transcreation: an implementation science framework for

DEBATE Open Access

Transcreation: an implementation scienceframework for community-engagedbehavioral interventions to reduce healthdisparitiesAnna María Nápoles1* and Anita L. Stewart2

Abstract

Background: Methods for translating evidence-based behavioral interventions into real-world settings seldomaccount for the special issues in reaching health disparity populations.

Main text: The objective of this article is to describe an innovative “transcreational” framework for designing anddelivering interventions in communities to reduce health disparities. We define transcreation as the process ofplanning, delivering, and evaluating interventions so that they resonate with the community experiencing healthdisparities, while achieving intended health outcomes. The Transcreation Framework for Community-engagedBehavioral Interventions to Reduce Health Disparities comprises seven steps: 1) identify community infrastructureand engage partners; 2) specify theory; 3) identify multiple inputs for new program; 4) design interventionprototype; 5) design study, methods, and measures for community setting; 6) build community capacity fordelivery; and 7) deliver transcreated intervention and evaluate implementation processes. Communities areengaged from the start and interventions are delivered by community-based interventionists and tested incommunity settings. The framework applies rigorous scientific methods for evaluating program effectiveness andimplementation processes. It incorporates training and ongoing technical assistance to assure treatment fidelity andbuild community capacity.

Conclusions: This framework expands the types of scientific evidence used and balances fidelity to evidence andfit to the community setting. It can guide researchers and communities in developing and testing behavioralinterventions to reduce health disparities that are likely to be sustained because infrastructure development isembedded in the research.

Keywords: Evidence-based interventions, Intervention adaptation, Translation, Health disparities, Community-engaged research, Implementation science

BackgroundHealth disparities in the United States (U.S.) based onrace, ethnicity, socioeconomic status, geography, gender,sexual orientation, and disability status continue to per-sist. According to the Centers for Disease Control andPrevention (CDC) Health Disparities and InequalitiesReport—U.S. 2013, the risk of premature death due to

cardiovascular disease, the leading cause of death, was atleast 50% higher among African Americans than Whites[1]. African Americans, Latinos, and those who are poor,less educated or disabled experienced a higher preva-lence of diabetes and obesity, compared to their counter-parts [1]. Risk of uncontrolled hypertension was higheramong Mexican Americans, immigrants and the unin-sured [2], and rates of HIV diagnosis were eight-foldhigher among African Americans and two-fold higheramong Latinos and Native Hawaiians/other Pacific Is-landers compared to Whites [3]. Regional variations

* Correspondence: [email protected] Institute on Minority Health and Health Disparities, 9000 RockvillePike, Building 3, Floor 5, Room E08, Bethesda, MD 20892, USAFull list of author information is available at the end of the article

© The Author(s). 2018 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.

Nápoles and Stewart BMC Health Services Research (2018) 18:710 https://doi.org/10.1186/s12913-018-3521-z

http://crossmark.crossref.org/dialog/?doi=10.1186/s12913-018-3521-z&domain=pdf

http://orcid.org/0000-0001-8838-2899

mailto:[email protected]

http://creativecommons.org/licenses/by/4.0/

http://creativecommons.org/publicdomain/zero/1.0/

were also notable, with Southeastern African Americanand Whites experiencing shorter life expectancy thantheir counterparts from other U.S. regions [1]. Between1992 and 2006, mortality rates among women increasedin 42.8% of U.S. counties while only 3.4% of countiesexperienced mortality increases among men [4]. Sexualand gender minority (SGM) groups experienced higherincidence and mortality rates due to chronic disease,more depression, anxiety and suicidal ideation, andgreater substance use than non-SGM groups [5]. Adultswith a disability were more likely to have cardiovasculardisease, be obese, be a current smoker, be physically in-active, and less likely to receive preventive health screen-ings than those without a disability [6].To address these inequities, the National Institutes of

Health (NIH) Health Disparities Strategic Plan recom-mends translating scientific discoveries including effectivebehavioral interventions to reach disparity populations, de-fined as subgroups that have worse health compared to thegeneral population [7]. Evidence-based interventions (EBIs)are research-tested behavioral interventions found to be ef-fective in achieving desired outcomes [8]. Most EBIs weredeveloped and tested in academic settings for mainstream,highly selected populations. Fewer EBIs were designed foror have been applied in disparity populations, thus they arenot reaching those most likely to benefit [9, 10].To address health disparities, we need translational

models/frameworks that can address the special issuesof communities experiencing such disparities, and pro-vide methodological guidelines to researchers. The NIHStage Model describes six general stages of translation(from basic research to implementation), but does notprovide methodological details for implementinginterventions [11]. Another model, the ConsolidatedFramework for Implementation Research focuses on ataxonomy used to classify characteristics of the interven-tion, the inner and outer settings, individuals, and imple-mentation processes, rather than specifying practicalmethodological steps [12].We identified five models for disseminating EBIs into

real-world settings from the fields of implementationscience and public health that describe methodologicalsteps. These include Intervention Mapping [13], theCenter for Disease Control and Prevention’s (CDC) Rep-licating Effective Programs (REP) [14], CDC’s Division ofHIV/AIDS Prevention (DHAP) Framework [15], theExploration, Preparation, Implementation, SustainmentFramework (EPIS) [16], and the Evidence Driven Com-munity Health Improvement Process (EDCHIP) [17].Strengths of these frameworks are that most addressmultilevel influences on program implementation,emphasize late-stage translation in real-world settings,and attempt to balance scientific evidence with fit to thecontext. Some emphasize involving communities in the

process [15–17]. However, none of these translationalmodels focus on the special methodological issues per-taining to reducing health disparities or recognize theproduct as a new intervention. Methods for transferringscientific knowledge into programs to reduce health dis-parities have just begun to receive attention. Chinmanand colleagues consider addressing disparities as a “spe-cial case” of implementation science and suggest blend-ing methods from health disparities research andimplementation science [9]. To reduce disparities inhypertension, Mueller and colleagues recommend “de-signing and testing pragmatic interventions in real-worldsettings using implementation research methods” (p.712) [18]. Lopez-Class and colleagues argue that “latestage” translation research (T4), i.e., efficacy and effect-iveness trials in community settings, holds potential forreducing health disparities [19]. These recent papersprovide a starting point for systematic efforts to advancetranslation of EBIs to reduce health disparities.

