient outcomes, and mayfunction best when applied to multiple factors.Successful adaptation is further complicated by thepossibility that research participants may be fundamen-tally different than those who choose not to participateat all, or those who are never even approached for re-search. Both groups could be different from researchparticipants on a variety of social, economic, and per-sonal factors. Medical mistrust is one plausible exampleof the potential unrepresentativeness of research sam-ples; if a person does not trust those in the medical fieldenough to seek help for their ailments it is unlikely thatthey would permit medical researchers to study them.Those most in need of culturally competent servicesmay not be involved in the development of culturallycompetent interventions designed to reach them.Statistical significanceIt is possible that some retained studies failed to observesignificant effects because the sample population wasnot as underserved as the initial population of interest.For example, Jandorf and colleagues, in a comparison ofcompleters to drop-outs, noted that all of their partici-pants (experimental and control) had insurance coverage[51]. The implication is that differential access tomedical information may not be as pronounced in thissample as it would be in a sample including uninsuredindividuals.Another commonly cited explanation for observed re-sults was that the statistical significance of the effectsmay have been obscured by the nature of the controlconditions; rarely is a standard intervention completelyculturally insensitive. The retained studies all tested cul-tural adaptations in isolation, but they often did so aboveand beyond pre-existing adaptations to improve effect-iveness. For instance, Holt et al. conducted both theirexperimental and control conditions within a church, soeven the standard condition was perceived as highlyspiritual by recipients [66, 67].Included studies also tended to focus on statistical sig-nificance rather than effect sizes. Small sample sizes inmany studies may have made it difficult for small effectsto achieve significance—effects which may have practicalimportance. For example, Lee et al. measured previous-month heavy drinking days, and although the outcomewas statistically non-significant, there was medium effectsize in favor of the adapted group [54]. It should benoted that even minimally increased efficacy could havepractical and financial effects when one takes intoaccount the millions of individuals that are currentlyinvolved with some form of health services.Mechanisms of influenceAlthough mechanisms of influence are not often elaboratedin the research reviewed, it was possible to discern fourmain pathways by which researchers sought to achievemore effective services for underserved populations: (1) ad-dressing systemic barriers such as location of service, trans-portation, language, child care, and affordability of services,(2) increasing community engagement to help identify andameliorate barriers, (3) integrating cultural perspectives andvalues directly into the intervention, and (4) enhancing theservice experience, thereby increasing satisfaction andideally the likelihood of access, uptake, and follow-throughto result in improved health outcomes. These mechanismsare not mutually exclusive but they highlight major path-ways to potentially successful intervention (increased ac-cess, uptake, follow-through, knowledge, and ultimately,changes in health status or behavior). The included re-search clearly indicates that one or a combination of theideal outcomes is not necessarily sufficient to ultimatelyHealey et al. BMC Health Services Research  (2017) 17:8 Page 20 of 26achieve changes in health. The mechanisms change tosome degree with the nature of the desired outcome.Overcoming systemic barriers Finances, geography, orcomprehension ability are some examples of systemic bar-riers to accessing adequate health care. Addressing thesesystemic barriers was an important focus in a few of theretained studies. The efficacy of Skaer and colleagues’mammography vouchers highlights the importance offinancial barriers, identified in prior community researchas the greatest barrier to treatment for their population ofinterest [57]. The addition of illustrations to simplify con-tent, and therefore address barriers of communication andunderstanding, has likewise been shown to be effective inimproving some health outcomes, particularly knowledgeand understanding [59, 72]. The question then becomeshow to translate increased understanding into behaviorchange, e.g., seeking treatment or adhering to medication[59, 72].Community engagement Community engagement wasoften the foundation upon which adapted interven-tions were developed. Researchers strove to involvecultural consultants and community members in thedevelopment of their interventions and focus-grouptesting of the interventions to ensure their validity[37, 46, 48, 49, 54–56, 60–67, 71, 72, 76, 77, 80]. Itwas not possible to determine to what degree this engage-ment involved true partnerships with the communityversus less involved approaches such as brief consultation.Such authentic partnerships are often identified as the firststep to making services more relevant to diverse peoples[82]. Interestingly, at least one researcher observed the ad-aptations developed in concert with the community werenot noticed by the target population. Pan et al. found in amanipulation check that every participant in the adaptedcondition believed they were receiving the standard, un-adapted treatment, in spite of the extensive evidence-based cultural adaptations made to psychotherapy [65].Whether or not recognition of adaptation is associatedwith success is unknown. In practice, of the studies thatincluded involvement of the community in the develop-ment of an intervention, a little less than half were foundto be effective in some way [46, 62–65, 71, 72, 76, 80].Integrating cultural context and perspectives withtreatment Some studies attempted to achieve betterhealth outcomes by integrating aspects of culture or cul-tural understanding with the treatment activity. Four ofthe nine adapted interventions that had a practical im-pact on the health outcomes of recipients implementedsupplemental cultural components that were not avail-able in the standard intervention that addressed directneeds of the service recipients [54–58]. Seven of thenine implemented packages of adaptations both modi-fied the content and the manner in which the contentwas delivered. Wording was simplified, group compos-ition was altered, and the treatment procedure changed,among other adaptations [38, 46, 55, 56, 72, 74, 80].Some attempted to increase the level of interactivity andrapport between provider and recipient, extending the con-cept of partnership to provider-recipient interactions. Panet al., for instance, adapted their style of communication(being more directive) to be more congruent with their cli-ents’ expectations [65]. These efforts were not necessarilymore effective than standard treatment, for example, Jan-dorf et al.’s use of peers as navigators [51, 52], but it did re-sult in some distinct findings. For example, although allfour studies that used Pathways to Freedom in their inter-ventions [69, 73–75] were judged as statistically effective, itwas the increase in quit rate among recipients in theadapted group in Orleans et al. that was considered prac-tically important. Orleans et al. implemented a counselingcomponent in addition to the Pathways to Freedom guidethat was culturally adapted to be more interactive [74].Likewise, the intervention group in the Kalichman et al.study that received both adapted content and a change inmanner in service delivery was the only condition thatresulted in recipients screening for HIV [46].The Havranek et al. study sought to improve the qualityof provider-recipient interaction. It was the only retainedstudy to examine both the service provider and service re-cipient outcomes simultaneously, as a pair [58]. A largeeffect was observed, with increased client communicationthat resulted in less provider-dominance. This study maysuggest that provider competency can be influencedthrough client-based interventions, but also that provider-client communication can be improved in a dyadic andrecursive way. Interventions targeting both service pro-viders and service recipients may not only be able to com-plement each other, but may also produce interactioneffects that have a wider impact than previously expected.Interestingly, the enhanced communication and patients’increased knowledge about their conditions was notaccompanied by requests for more information about ser-vices or increases in reported satisfaction or increased trustin the provider.Service satisfaction Client or patient engagement maybe fostered by feelings of satisfaction with the servicesprovided (e.g., [83]). The literature on service satisfactionsuggests that the construct is multi-faceted and includesexpectations of the service user, the quality of the servicesreceived, feelings about the experience, the degree towhich the service recipients’ beliefs about the services areconfirmed in a positive or negative way, and whether thetreatment was fair or equitable [84]. It may be measuredglobally with respect to overall satisfaction or in moreHealey et al. BMC Health Services Research  (2017) 17:8 Page 21 of 26detailed fashion, with indicators of several facets of satis-faction. Although some retained studies assessed servicesatisfaction, not all studies considered mechanisms of in-fluence related to satisfaction. Burrow-Sanchez et al., forexample, measured satisfaction as a health outcome inand of itself as a component of intervention feasibility[76]. In some cases attempts at increasing satisfaction maybe ineffective. In the case of Jandorf et al., the professionalnavigators scored higher in satisfaction, and also resultedin a higher colonoscopy completion rate, though not sta-tistically significant [51]. In studies of cultural adaptations,it is possible that recipients’ expectations may contributeto the improvement of such outcomes as client satisfac-tion or trust in providers [85] — effects which may go un-noticed in studies that use participant-blinding to enhanceinternal validity.Implications for health policy and service providersImproved health outcomes for diverse populations canbe fostered in many ways. In addition to larger societalchanges, such as more equitable access to nutrition,health, and education, providers and policy makers canalso influence the service experience and outcomes.The paradoxOne of the essential lessons is that wholesale changes inmaterials provided or even service delivery personnelwill not automatically be helpful. The paradox of tryingto make services more culturally sensitive is that it canresult in over-generalization regarding what is importantto service users. The research suggests service recipientsin this modern, socially interconnected world likely be-long to multiple cultural or social groups simultaneously[86]. No one adaptation or package was found to be acure-all in this review, nor is it plausible that a singleadaptation will ever be the answer to cultural compe-tency. Each theory, each intervention, and each outcomeis simply a piece of a larger puzzle