Why we need a new implementation science frameworkto address health disparitiesAddressing health disparities requires delivering effectiveinterventions that are acceptable, practical, and designed toaddress the health needs of at-risk populations. Transla-tional models need to account for large differencesbetween the original EBI conditions (academic setting,mainstream population, professionals as interventionists)and the health disparity population and community [9, 10].Interventions need to be delivered in community as

well as health care settings to reach disparity populationsthat tend to have limited access to health care. Transla-tion models should account for community strengthsand incorporate the resourceful solutions they have devel-oped. The communities’ knowledge of the populations andthe drivers of disparities can inform development of effect-ive programs. To address resource constraints, interven-tions need to strategically build on existing infrastructureand develop new infrastructure to facilitate sustainability[20]. Efforts to bolster community resources and capacityincrease the likelihood of success and sustainability [17]. Totake full advantage of these resources and address dispar-ities in a sustainable way, translational models need to in-corporate methods for engaging communities throughoutthe entire process.Regarding populations, one needs to consider the

characteristics of the disparity populations and how theymight differ from mainstream populations in whichmost EBIs have been tested. Key differences can includeculture, language, English proficiency, literacy, poverty,health beliefs, access to resources, and others. Interven-tions resulting from this process need to be sensitive tothese characteristics.

Nápoles and Stewart BMC Health Services Research (2018) 18:710 of 15

To address these issues, we need a paradigm that en-ables us to design and evaluate interventions deliveredin real-world community settings from the outset. Thereare numerous EBIs that have been shown to be effica-cious under optimal conditions. We can expedite thesuccessful translation of these interventions to new set-tings and populations experiencing health disparities,but this process involves unique methodological steps.These steps help to ensure that the intervention reflectsrelevant scientific evidence (including EBIs) and simul-taneously fits the target audience and community fromthe start.In this paper, we propose a new paradigm that we refer

to as “transcreation” for designing and implementing be-havioral interventions specifically for communities experi-encing health disparities. We present the TranscreationFramework for Community-engaged Behavioral Interven-tions to Reduce Health Disparities (referred to henceforthas the transcreation framework) that describes methodsfor designing, delivering, and evaluating interventions toreach disparity populations in community settings. Ourmethods are intended to address limitations of traditionalresearch-to-practice models for translation of scientificknowledge, specifically when implementing behavioral in-terventions to reduce health disparities.

Main textThe transcreation frameworkThe term transcreation has been applied widely in thefield of marketing. Among global marketing profes-sionals, transcreation is recognized as the processwhereby messaging and content are developed oradapted for an audience so that they resonate in localmarkets and yet deliver the same impact as the original[21]. Use of the term “transcreation” in the health arenahas had a narrower scope, referring to the adaptation ofhealth education materials for improved understandingand cultural relevance to specific language and ethnicgroups [22]. We define transcreation as the processes ofplanning and delivering interventions to reduce healthdisparities so that they resonate with the targeted com-munity, while achieving intended health outcomes.The transcreation framework aims to guide researchers

and community partners to reduce health disparities bydeveloping and testing behavioral interventions that aregrounded in scientific evidence and build on communitystrengths. It shifts the emphasis from adapting and trans-lating one EBI for delivery in a new setting, to the designof a new intervention that better fits local needs and con-texts. Our transcreation framework calls for research inwhich evidence building starts with interventions that aretested initially in community settings and disparity popu-lations, rather than beginning with an efficacy trial con-ducted under optimal but constrained circumstances.

Onken and colleagues suggest that this type of researchlies somewhere between traditional efficacy research(tested using optimal conditions) and effectiveness re-search (tested in real world settings) [11]. Chinman andcolleagues refer to this as a “hybrid” design that can accel-erate the pace for translating EBIs into the real world [9].Our framework emphasizes scientific rigor, starting

with theory, integrating scientific evidence in interven-tion design, utilizing scientific methods to evaluate ef-fectiveness of interventions delivered in communitysettings, and applying stringent methods of evaluatingimplementation processes. It involves engaging commu-nities throughout the research process - researcherswork with community stakeholders to transcreate EBIsto be meaningful and deliverable with fidelity in commu-nity settings. This includes methods for capitalizing oncommunity resources and building capacity by engagingcommunity members in intervention delivery, and evenin recruitment and evaluation. This strategy increasesthe likelihood that interventions will be culturally sensi-tive, appealing, and sustainable. The seven steps of thetranscreation framework are listed in Fig. 1.This framework builds on our prior methodological

framework and training resource for adapting behavioralinterventions to address health disparities [23]. We havemade several major advances in our framework. Wenow include two types of theory: a socio-ecological the-ory of the determinants of the health disparity being tar-geted, and an appropriate behavior change theory. Weplace a greater emphasis on balancing fit to the contextwith fidelity to scientific evidence. We have incorporatedmethods for hiring and training community members asinterventionists and study staff, including developingstandardized intervention manuals. Finally, we specifymethods for three types of evaluation: formative to de-velop the intervention, process to understand factors thataffect implementation, and summative after the trial toinform future implementation and dissemination efforts.Perhaps the most important advance is shifting awayfrom the concept of translation and adaptation of EBIsto transcreation, since it has become apparent to us thatapplication of our framework results in a new interven-tion designed specifically with community partners to re-duce a targeted disparity.The revised framework incorporates our experience

and research with several behavioral interventions to re-duce disparities since that publication, as well as emer-ging literature on the intersection of implementationscience and health disparities research [9]. Our own ex-perience includes conducting community-based trials inmultiple diverse minority and underserved communitiesexperiencing health disparities (African Americans, Lati-nos, poor Whites, rural and urban, low-income cancer pa-tients, and low-income individuals at risk for diabetes).


We have tested several types of behavioral interventions(cognitive stress management, physical activity, and diet),using a variety of delivery modes (trained peers, groupsled by public health practitioners, mobile phones, healthcoaches) that have improved health behaviors, physicaland emotional well-being, and other outcomes [24–26].Use of the framework assumes that an academic/com-

munity partnership has been established. Prior to usingthe framework, there should be a shared understandingof the specific health disparity that will be addressed bythe partnership and intervention, e.g., higher prevalenceof a risk factor in a disparity population group or lack ofaccess to culturally competent services for a specificcondition that are a shared concern. In Table 1, wepresent a detailed description of the seven steps. Al-though these steps are described sequentially, in prac-tice, they may be iterative or occur at multiple timepoints in the transcreation process.

Step 1: Identify community infrastructure and engagepartnersTranscreation research builds on existing communityinfrastructure and expands it for implementation. Step1A is to identify intervention delivery settings andacademic partners that share a targeted health priorityand have appropriate infrastructure and experience.The composition of the academic-community coali-tion that is formed in this step will vary dependingon the health priority being addressed, community

needs, infrastructure and the ability to convenegroups and individuals [17]. This infrastructure mightinclude organizations or public health departments fo-cused on chronic diseases and academic institutions.Settings serving disparity populations can includecommunity-based organizations, government entities,community-based health care clinics, senior centers,public health departments, faith-based organizations,and other systems. Optimal settings can include thosealready delivering an intervention to address the tar-geted problem, e.g., to develop a stress managementintervention for Latinas with breast cancer, we part-nered with a community organization that providedcancer support programs for low-income Latinos [27].Step 1B focuses on identifying individuals who are

stakeholders with deep knowledge of the targetedhealth disparity and population and who are inter-ested in implementing an intervention to reduce thetargeted disparity. Partners can include multisectorrepresentatives, e.g., administrators, staff members,clients, community residents, and civic and privatesector partners.Partners form an academic-community coalition

that is based on community-based participatory re-search (CBPR) principles (e.g., trust, shared decision-making, equal value placed on scientific and commu-nity knowledge) and is responsible for all subsequentsteps. Using CBPR principles of engagement, partnerswill be engaged in all phases of research including

Fig. 1 The Transcreation Framework: A Seven-Step Process. This figure depicts the seven steps involved in designing, delivering and evaluatingbehavioral interventions in communities to reduce health disparities


Table 1 Transcreation framework for community-engaged behavioral interventions to reduce health disparities

Step Description and Methods

1. Identify community infrastructure & engage partners

A. Identify optimal community settings Identify settings with infrastructure for delivering and sustainingintervention, experience providing similar services, and thatare convenient for the target population.

B. Identify partners and form an academic-communitycoalition to address the health disparity

Identify persons in the community with intimate knowledgeof disparities and contributing factors and academic partnerswith relevant interest and experience; involve them in allphases of planning and execution.Develop an academic-community coalition based on CBPRprinciples (e.g., trust, shared decision-making, value scientificand community knowledge equally) that is responsible forall subsequent steps.

2. Specify theory

A. Identify framework of social-ecological determinants of disparities Identify, review, and select theoretical framework of determinantsof disparities that fits the health priority and context. This involvesusing CBPR engagement principles.

B. Determine theoretical basis for behavioral changes Working with the coalition identified in Step 1, identify, reviewand select one or more theories of behavior change relevantto the targeted health disparity population.

3. Identify multiple inputs for new program

A. Identify scientific evidence relevant to planned intervention toaddress targeted health disparity

Review evidence-based interventions (EBIs), systematic reviews,evidence-based guidelines, and other evidence (e.g., optimaldelivery modes) that can inform the intervention.For EBIs, identify core components that reflect mechanismsof action.

B. Obtain input from community on locally developed programs,community resources, and target population (formative evaluation)

Review locally developed programs relevant to targeted disparity.

Identify cultural, practical, organizational, and contextual factorsaffecting intervention design, success, and sustainability throughformative research including community resources, populationcharacteristics and needs, and knowledge of determinants ofdisparity.

4. Design intervention prototype

A. Design intervention to incorporate scientific evidence and locallydeveloped programs.

Examine commonalities of EBIs and other scientific evidenceand locally developed programs and synthesize.

Build in intervention components and delivery methods thatare supported by scientific evidence; include componentsthat address hypothesized mechanisms of action.

B. Design intervention for fit to community setting and population Build consensus on fit of potential intervention to communityand potential to address targeted disparity. Assure thatintervention is practical, accessible, and can be deliveredwithin existing resources or with some capacity building.

Incorporate content to fit population including culture, language,and learning styles, and format to fit reading level and preferredcommunication channels.

C. Integrate 4A and B to develop intervention components; vetprototype for relevance and potential for success.

Design specific components and content by weighingtradeoffs between scientific evidence and fit to setting(community, population).

Specify delivery format (e.g., phone, in-person, group), locationof in-person sessions, who delivers the program, sessionformat/content, and dose.

Document components and rationale for each; have keystakeholders review prototype, modify as indicated.

D. Manualize intervention to assure standardization Develop detailed program manual for interventionists that specifiesprogram delivery components, content, and procedures. Manualprovides guidance to interventionists on methods to increase thefidelity of program delivery.Develop participant manual using principles for low-literacyparticipants.


providing inputs to program design, formative re-search, creation and vetting of a prototype, implemen-tation processes, summative (post-implementation)evaluation, and dissemination of results and successfulprograms [28]. Partners need to be compensatedthrough direct funding in proportion to the time and

effort expended on the project. Employing CBPRprinciples to ensure equitable partnerships throughoutand remunerating community partners and stake-holders helps build trust over time and ensures thatongoing community input is integrated and culturallyappropriate approaches are discussed [28].

Table 1 Transcreation framework for community-engaged behavioral interventions to reduce health disparities (Continued)

Step Description and Methods

5. Design study, methods, & measures for community setting

A. Develop rigorous study design that is appropriate for interventiondelivered in community setting

Randomized controlled trials, especially individual-level, maynot be the most appropriate design. Identify alternatives thatretain scientific rigor, e.g., cluster randomized trials, pragmatictrials, rigorous quasi-experimental designs.