that needs to be in-crementally assembled to build culturally competent ser-vices. It is critically important to be aware of variationswithin cultural groups. With such awareness, for example,one might identify the locus of the problem in a particularservice or geographic area, identify barriers to access anduptake and the existing service elements that reinforcethese barriers, and understand the variety of paths toachieving service effectiveness for all concerned.Potential pathways to changeThere appear to be numerous possible pathways to successbut few, if any, have been systematically replicated overtime. For policy makers, one possible first step is to assurethat pilot projects have sufficiently controlled studies to in-form the question of efficacy. Additionally, reaching out tothe community to determine their needs/barriers toservice and then clearly addressing those needs begins tobuild the reciprocal relationship that is necessary at boththe community and individual levels of service. Supple-mental services to address barriers may provide some hopefor change if they are actually responsive to the felt need.Attending to the recipient’s experience with services deliv-ered, whether or not they recognize any cultural adapta-tions, may also support service engagement and laterfollow-through. Attention to and reporting of cost analysesin all efforts would be enormously helpful for future deci-sion making.In terms of practice, service for diverse populations canbe optimized by taking great care in individualizing, un-derstanding, and showing respect for each person’s indi-vidual needs and barriers [87]. Inquiring about a person’sculture and what is important to them is a major first step.At least one study showed that by strengthening the indi-vidual service recipient’s self-assurance and affirming theirvalues, improved communication is possible.Interventions not included in the reviewThe studies reviewed were necessarily limited. Interven-tions for which the adaptation cannot be isolated or cul-tural alternatives to standard adaptations that do not havea reasonable comparison (e.g. cultural treatments like FirstNations art or elements of tai chi) should not be ignored,and should be interpreted and incorporated into all dia-logue relating to cultural competence, albeit in a way thatdiffers from those retained in this review of isolated cul-tural adaptations. See Additional file 1 for further reviews.Implications for health researchThe lack of available detail on the adaptations studiedwas disconcerting during this review. Promoting the useof an adapted intervention requires the provision ofsufficient detail such that adaptations can be effectivelyincorporated into practice, for instance, provision of easyaccess to treatment manuals to ensure fidelity. Addition-ally, the implementation of tested interventions withinexisting systems should be plausible, or at least providesuggestions regarding how different systems could beadapted to incorporate these interventions.Specificity of adaptations and outcomesUnderstanding in this field can be improved with specificfocus on a variety of health outcomes, using methods thatisolate and analyze adaptations differing in number, type,and depth. Such focus will help develop a deeper under-standing of recipients’ and providers’ health experiencesand maximize the effectiveness of health interventions.Studies that provide more explicit detail and documenta-tion of mechanisms of change (e.g., including client satis-faction) would make major contributions. As in mostsystematic reviews, the dearth of attention to cost analysesHealey et al. BMC Health Services Research  (2017) 17:8 Page 22 of 26is a hindrance to future progress. Research should beundertaken in a manner that permits and emphasizes theeffects that services have on the health experience of theirclients, in as much detail as possible throughout their in-volvement with health services.Building the evidenceCareful, generative research will aid in the systematic de-velopment of effective interventions. The lack of researchthat sufficiently controls for other influences is delayingthe development of the most effective cultural adapta-tions. Further, additional systematic replications with vari-ations in sample and methods may illuminate anypotential differential patterns of results related to cultureand/or geographical region, help support the developmentof adaptations that are not amenable to highly controlledconditions, and inspire novel methods of cultural adapta-tion. Some efforts may actually have a deleterious effecton service recipients, such as inducing stereotype threatrather than combatting it or correcting for it, and havemixed or reverse impacts. Interventions must be shown tobe congruent with the needs of the population in whichthey are being implemented through evidenced based re-search and thorough needs assessments.Limitations of this reviewRCTs and very rigorous quasi-experimental designs, bytheir very nature, limit the participants and contextswithin which the studies are conducted. For example, thetypes of organizations that are able to conduct such astudy, the types of adaptations that can be evaluated withtightly controlled research designs, and even the countriesthat have sufficient resources to run such studies will bias,to some extent, the knowledge resulting from such studies(see, e.g., [88, 89]). This project focused on more stringentdesigns due to the amount of literature available regardingless rigorous approaches (Additional file 1) and the dearthof information available about actual effectiveness of cul-tural adaptations. The inclusion of more rigorous studiesyielded