B. Develop outreach, recruitment, and data collection strategiesappropriate for population and setting

Design strategies based on evidence of effectiveness in disparitypopulations and perspective of community.

Utilize community members for outreach, recruitment, and datacollection in community settings to the extent possible; developtraining protocol.

C. Select measures of outcomes, mediators, and moderators that arerelevant and appropriate for population

Select outcomes based on conceptual framework linkingcomponents to outcomes. Identify measures that are responsiveto similar interventions. Identify measures of moderators andmediators of effects.Assure that all measures are appropriate for the target populationand meet stringent psychometric criteria in that population.

6. Build community capacity for delivery

A. Enhance infrastructure and expertise Compensate community organizations and members forresearch involvement.Train community organization staff on skills that can be appliedto fund, deliver, and evaluate programs in the future.

B. Select/train community-based interventionists(an interim intervention)

Establish qualifications for community-based interventionists(e.g., CHWs) including desired level of competence/knowledgeafter training.Hire and train community-based interventionists. Utilizinginterventionist manual developed in step 4D, create trainingprotocol including: 1) content, format, theory, and protocolof “transcreated” intervention, 2) delivery skills includingcommunication skills, handling problems, and 3) importanceof fidelity to the protocol.

7. Deliver “transcreated” intervention & monitor implementation processes

A. Create and implement methods for monitoring delivery ofintervention and providing ongoing technical assistance tointerventionists

Create ongoing technical assistance plans for interventionistsas they deliver the intervention.Develop structured assessment for monitoring fidelity ofinterventionists to protocol.Establish system for providing feedback and support tointerventionists if needed to improve adherence or preventburnout.Establish system for modifying intervention if needed to addressunanticipated situations.

B. Create and implement methods for assessing other processes of deliveringintervention (summative evaluation)

Design specific procedures and data collection strategies toassess implementation processes.Interventionists: Suggested improvements, difficulties deliveringprogram, acceptability of training and manual.Participants: Real-time - Program receipt (attendance, how wellthey learned components) and enactment (can demonstrate skills).Retrospective - perceived benefits of program, suggestedimprovements, perceived usefulness and ease of use of programcomponents and materials.Stakeholders such as program managers, executive directors:issues in implementing program in that setting; successes andchallenges of implementing program, suggestions forimprovement.


Step 2: Specify theoryBeginning with a strong theoretical basis for the targeteddisparity and intervention aids in understanding mecha-nisms of action [29], assuring that core components areincluded. Step 2A focuses on theoretical frameworksspecifying social-ecological determinants of health,health behavior, and health disparities [19, 30, 31]. Aca-demic researchers and community stakeholders work to-gether to identify, review and select appropriate healthdisparities and health behavior change theories andframeworks [28]. Selection of these frameworks will bebased on their relevance to the health priority, popula-tion and context. For example, working with communitypartners in one of our studies, we identified importantdeterminants of poor emotional well-being in the targetpopulation as marital/family discord due to miscommu-nication and financial hardship, so these became part ofour larger health disparity framework. Selection of theappropriate theories is based on both scientific and com-munity data and knowledge about the health priorityand population. Western-based theoretical approachescan be blended with indigenous-based theories [28].Step 2B is to apply theories of behavior change that

account for individual and social factors; this assuresthat the intervention will be theory-based. For example,Social Cognitive Theory [32] has been applied in behav-ioral interventions for disparity populations [33, 34], be-cause of the potency of its principles of self- monitoring,goal setting, and problem solving that can increaseself-efficacy, which often mediates improved outcomes.Delineating the theory underlying the program may in-clude a logic model describing how the program works.A compendium of health behavior change theories canguide these discussions [35].

Step 3: Identify multiple inputs for new programTwo basic types of inputs to the planning process includescientific evidence and input from the community includ-ing locally developed programs and knowledge. Step 3Aincludes identifying and reviewing scientific evidence,which can include EBIs as well as other types of evidence.EBIs can be identified via articles and reviews, meta ana-lyses, and websites. Of importance is to determinewhether any minority or other disparity groups were in-cluded in EBI testing. It is helpful to secure program ma-terials for candidate EBIs via websites or by contactingprogram developers. For each EBI, identifying core com-ponents that reflect specific mechanisms of action is key[11]. This should result in a conceptual framework of howeach component leads to hypothesized outcomes, i.e.,pathways/key mechanisms of effects.Other types of scientific evidence can be as important as

EBIs, including systematic reviews and evidence-based guide-lines. Examples include evidence-based recommendations

about community preventive services and interventions toimprove health [17], and a synthesis of evidence on the roleof self-efficacy for achieving optimal outcomes of self-man-agement interventions [36].Of special importance to health disparities is the siz-

able amount of scientific evidence supporting the effect-iveness of interventions delivered by community healthworkers (CHWs). CHW interventions can improve pre-ventive behaviors and health outcomes in disparity pop-ulations, most notably in under-resourced areas.Systematic reviews of CHW interventions have substan-tiated their effectiveness in improving glucose and lipidlevels, systolic and diastolic blood pressures, physical ac-tivity, dietary behaviors, and mental health outcomes[37, 38]. CHWs can enhance intervention effectivenessand reach because they are trusted by and share the ex-periences of individuals in disparity communities. Evi-dence on the value of the roles of other communitypractitioners (e.g., social workers, community health careproviders) needs to be considered.Step 3B is to use formative research to obtain input

from the community and the disparity population to beserved to understand cultural, practical and contextualfactors that might affect intervention design, implemen-tation, and sustainability [39]. Community stakeholderinputs include locally developed programs, deep know-ledge of the specific health disparity, frameworks andmeasures, and organizational and community resourcesthat can be harnessed to increase the likelihood of suc-cess and sustainability [28]. Interventions that build onexisting programs and resources will enhance the fit ofthe intervention to the setting. A review by theacademic-community coalition of organizational struc-tures, staff, skills, and inter-organizational networkswithin which agencies deliver interventions helps iden-tify resources (e.g., community asset mapping). A thor-ough assessment of the target population, its risk profile,and determinants of health disparities that incorporatesthe knowledge and experience of local community mem-bers and practitioners is needed. Formative research caninclude focus groups and semi-structured interviewswith community key informants.