more information about potential effectiveness ofcultural adaptation than is available in other publicationsto date and complements existing knowledge in the field.Other limitations stem from the procedures requiredfor such a sizeable study and the limited source of thematerials that were qualified for inclusion. For instance,several teams of reviewers conducted the screening.Though several safeguards for consistency were in place,differences among the teams are possible. Tests for pub-lication bias were not conducted due to the extraordin-ary variety of adaptations, outcomes, and designs used inthe studies. Limiting the search to reports in English alsoprecludes all studies published in only non-English lan-guages. Lastly, all 31 retained studies were conducted inthe United States, so the application of their results, andthe corresponding conclusions of this review, may notgeneralize to other nations with differing social, political,cultural, and/or economic structures.ConclusionThis review identifies the most rigorous research in thefield of cultural adaptations of health and mental health ser-vices, and presents study findings within two conceptualframeworks. These frameworks allow for more systematiccategorization of health outcomes and cultural adaptationsto inform and support future research and practice in thisarea. The results suggest several important directions fordevelopment of future practice, policy, and research. Forpractitioners, the literature suggests that high quality ser-vices are the result of engaging communities, understand-ing the needs and desires of the patient or clientpopulations, and adapting to their needs as much as pos-sible in each service encounter. Policy should likewise becongruent with the needs of those involved, and should beinformed by both the impacts on individuals and a macro-scopic understanding of local communities as a whole.Policies should maximize benefit and minimize harm for allthose they affect.As efforts to produce culturally competent services con-tinue, future research should focus on the isolated studyof cultural adaptations, alone and in packages, to identifywhich among them augment efficacy. The moderating roleof acculturation could be explored to a greater extent toyield a more complete understanding of the role of tailor-ing in health and mental health. The variations withingroups also appears to support the individualization ofservices. The exploration of which interventions are ef-fective, for whom, and what sort of outcomes they influ-ence continues to be of importance in health and mentalhealth service delivery, and is critical to establishing cul-tural competence and promoting health and mental healthin our diverse, multicultural societies.Endnote1Kalichman et al. tested both an adaptation in isolation,but also a package of adaptations in a second comparison[46]. As a result, this study has been counted twice.Additional filesAdditional file 1: Similar Reviews and Meta-analyses. Other reviews oncultural health disparities and service adaptations located during searchand screening process. (DOCX 36 kb)Additional file 2: Glossary. A list of main terms used in this report withtheir corresponding operational definition. (DOCX 23 kb)Additional file 3: Complete Electronic Database Search Strategy for theSystematic Review. A record of search strategy and dates for theelectronic database searches. (DOCX 48 kb)Healey et al. BMC Health Services Research  (2017) 17:8 Page 23 of 26Additional file 4: Grey Literature Sources. A list of sources used at thetime of the grey literature search, with year searched if the link changedbetween 2012 and 2015. (DOCX 20 kb)Additional file 5: Detailed Table of Tested Adaptations and Outcomesby Report. A table sorting retained studies, their adaptations, and theiroutcome variables in accordance with the conceptual adaptation andoutcome frameworks (Tables 3 and 4). (DOCX 83 kb)Additional file 6: One Example of a Conceptual Hierarchy of ClientEngagement. An example of another potential dimension by whichadaptations could be evaluated. (DOCX 20 kb)Additional file 7: Forest Plot - Intervention Effects by Number ofAdapted Elements. A forest plot illustrating treatment efficacy sorted bynumber of adapted elements. (DOCX 84 kb)AbbreviationsADHD: Attention deficit hyperactivity disorder; CA: Culturally adapted;CBCL: Child behavior checklist; CD: Conduct disorder; CRC: Colorectal cancer;DrInC: Drinkers’ inventory of consequences; DSM-IV TR: Diagnostic and statisticalmanual of mental disorders, 4th edition text revision; ECBI: Eyberg childbehavior inventory; ECI: Early Childhood inventory; FOBT: Fecal occult bloodtest; ODD: Oppositional defiant disorder; PCIT: Parent child interaction therapy;PI: Principal investigator; PLOC: Parental locus of control; PSI: Parenting stressindex; RCT: Randomized controlled trial; STD: Standard; TAU: Treatment as usual;USDHHS: U.S. Department of Health and Human ServicesAcknowledgementsJennifer Bitz, project manager par excellence, kept the project on time andwithin budget. Consultation in designing the study and interpreting theresults was sought from our advisory panel and knowledge users. Theknowledge users included: 1) nationally and internationally known educators,Drs. Shirley Chau and Dixon Sookraj; 2) experts in large system analysis(Director of the American Bar Association Center on Children and the Law,Mr. Howard Davidson, and Director of the Child Welfare League of Canada,Mr. Peter Dudding; and 3) a representative of managers of service delivery,Mr. Daniel Sheriff. The project consultants included: 1) Drs. Allan Best andJulia Littell, internationally known experts in the field of knowledge translationand research synthesis; 2) researchers from Canada’s National Aboriginal HealthOrganization (NAHO), Dr. Simon Brascoupé (Director, NAHO First Nations Centre),and Ms. Nicole Robinson; 3) an expert information scientist with many years’experience in systematic reviews, Ms. Mimi Doyle-Waters; and 4) Dr. DonnBaumann, a social psychologist who has undertaken system-wide researchon state-wide cultural adoptions of service delivery systems in the State ofTexas, USA. Additionally, the investigators sought out the voice of the Indigenouscommunity through representatives of Kelowna’s Indigenous serviceorganizations, Ms. Molly Brewer and Ms. Niki Stevenson of Ki-Low-Na FriendshipSociety, Ms. Valerie Richards of Métis Community Services, and Ms. Kama Steligaand Ms. Margaret Eli of Westbank First Nation Community Services. Their inputwas invaluable over the entire course of the project from the formulation of thestudy questions to the review of this document. Research assistantswho contributed to the search and screening process included: DavidMcCarty-Caplan, Ian Jantz, and Jacqueline May Kanippayoor.FundingThis study was supported by the Canadian Institutes of Health Research(www.cihr-irsc.gc.ca), grant # KIM-111609.Availability of data and materialsScreening records are available by request from the corresponding author.Electronic databases available online: MEDLINE (https://www.ncbi.nlm.nih.gov/pubmed/), PsycINFO (http://search.proquest.com/advanced). Databasesaccessed through UBC’s server: CINAHL (https://health.ebsco.com/products/the-cinahl-database), EMBASE (https://www.elsevier.com/solutions/embase-biomedical-research). Grey literature links are available in Additional file 4.Authors’ contributionsSJW and AJD were responsible for the conception, design, and implementationof the study throughout the entire project including leading the writing of thearticle for publication. RJ designed the search methodology, ran each iterationof database searches, and provided methodological content for the article. PHand AV managed and implemented the initial study and wrote the 2012report. PH wrote the first draft of this article with SJW and AJD. MLS andKW managed and conducted the 2015 update and revised the article forpublication with SJW and AJD. All authors read and approved the finalmanuscript.Authors’ informationPH – BA Honours, MA CYC Student (UVic). Program coordinator for the SafeHome and Employment Services Programs, South Okanagan Women inNeed Society.MLS – BA Honours student (UBC). Conducts research for the Centre for theStudy of Services to Children and Families at the University of BritishColumbia’s Okanagan Campus.KW – BA Honours. Conducts research for the Centre for the Study of Servicesto Children and Families at the University of British Columbia’s OkanaganCampus.AJD – PhD, MSW. Dean and Maconda Brown O’Connor Endowed Dean’sChair of the Graduate College of Social Work, University of Houston.AV – BSc Honours. A probation officer with Manitoba Justice and youthsupport worker with Macdonald Youth Services.RJ – BA, MLIS. Associate Chief and Nursing Liaison Librarian at the Universityof British Columbia’s Okanagan Campus.SJW – PhD, MSW. Joint appointment in the Department of Psychology and theSchool of Social Work and Director of the Centre for the Study of Services toChildren and Families at the University of British Columbia’s Okanagan Campus.Competing interestsThe authors declare that they have no competing interests.Consent for publicationNot Applicable.Ethics approval and consent to participateNot Applicable.Author details1Centre for the Study of Services to Children and Families, University ofBritish Columbia, ASC 453, 3187 University Way, Kelowna, BC V1V 1V7,Canada. 2University of Houston Graduate College of Social Work, 3511 CullenBlvd Room 110HA, Houston, TX 77204-4013, USA. 3University of BritishColumbia, Okanagan Campus Library, LIB 241, 3287 University Way, Kelowna,BC V1V 1V7, Canada.Received: 15 January 2016 Accepted: 15 December 2016References1. Canadian Institute for Health Information. Health Indicators 2013. 2013.https://secure.cihi.ca/free_products/HI2013_EN.pdf. Accessed 20 Dec 2016.2. U.S. Department of Health and Human Services; Centers for Disease Controland Prevention. CDC health disparities and inequalities report—UnitedStates, 2013. 2013. http://www.cdc.gov/mmwr/pdf/other/su6203.pdf.Accessed 20 Dec 2016.3. Lê Cook B, Doksum T, Chen C, Carle A, Alegría M. The role of provider supplyand organization in reducing racial/ethnic disparities in mental health care inthe U.S. Soc Sci Med. 2013. doi:10.1016/j.socscimed.2013.02.006.4. Lee HY, Vang S. Barriers to cancer screening in Hmong Americans: Theinfluence of health care accessibility, culture, and cancer literacy. J CommunityHealth. 2010. doi:10.1007/s10900-010-9228-7.5. Benjamins MR, Whitman S. Relationships between discrimination in healthcare and health care outcomes among four race/ethnic groups. J BehavMed. 2014. doi:10.1007/s10865-013-9496-7.6. Shavers VL, Fagan P, Jones D, Klein WMP, Boyington J, Moten C, Rorie E.The state of research on racial/ethnic discrimination in the receipt of healthcare. Am J Public Health. 2012. doi:10.2105/AJPH.2012.300773.7. Tejeda S, Darnell JS, Cho YI, Stolley MR, Markossian TW, Calhoun EA. Patientbarriers to follow-up care for breast and cervical cancer abnormalities.J Womens Health. 