Step 4: Design intervention prototypeIntervention design involves the academic-communitycoalition synthesizing/integrating scientific evidence andcommunity input (inputs identified in Step 3) into a pro-gram description. Step 4A is to assure that the interven-tion reflects relevant scientific evidence. By includingcore components of EBIs, program content can addressmechanisms while integrating other scientific evidence.For example, a peer-delivered stress management pro-gram for Latinas with breast cancer incorporated com-ponents reflecting known mechanisms for improving


quality of life, namely, strategies to enhance self-efficacyfor managing cancer, and peer support from a Latinacancer survivor [27]. Use of peer delivery was based onevidence of the effectiveness of CHWs in diverse com-munities and on a specific peer support model devel-oped by one of the partner community agencies. Inanother study, evidence that action plans were effectivein improving self-efficacy [40] was the basis for usingthem in a diabetes risk reduction intervention forlower-socioeconomic status and minority adults [33].When translating interventions for sexual and genderminority older adults, Fredriksen-Goldsen and col-leagues suggest including components directly address-ing the higher rates of relevant risk factors in thesepopulations (e.g., stigma, discrimination, victimization),which also need to be included in the relevant healthdisparity framework [41].Step 4B includes designing the intervention to fit the

population and community, and involves utilizing pro-gram delivery methods, staffing models, and communityassets identified in Step 3B. Again, this involves iterativeprocesses by the academic-community coalition mem-bers to review all candidate sources of knowledge andscientific evidence thus far, until consensus is reachedon core components. Materials and strategies need toaccommodate cultural and language/literacy issues, con-venience, and accessibility, to improve participation. Forexample, knowing that the target population has limitedtransportation and discretionary time can result in a de-sign in which group-based components are delivered inneighborhood settings to increase access [33]. Step 4Bcan involve applying methodological frameworks for cul-turally adapting interventions such as identifying mes-sages that support ethnic practices or the uniquecontexts of persons with disabilities [42]. Low-literacymaterials and visual aids can improve knowledge amongindividuals with a range of literacy levels [43]. Regardingfit to the community, practical issues can affect imple-mentation, adoption and sustainability [44], requiringstrategies to address resource needs. For example, to re-duce costs, translation of a caregiver support programfor Area Agencies on Aging resulted in fewer sessionsthan the original [45].Step 4C is for the coalition to integrate 4A-B into a

prototype and vet the components. The integration caninvolve comparing form (key elements/components) andfunction (mechanisms and processes that lead to desiredoutcomes) of the EBIs [46]. Such an approach allows“context level adaptation” of the intervention form, e.g.,varying the literacy level of materials, while preservingthe function, e.g., mechanisms of action specified by theintervention’s underlying behavior change theory [46].The academic-community coalition considers and de-cides together what aspects of, and how, EBIs need to be

adapted to maximize potential effectiveness and adop-tion, while preserving their scientific integrity. Typically,for interventions aimed at reducing health disparities,this step is heavily influenced by community membersand stakeholders that include individuals struggling withthe specific disparity being targeted.Nápoles and colleagues describe the entire process of

integrating an EBI tested in white women, a locally de-veloped program for Latinas, and formative research tocreate Nuevo Amanecer, a stress management interven-tion for Latinas with breast cancer [27]. Mejia and col-leagues describe a variety of programs that have resultedfrom processes of adapting EBIs to local contexts, ran-ging from minor surface structure changes to substantialadaptations [47]. In another case, they describe adaptingevidence-based principles (functions) selected by com-munity stakeholders from a number of EBIs, rather thanthe standard model of adapting a single EBI [47]. Thisstep may result in a new program that “may bear littleresemblance to the original EBI as local consumers makedecisions as to what aspects of existing EBIs they deemrelevant to their local communities” (p. 691) [48].Step 4D is to operationalize the program by defining

the specific components. Examples of components in-clude small-group introductory session, written programmanuals, one-on-one planning sessions, telephone coun-seling/coaching, audio-visual materials, and group work-shops [27, 33]. Each component or session can bedescribed in terms of its objectives, content, format (e.g.,phone, in-person), timing, dose (number and duration ofsessions), location, and who delivers it. The program de-scription should include behavioral change strategies beingused such as education, skills training, and motivationalstrategies [33, 34]. This step includes manualizing the inter-vention to assure standardization of key core componentsand functions. Standardization of the manual and materialsprovides a basis for training interventionists and assuring fi-delity of delivery to the transcreated intervention [11]. Con-sultations with the community allow for vetting of programcomponents and final adaptations.

Step 5: Design study, methods, and measures forcommunity settingStep 5 focuses on specialized methodological strategiesfor evaluating the transcreated intervention, i.e., con-ducting experimental studies in community settings.Given that the transcreated program and the settingare new, it is critical to implement a stringent scien-tific evaluation of the intervention being delivered ina community setting where optimal control of studyconditions is not possible. Traditional researchmethods usually require augmentation to be appropri-ate for ethnic minority and lower-socioeconomic sta-tus populations [49].


Step 5A concerns finding the most rigorous study designthat is acceptable to the community and practical. Identify-ing acceptable control groups requires special consideration.Designs in which individuals are the unit of randomizationmay not be appropriate for community-based interventionsdue to practical concerns (e.g., health system-level changesthat may not be under investigator control), ethical reasons(e.g., minority communities experiencing unequal treatmentmay resist the intentional withholding of a treatment to indi-viduals), or scientific considerations (e.g., contamination).Implementation science uses a range of study designs thatallow for causal inferences and approximate experimentaldesigns, with tradeoffs between internal and external valid-ity, e.g., group randomized trials, interrupted time series de-signs. For example, group randomized trials addresspotential bias by matching groups on stable correlates of theoutcome, such as age, in contexts where individualrandomization is culturally unacceptable or not feasible. Aninterrupted time series design is indicated when the specifictime point that an intervention occurs is known, longitu-dinal outcome data are available for a period of time beforeand after intervention onset, and randomization is not feas-ible. Interrupted time series allows assessment of trends be-fore, during and after implementation [50]. Excellentoverviews of the issues involved and alternative designs areavailable [50, 51].Step 5B pertains to development of study methods