2013. doi:10.1089/jwh.2012.3590.8. Rooks RN, Wiltshire JC, Elder K, BeLue R, Gary LC. Health information seekingand use outside of the medical encounter: Is it associated with race andethnicity? Soc Sci Med. 2012. doi:10.1016/j.socscimed.2011.09.040.Healey et al. BMC Health Services Research  (2017) 17:8 Page 24 of 269. Sorkin DH, Ngo-Metzger Q, De Alba I. Racial/ethnic discrimination inhealth care: Impact on perceived quality of care. J Gen Intern Med.2010. doi:10.1007/s11606-010-1257-5.10. Van Herk KA, Smith D, Andrew C. Identity matters: Aboriginal mothers’experiences of accessing health care. Contemp Nurse. 2011. doi:10.5172/conu.2011. González HM, Vega WA, Williams DR, Tarraf W, West BT, Neighbors HW.Depression care in the United States: Too little for too few. Arch GenPsychiatry. 2010. doi:10.1001/archgenpsychiatry.2009.168.12. Lee C, Ayers SL, Kronenfeld JJ. The association between perceived providerdiscrimination, health care utilization, and health status in racial and ethnicminorities. Ethn Dis. 2009;19:330–7.13. Cuffee YL, Hargraves JL, Rosal M, Briesacher BA, Schoenthaler A, Person S,et al. Reported racial discrimination, trust in physicians, and medicationadherence among inner-city African Americans with hypertension. Am JPublic Health. 2013. doi:10.2105/AJPH.2013.301554.14. Beach MC, Price EG, Gary TL, Robinson KA, Gozu A, Palacio A, et al. Culturalcompetence: A systematic review of health care provider educationalinterventions. Med Care. 2005. doi:10.1097/01.mlr.0000156861.58905.96.15. Bhui K, Warfa N, Edonya P, McKenzie K, Bhugra D. Cultural competence inmental health care: A review of model evaluations. BMC Health Serv Res.2007. doi:10.1186/1472-6963-7-15.16. Nierkens V, Hartman MA, Nicolaou M, Vissenberg C, Beune EJAJ, Hosper K,et al. Effectiveness of cultural adaptations of interventions aimed at smokingcessation, diet, and/or physical activity in ethnic minorities. A systematicreview. PLoS One. 2013. doi:10.1371/journal.pone.0073373.17. Chipps JA, Simpson B, Brysiewicz P. The effectiveness of cultural-competence training for health professionals in community-basedrehabilitation: A systematic review of literature. Worldv Evid-Based Nu. 2008.doi:10.1111/j.1741-6787.2008.00117.x.18. Hasnain R, Kondratowicz, DM, Portillo N, Borokhovski E, Balcazar F, Johnson T,et al. The use of culturally adapted competency interventions to improverehabilitation service outcomes for culturally diverse individuals withdisabilities. Submitted to the Campbell Collaboration, Education CoordinatingGroup. 2009. http://www.ncddr.org/partners/subgroup/resources/hasnain_competency_interventions_review_2010.pdf. Accessed 20 Dec 2016.19. Hawthorne K, Robles Y, Cannings-John R, Edwards AGK. Culturally appropriatehealth education for type 2 diabetes in ethnic minority groups: A systematic andnarrative review of randomized controlled trials. Diabet Med. 2010. doi:10.1111/j.1464-5491.2010.02954.x.20. Pottie K, Hadi A, Chen J, Welch V, Hawthorne K. Realist review to understand theefficacy of culturally appropriate diabetes education programmes. Diabet Med.2013. doi:10.1111/dme.12188.21. Bailey EJ, Cates CJ, Kruske SG, Morris PS, Brown N, Chang AB. Culture-specific programs for children and adults from minority groups who haveasthma. Cochrane DB Syst Rev. 2009;2:CD006580.22. Herbst JH, Kay LS, Passin WF, Lyles CM, Crepaz N, Marin BV. A systematic reviewand meta-analysis of behavioral interventions to reduce HIV risk behaviors ofHispanics in the United States and Puerto Rico. AIDS Behav. 2007;11:25–47.23. Wilson BDM, Miller RL. Examining strategies for culturally grounded HIVprevention: A review. AIDS Educ Prev. 2003;15:184–202.24. Fitzgibbon ML, Tussing-Humphreys LM, Porter JS, Martin IK, Odoms-YoungA, Sharp LK. Weight loss and black women: A systematic review of thebehavioral weight loss intervention literature. Obes Rev. 2012. doi:10.1111/j.1467-789X.2011.00945.x.25. Griner D, Smith TB. Culturally adapted mental health intervention: A meta-analytic review. Special issue: Culture, race, and ethnicity in psychotherapy.Psychother. 2006;43:531–48.26. Smith TB, Domenech Rodríguez M, Bernal G. Culture. J Clin Psychol: InSession. 2010. doi:10.1002/jclp.20757.27. Lie DA, Lee-Rey E, Gomez A, Bereknyei S, Braddock III CH. Does culturalcompetency training of health professionals improve patient outcomes? Asystematic review and proposed algorithm for future research. J Gen InternMed. 2010. doi:10.1007/s11606-010-1529-0.28. Markus HR, Moya PML, editors. Doing race: 21 essays for the 21st century.New York: W W Norton & Company; 2010.29. Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews ofinterventions, Version 5.1.0. Chicester. UK: Wiley-Blackwell; 2011.30. Greenhalgh T, Peacock R. Effectiveness and efficiency of search methods insystematic reviews of complex evidence: audit of primary sources. BMJ.2005;331:1064–5.31. Grimshaw J. A guide to knowledge synthesis. Ottawa: Canadian Institutes ofHealth Research; 2010.32. Littell JH, Corcoran J, Pillai V. Systematic reviews and meta-analysis. NewYork: Oxford University Press; 2008.33. Popay J, Roberts H, Sowden A, Petticrew M, Arai L, Rodgers M, et al.Guidance on the conduct of narrative synthesis in systematic reviews.Lancaster: Institute for Health Research, Lancaster University; 2006.34. U. S. Department of Health and Human Services, Office of Minority Health.National standards for culturally and linguistically appropriate services in healthcare: Final report. Washington, DC: Author. 2001. http://minorityhealth.hhs.gov/assets/pdf/checked/finalreport.pdf. Accessed 20 Dec 2016.35. Balshem H, Helfand M, Schünemann HJ, Oxman AD, Kunz R, Brozek J, et al.GRADE guidelines: 3. Rating the quality of evidence. J Clin Epidemiol. 2011.doi:10.1016/j.jclinepi.2010.07.015.36. Breitkopf CR, Dawson L, Grady JJ, Breitkopf DM, Nelson-Becker C, Snyder RR.Intervention to improve follow-up for abnormal papanicolaou tests: Arandomized clinical trial. Health Psychol. 2014. doi:10.1037/a0032722.37. Johnson CA, Unger JB, Ritt-Olson A, Palmer PH, Cen SY, Gallaher P, Chou C.Smoking prevention for ethnically diverse adolescents: 2-year outcomes ofa multicultural, school-based smoking prevention curriculum in SouthernCalifornia. Prev Med. 2005;40:842–52.38. La Roche MJ, Koinis-Mitchell D, Gualdron L. A culturally competent asthmamanagement intervention: A randomized controlled pilot study. Ann AllergAsthma Im. 2006;96:80–5.39. Chowdhary N, Jotheeswaran AT, Nadkarni A, Hollon SD, King M, Jordans MJ,Rahman A, Verdeli H, Araya R, Patel V. The methods and outcomes ofcultural adaptations of psychological treatments for depressive disorders:a systematic review. Psychol Med. 2014;44:1131–46.40. Castro FG, Barrera Jr M, Martinez Jr CR. The cultural adaptation of preventioninterventions: Resolving tensions between fidelity and fit. Prev Sci. 2004;5:41–5.41. Castro FG, Barrera Jr M, Steiker LK. Issues and challenges in the design ofculturally adapted evidence-based interventions. Annu Rev Clin Psycho.2010;6:213–39.42. Barrera M, Castro FG. A heuristic framework for the cultural adaptation ofinterventions. Clin Psychol-Sci Pr. 2006;13:311–6.43. Bernal G, Bonilla J, Bellido C. Ecological validity and cultural sensitivity foroutcome research: issues for the cultural adaptation and development ofpsychosocial treatments with Hispanics. J Abnorm Child Psych. 1995;23:67–82.44. Resnicow K, Soler R, Braithwaite RL, Ahluwalia JS, Butler J. Cultural sensitivityin substance use prevention. J Community Psychol. 2000;28:271–90.45. Newton Jr RL, Perri MG. A randomized pilot trial of exercise promotion insedentary African-American adults. Ethn Dis. 2004;14:548–57.46. Kalichman SC, Kelly JA, Hunter TL, Murphy DA, Tyler R. Culturally tailoredHIV-AIDS risk-reduction messages targeted to African-American urbanwomen: Impact on risk sensitization and risk reduction. J Consult Clin Psych.1993;61:291–5.47. Gondolf EW. Program completion in specialized batterer counseling forAfrican-American men. J Interpers Violence. 2008;23:94–116.48. Wang JH, Schwartz MD, Brown RL, Maxwell AE, Lee MM, Adams IF,Mandelblatt JS. Results of a randomized controlled trial testing the efficacyof a culturally targeted and a generic video on mammography screeningamong Chinese-American immigrants. Cancer Epidem Biomar. 2012a.doi:10.1158/1055-9965.EPI-12-0821.49. Wang JH, Schwartz MD, Luta G, Maxwell AE, Mandelblatt JS. Interventiontailoring for Chinese American women: Comparing the effects of twovideos on knowledge, attitudes and intentions to obtain a mammogram.Health Educ Res. 2012b. doi:10.1093/her/cys007.50. Fitzgibbon ML, Stolley MR, Ganschow P, Schiffer L, Wells A, Simon N,Dyer A. Results of a faith-based weight loss intervention for black women.J Natl Med Assoc. 2005;97:1393–402.51. Jandorf L, Cooperman JL, Stossel LM, Itzkowitz S, Thompson HS, Villagra C,et al. Implementation of culturally targeted patient navigation system forscreening colonoscopy in a direct referral system. Health Educ Res. 2013a.doi:10.1093/her/cyt003.52. Jandorf L, Braschi C, Ernstoff E, Wong CR, Thelemaque L, Winkel G, et al.Culturally targeted patient navigation for increasing African American’sadherence to screening colonoscopy: A randomized clinical trial. CancerEpidem Biomar. 2013b. doi:10.1158/1055-9965.EPI-12-1275.53. Ard JD, Kumanyika S, Stevens VJ, Vollmer WM, Samuel-Hodge C, Kennedy B,et al. Effect of group racial composition on weight loss in AfricanAmericans. Obesity. 2008. doi:10.1038/oby.2007.49.Healey et al. BMC Health Services Research  (2017) 17:8 Page 25 of 2654. Lee CS, López SR, Colby SM, Rohsenow D, Hernández L, Borrelli B, CaetanoR. Culturally adapted motivational interviewing for Latino heavy drinkers:Results from a randomized clinical trial. J Ethn Subst Abuse. 2013. doi:10.1080/15332640.2013.836730.55. McCabe K, Yeh M. Parent–child interaction therapy for Mexican Americans:A randomized clinical trial. J Clin Child Adolec. 2009. doi:10.1080/15374410903103544.56. McCabe K, Yeh M, Lau A, Argote CB. Parent–child interaction therapy forMexican Americans: Results of a pilot randomized clinical trial at follow-up.Behav Ther. 2012. doi:10.1016/j.beth.2011.11.00.57. Skaer TL, Robison LM, Sclar DA, Harding GH. Financial incentive and the useof mammography among Hispanic migrants to the United States. HealthCare Women In. 1996;17:281–91.58. Havranek EP, Hanratty R, Tate C, Dickinson LM, Steiner JF, Cohen G,Blair IA. The effect of values affirmation on race-discordant patient-provider communication. Arch Intern Med. 2012. doi:10.1001/2013.jamainternmed.258.59. Mohan A, Riley B, Schmotzer B, Boyington DR, Kripalani S. Improvingmedication understanding among Latinos through illustrated medicationlists. Am J Manag C. 2014;20:e547–55.60. Kreuter MW, Skinner CS, Steger-May K, Holt CL, Bucholtz DC, Clark EM,Haire-Joshu D. Responses to behaviorally vs culturally tailored cancercommunication among African American women. Am J Health Behav.2004;28:195–207.61. Kreuter MW, Steger-May K, Bobra S, Booker A, Holt CL, Lukwago SN,Skinner CS. Sociocultural characteristics and responses to cancer educationmaterials among African American women. Cancer Control. 2003;10:69–80.62. Kreuter MW, Sugg-Skinner C, Holt CL, Clark EM, Haire-Joshu D, Fu Q, et al.Cultural tailoring for mammography and fruit and vegetable intake amonglow-income African-American women in urban public health centers.Prev Med. 2005;41:53–62.63. Resnicow K, Davis R, Zhang N, Tolsma D, Alexander G, Wiese C, et al.Tailoring a fruit and vegetable intervention on ethnic identity: Results of arandomized study. Health Psychol. 