such as recruitment and data collection that provide sci-entific rigor and also meet community needs. Becauseconvenience for participants implies conducting recruit-ment and data collection in the community, methodsneed to be well designed and clear, including monitoringto assure consistency with protocols. There is much lit-erature on recruitment of disparity populations. Strat-egies with demonstrated effectiveness include using aconceptual framework of determinants of recruitment toguide efforts and engaging community change agents[52]. For studies targeting participants with specific riskfactors, one strategy is to conduct community-basedhealth education and screening and then recruit thoseidentified as “at risk” [53]. Step 5B includes identifyingstrategies for overcoming common barriers to data col-lection in the community. For example, we developedlanguage and literacy-level appropriate videos demon-strating self-collection procedures for repeated salivarysamples to assess diurnal cortisol rhythms among ruralLatinas with breast cancer. Conducting data collectionin the community is feasible and increases retention ofdisparity populations, but requires considerations suchas obtaining space in community centers and havingportable data collection materials or home-based assess-ment technologies [25].Step 5C focuses on identifying outcome measures that

are responsive to change, linked specifically to

intervention components, and appropriate for the popu-lation. Once academic and community partners haveworked together to identify the health priority and popu-lation being targeted, the health disparity and health be-havior frameworks that will guide measures selection,and the appropriate intervention components that are cul-turally appropriate, feasible and acceptable, outcomes areselected, or developed and pretested if no relevant mea-sures exist. This occurs through a review of candidatemeasures, discussion of alternatives, and co-selection offinal measures so that they will be sensitive to changes tar-geted by the intervention and can be collected in commu-nity settings given resource and time constraints.Selecting measures requires attention to hypothesized

mechanisms, thus a mechanism and its measure shouldbe stated for how each component is linked to a desiredoutcome. A peer-delivered stress management programfor Latinas with breast cancer included components toincrease social support and self-efficacy, thus measuresof social support and self-efficacy were used to evaluatethese as mediators of change [34]. The theoreticalframework underlying an intervention may also point tothe need to measure potential moderators of interven-tion effects, e.g., characteristics of subgroups that mayrespond differentially to treatment effects.All measures need to be culturally appropriate and

meet stringent psychometric criteria in the targeted pop-ulations [54]. This involves defining concepts from theperspective of the target population, locating measuresof those concepts, reviewing whether measures reflectthe concept and have good psychometric properties inthe target population, and choosing the best measure(s)and adapting them if needed [54]. Measures may needto be culturally adapted, translated into other languages,and pretested using cognitive interviewing methods.Measures may need to be modified or simplified in for-mat to accommodate limited literacy or low socioeco-nomic status. It is important to obtain writtenpermission to use or modify existing measures.

Step 6: Build community capacity for deliveryBuilding capacity focuses on enhancing community in-frastructure (identified in Step 3B) by developing acommunity-based workforce that can blend communityknowledge with scientific principles. Community staffcan obtain training in research and evaluation methods,while researchers gain knowledge of a community’s bestpractices, assets, and processes of implementing success-ful programs. Together, this shared exchange of informa-tion enhances opportunities to expand transcreation ofeffective programs.Step 6B is to identify and train qualified individuals

from the community (practitioners, CHWs, communitymembers who have experienced the health problem


being addressed) who can deliver the intervention. Hir-ing interventionists from the community promotes cul-tural sensitivity and trust, and builds communitycapacity [55]. Training these interventionists is an inter-vention in itself, requiring a training protocol and evalu-ation, as well as ongoing technical assistance (step 7A)to assure competency. Training community interven-tionists involves using active learning strategies (e.g., roleplaying and modeling) and detailed interventionist andparticipant program manuals [45]. Training may beneeded on organizational and delivery skills, communica-tion skills, role playing, role modeling desired behaviorchange, troubleshooting implementation problems, main-taining confidentiality, and tracking program activities[56]. Training needs to cover special considerations whenworking with disparity populations, such as limited Eng-lish proficiency, low literacy, and financial hardship. It isimportant to provide interventionists with information oncommunity resources so they can make appropriate refer-rals. Interventionists need to learn how to track partici-pant receipt of the intervention (see step 7), e.g.,attendance, receipt of components, and problems experi-enced by participants with program activities. To assureadequate preparation to deliver the intervention, corecompetencies need to be defined and assessed [45]. Redu-cing health disparities usually requires training commu-nity practitioners and researchers in patient-centeredresearch and CBPR approaches. For CHWs located inrural areas, videoconferencing can be used for trainingand technical assistance support.

Step 7: Deliver “transcreated” intervention and monitorimplementation processesStep 7 involves having trained interventionists deliverthe transcreated program to participants, and monitor-ing implementation processes. When testing the inter-vention first in a community setting, the process ofmonitoring its delivery, termed implementation fidelity[57], is crucial to its success due to the influence of fac-tors outside of researchers’ control. We have built on themethods for monitoring these processes developed bythe NIH Behavior Change Consortium [58], noting howthese methods apply to health disparity interventions.Step 7A is to design and implement methods for monitor-

ing intervention delivery with feedback loops. Demonstrat-ing that an intervention was delivered as intended(treatment fidelity) is important in any effectiveness trial, butis essential when interventions are delivered by trained com-munity members, practitioners or clinicians. Structured fi-delity assessments can occur through ratings of directlyobserved, audiotaped or videotaped sessions. This step in-cludes plans for ongoing technical assistance and/or trainingto interventionists to assure fidelity and prevent unantici-pated implementation challenges [57]. Interventionists need

a way to contact researchers or community partners to re-port problems and obtain help or suggestions. Regularphone calls with interventionists can help. An evaluation ofa church-based nutrition program for African Americansnoted the importance of providing ongoing technical assist-ance to assure effectiveness [59]. Issues can be substantive(procedures for dealing with extremely depressed partici-pant) or logistical (difficulty reaching participants). Re-searchers and community partners can provide help in theform of ideas or resources that are available. For example,similar to our experiences, community interventionists re-ported needing more information on how to manage partic-ipants who were not attending sessions [56]. In somesituations, problems may necessitate modifying the interven-tion protocol.Should fidelity assessments identify concerns, this