2009. doi:10.1037/a0015217.64. Huey SJ, Pan D. Culture-responsive one-session treatment for phobic AsianAmericans: A pilot study. Psychother (Chic). 2006. doi:10.1037/0033-3204.43.4.549.65. Pan D, Huey SJ, Hernandez D. Culturally adapted versus standard exposuretreatment for phobic Asian Americans: Treatment efficacy, moderators, andpredictors. Cult Divers Ethn Min. 2011. doi:10.1037/a0022534.66. Holt CL, Scarinci IC, Debnam K, McDavid C, Litaker M, McNeal SF, et al.Spiritually based intervention to increase colorectal cancer awarenessamong African Americans: Intermediate outcomes from a randomized trial.J Health Commun. 2012a. doi:10.1080/10810730.2012.665418.67. Holt CL, Litaker MS, Scarinci IC, Debnam KJ, McDavid C, McNeal SF, et al.Spiritually based intervention to increase colorectal cancer screeningamong African Americans: Screening and theory-based outcomes from arandomized trial. Health Educ Behav. 2012b. doi:10.1177/1090198112459651.68. Halbert CH, Kessler L, Troxel AB, Stopfer JE, Domchek S. Effect of geneticcounseling and testing for BRCA1 and BRCA2 mutations in AfricanAmerican women: A randomized trial. Public Health Genomi. 2010.doi:10.1159/000293990.69. Webb MS, Baker EA, Rodriguez DY. Effects of culturally specific cessationmessages on theoretical antecedents of behavior among low-income AfricanAmerican smokers. Psychol Addict Behav. 2010. doi:10.1037/a0018700.70. Sanders Thompson VLS, Kalesan B, Wells A, Williams S, Caito NM. Comparingthe use of evidence and culture in targeted colorectal cancer communicationfor African Americans. Patient Educ Couns. 2010;81 Suppl 1:S22–33.71. Holt CL, Wynn TA, Litaker MS, Southward P, Jeames SE, Schulz EK. Acomparison of a spiritually based and non-spiritually based educationalintervention for informed decision making for prostate cancer screeningamong church-attending African American men. Urol Nurs. 2009;29:249–58.72. Unger JB, Cabassa LJ, Molina GB, Contreras S, Baron M. Evaluation of afotonovela to increase depression knowledge and reduce stigma amongHispanic adults. J Immigr Minor Health. 2013. doi:10.1007/s10903-012-9623-5.73. Nollen N, Ahluwalia JS, Mayo MS, Richter K, Choi WS, Okuyemi KS, ResnicowK. A randomized trial of targeted educational materials for smokingcessation in African Americans using transdermal nicotine. Health EducBehav. 2007;34:911–27.74. Orleans CT, Boyd NR, Bingler R, Sutton C, Fairclough D, Heller D, et al. A self-help intervention for African American smokers: Tailoring cancer informationservice counseling for a special population. Prev Med. 1998;27:S61–70.75. Webb MS. Culturally specific interventions for African American smokers:An efficacy experiment. J Natl Med Assoc. 2009;101:927–35.76. Burrow-Sanchez JJ, Wrona M. Comparing culturally accommodated versusstandard group CBT for Latino adolescents with substance use disorders: Apilot study. Cult Divers Ethn Min. 2012. doi:10.1037/a0029439.77. Burrow-Sanchez JJ, Minami T, Hops H. Cultural accommodation of groupsubstance abuse treatment for Latino adolescents: Results of an RCT.Cult Divers Ethn Min. 2015. doi:10.1037/cdp0000023.78. Shoptaw S, Reback CJ, Peck JA, Yang X, Rotheram-Fuller E, Larkins S, et al.Behavioral treatment approaches for methamphetamine dependence andHIV-related sexual risk behaviors among urban gay and bisexual men. DrugAlcohol Depen. 2005;78:125–34.79. Chiang C, Sun F. The effects of a walking program on older Chinese Americanimmigrants with hypertension: A pretest and posttest quasi-experimentaldesign. Public Health Nurs. 2009. doi:10.1111/j.1525-1446.2009.00776.x.80. Hwang WC, Myers HF, Chiu E, Mak E, Butner JE, Fujimoto K, et al. Culturallyadapted cognitive-behavioral therapy for Chinese Americans with depression: Arandomized controlled trial. Psychiat Serv. 2015. doi:10.1176/appi.ps.201400358.81. Resnicow K, Zhou Y, Hawley S, et al. Communication preference moderatesthe effect of a tailored intervention to increase colorectal cancer screeningamong African Americans. Patient Educ Couns. 2014;97:370–5.82. Ashcraft L. Successful service integration depends on authenticpartnerships. Behav Healthc. 2011;31(3):6–8.83. Kendra M, Weingardt K, Cucciare M, Timko C. Satisfaction with substanceuse treatment and 12-step groups predicts outcomes. Addict Behav. 2015;40:27–32.84. Fraser MW, Wu S. Measures of consumer satisfaction in social welfare andbehavioral health: a systematic review. Res Social Work Prac. 2015.doi:10.1177/1049731514564990.85. Zhang Z, Gerstein DR, Friedmann PD. Patient satisfaction and sustainedoutcomes of drug abuse treatment. J Health Psychol. 2008.doi:10.1177/1359105307088142.86. Perez AD, Hirschman C. The changing racial and ethnic compositionof the US population: Emerging American identities. Popul Dev Rev. 2009.doi:10.1111/j.1728-4457.2009.00260.x.87. Saha S, Beach MC, Cooper LA. Patient centeredness, cultural competenceand healthcare quality. J Natl Med Assoc. 2008;100(11):1275–85.88. Clay RA. More than one way to measure. Monit Psychol. 2010;41:52–5.89. Cartwright N, Munro E. The limitations of randomized controlled trials inpredicting effectiveness. J Eval Clin Pract. 2010;16:260–6.•  We accept pre-submission inquiries •  Our selector tool helps you to find the most relevant journal•  We provide round the clock customer support •  Convenient online submission•  Thorough peer review•  Inclusion in PubMed and all major indexing services •  Maximum visibility for your researchSubmit your manuscript atwww.biomedcentral.com/submitSubmit your next manuscript to BioMed Central and we will help you at every step:Healey et al. BMC Health Services Research  (2017) 17:8 Page 26 of 26


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items