step includes a plan to provide sensitive yet directfeedback to interventionists to improve protocol ad-herence. A feedback mechanism can improve practi-tioners’ performance and inform ongoing coachingand supervision [57]. A foundation of trust betweenacademic and community partners is needed whenacademic personnel provide feedback to intervention-ists who are not directly employed by the study.Community organizations often retain control oftranscreated programs, thus need to feel engaged andresponsible for insuring that the program is imple-mented according to protocol.Step 7B is to create and implement methods for moni-

toring an array of intervention delivery processes. Theseinclude quantitative (structured surveys, administrativetracking data) and qualitative (direct observation,open-ended interviews) mixed methods approaches inreal time or retrospectively. Sources of information in-clude interventionists, participants, stakeholders, andobservers. Community partners and stakeholders are in-volved directly, providing feedback and troubleshootingalternatives with academic partners. Results can be trian-gulated across methods and sources. Some of these mea-sures of intervention implementation need to be studyspecific to capture relevant indicators of implementa-tion, e.g., participant mastery of key content. General ap-proaches include monitoring dropouts and their reasons.Semi-structured interviews of participants can assess theperceived usefulness and ease of use of specific compo-nents, and suggested improvements, while probing forin-depth understanding of complex contextual influ-ences on program implementation [24]. Mixed methodscan be used to assess interventionists’ perspectives onpractical and cultural factors affecting participation,acceptability of training and manuals, and perceived pro-gram effectiveness, For example, interventionists deliver-ing a lifestyle intervention to overweight older ruralparticipants identified self-monitoring (e.g., starting each


session with a weigh-in) as important to program suc-cess [56].Stakeholders such as administrators, program direc-

tors, and staff from the community settings hosting theintervention can provide “system-level” feedback [44].This can be done regularly throughout the implementa-tion to enable course corrections, such as reducing theamount of telephone support provided to participantsdue to organizational burden.

ConclusionsThe Transcreation Framework for Community-engagedBehavioral Interventions to Reduce Health Disparitiesincludes the major advantages of other implementationscience frameworks while adding important methodo-logical steps critical for reducing health disparities. Per-haps the most novel feature is recognizing that whenresearchers address social determinants and health dis-parities with the full engagement of community partnersexperiencing the disparity, the intervention produced isnot an adapted EBI, but a new intervention because ofthe extensive adaptations required to fit the communitycontext in the presence of these disparities.We discuss five defining features of our framework: 1)

balances fidelity to scientific evidence and fit to developa new “transcreated” intervention, (2) tests the transcre-ated intervention in community settings, 3) engages thecommunity throughout and builds capacity, 4) usesrigorous scientific methods in community settings, and5) expands the types of evidence used for transcreationof behavioral interventions. We also highlight differencesbetween our framework and the other implementationscience models that we reviewed.

Balances fidelity to scientific evidence and fit tocommunity settingOur framework includes attention to treatment fidelityas well as focusing on the fit of the intervention to thecommunity and testing the intervention in the commu-nity from the outset. Our approach shifts the focusfrom translation of a single tested EBI to the synthesisprocess we call transcreation in which varyingstakeholders build new programs that retain the corecomponents of successful programs and communityknowledge, while making adaptations that improve thefit to local contexts. A major difference between ourframework and the other implementation frameworkswith methodological steps that we reviewed is that theyview the resulting program as a translation or adapta-tion of one EBI, and not a new program that integratesvarious EBIs and other types of evidence, along withcommunity knowledge.

Tests the transcreated intervention in community settingsOur framework calls for research in which interventionsare tested initially in community settings and disparitypopulations rather than beginning with an efficacy trialconducted under optimal but constrained circumstances.This approach means focusing on later stages (T3 andT4) of translation to circumvent the extensive and itera-tive adaptations that might result as researchers proceedthrough earlier translation stages (T1 and T2). As statedby Onken and colleagues, “The intervention develop-ment process is incomplete until an intervention is opti-mally efficacious and implementable with fidelity bypractitioners in the community” (p. 22) [11]. An import-ant consideration in late stage translation is that com-munity partners are the foundation and need to befunded for these efforts. Broad community involvementon the part of affected communities was emphasized inonly two of the five frameworks that offered methodo-logical steps [13, 17].

Engages the community throughout and builds capacityPrimarily, our model differs from other translationmodels in that we propose starting the transcreationprocess with the full integration of the community’sknowledge, local programs, and participants from thebeginning, which is necessary to address social determi-nants of health and maximize feasibility. This approachrecognizes that EBIs may not demonstrate the same levelof efficacy as originally tested unless we include commu-nity members and practitioners in ensuring their fit tolocal contexts. Our framework adheres to CBPR princi-ples and integrates these with implementation processes.Other implementation frameworks tend to include com-munity representatives in limited or early stages of inter-vention adaptation to design and pretest the programprototype. Similar to two of the models with methodo-logical steps that we reviewed, we emphasize building onexisting community resources such as locally developedprograms and services, thus increasing sensitivity to thelocal culture and the likelihood of sustainability [15, 16].This includes training qualified community members orpractitioners as interventionists or as research staff(helping with recruitment and data collection). With on-going training and technical assistance, this unique fea-ture builds community capacity.

Uses rigorous scientific methods in community settingsOur model stresses the use of rigorous scientificmethods for conducting effectiveness trials, outreachand recruitment, and process evaluation in communitysettings. Although the methods used in effectiveness tri-als have been increasingly applied in health disparitiesresearch, they have not focused on the unique methodo-logical considerations when these are applied in


populations experiencing health disparities. The mostprominent models either focus on engaging communi-ties in the process of implementation and capacity build-ing to address multilevel determinants of behavior, [14–17] or emphasize health behavior theory, core compo-nents, and mechanisms of action [13]; none address allof these features or health disparities frameworks. Also,we feature prominently state-of-the-art methods formonitoring implementation, sometimes referred to asprocess evaluation or summative evaluation. We haveemphasized the need for carefully designed yet practicalqualitative and quantitative strategies that fit the context.Formative evaluation serves to secure input from experi-enced stakeholders prior to and during transcreationprocesses, while summative evaluation assesses imple-mentation processes to inform subsequent implementa-tion efforts. Such evaluation is critical, especially whenEBIs are being tested in vulnerable populations typicallyunderrepresented in clinical research [9]. Three of thefive implementation science models that specify meth-odological steps mention stakeholder input at differentstages, but most often stakeholders tend to be agencyplanners or delivery system representatives, and not thecommunity individuals faced with the health disparities;these frameworks also do not emphasize the greaterneed for rigorous evaluation in the context of health dis-parities settings because adaptations tend to be extensive[13, 14, 16].

Expands the types of evidence used for transcreation ofbehavioral interventionsOur model redefines the term “evidence” in “evidence--based.” The term “evidence-based” usually is applied todescribe research-tested programs for which there is em-pirical evidence that they improve outcomes. This nar-row interpretation of “evidence” impedes our ability toreduce disparities because it discounts the equally valu-able evidence and knowledge that public health andcommunity sectors contribute. We encourage health dis-parities research that integrates evidence from system-atic reviews, community guidelines, and best practices.Only two of the five implementation science frameworksthat specify steps consider these additional sources ofevidence [13, 15] and they assume that only one EBI isbeing translated or do not specify that components frommore than one EBI can be utilized.

Implications for reducing health disparitiesProgress on reducing health disparities in the U.S. hasbeen slow due to growing income inequality, systemic dis-crimination, and unequal access to resources, as well asineffective translation models. Achieving population-levelreductions in health disparities will involve concerted ef-forts that focus on where we intervene, how we intervene,

and how we sustain effective programs. Based on emer-ging literature and our experience, we note some researchand policy implications of our framework for reducinghealth disparities through transcreation research on be-havioral interventions.

Accelerate progress toward reducing health disparities byusing transcreation and beginning with later translationstagesCommunity-engaged transcreation takes behavioral in-terventions to the communities where they are needed.Applying our transcreation framework in the communi-ties experiencing pernicious health disparities can en-hance the design of interventions to maximize fit to thepopulation, while preserving the scientific basis of be-havioral interventions, and rigorously evaluating pro-gram outcomes and implementation processes. The goalis to create programs and design implementation strat-egies that reduce health disparities. Once programs andimplementation strategies for delivery in community set-tings are shown to be effective, they can be scaled upand disseminated.

Build community capacity to address health disparitiesthrough transcreation of behavioral interventionsEmbedding programs within existing community servicesand health care delivery networks can extend their reachand acceptability in communities with disparity popula-tions. Implementation strategies that build on communityassets and infrastructure maximize potential scalability andimpact. Engaging communities that are experiencing dis-parities at each stage of the translational process “considersculture and diversity from a community-engaged researchperspective” (p. 115) [20]. Training community practi-tioners or community health workers on cutting-edge EBIsand evaluation methods can enhance their ability to tran-screate effective programs for local communities.Enhanced training of CHWs to deliver effective inter-

ventions across a broad range of health issues is an ex-cellent strategy for building capacity (see Step 3A) [34,60]. Studies have supported the cost-effectiveness ofCHW interventions [61]. CHWs are well-suited for pro-viding higher intensity services needed to reduce dispar-ities in highly impoverished populations [62]. They canbe especially effective when integrated into health caresystems with access to health care teams and electronichealth records [62]. Due to widespread adoption ofsmartphones and internet access by minority popula-tions, technology-enhanced CHW interventions mayhelp address time, cost, and transportation barriers evi-dent in disparity populations and allow interventions tobe interactive and individually tailored for greater patientactivation and effectiveness [63]. More research isneeded to identify populations for whom approaches


using mobile technology tend to be most effective, andthe types of additional support needed to maximize theireffectiveness in specific disparity populations.

Funding models that support sustainability of communityprogramsOne of the biggest challenges to reducing health disparitiesby harnessing the potential of transcreated community-en-gaged interventions is sustainability funding. Sustainabilityfunding and reimbursement opportunities are practicallynon-existent. State-level programs that utilize innovativefunding models could significantly reduce disparities if im-plemented more broadly, but they require substantial re-sources and concerted effort to set up, and may requirespecial legislation [64]. For example, the state of Montanapartnered with Medicaid to reimburse community sites de-livering an adapted Diabetes Prevention Program, achievingexcellent enrollment and weight reduction targets [10]. Thestate of Delaware financed universal community-basedscreening, treatment, and patient navigation for colorectalcancer, resulting in the elimination of statewide disparitiesin colorectal cancer screening, incidence, and mortalitybetween whites and African Americans [65]. In NewMexico, use of Medicaid-funded CHWs to provide naviga-tion services to complex, high-need patients has beencost-effective, reducing emergency room visits, hospitaladmissions, and prescriptions [64]. Greater community-and population-level funding of effective transcreated pro-grams could help reduce health disparities.Our failure to identify effective implementation strat-

egies of proven interventions contributes to the persist-ence of health inequities. Conversely, identifying highlyeffective implementation methods that consider the so-cial determinants of health and community-based solu-tions may contribute significantly to achieving healthequity. Working together, scientists, community practi-tioners, community organizations, and consumers candevelop and test transcreated programs that are sensitiveto contextual factors in communities experiencing dis-parities, build on and strengthen community assets, andare practicable and acceptable. Detailed evaluation oftranscreated program processes and outcomes usingrigorous methods can identify components and contextualfactors that lead to desired behavioral changes and healthimprovements. Such methods for implementing behav-ioral interventions in real-world communities harness sci-entific evidence and community knowledge, apply itdirectly in community settings, and build community cap-acity to implement and evaluate interventions, thereby ac-celerating progress in reducing health disparities.

AbbreviationsCBPR: Community-based participatory research; CDC: Centers for DiseaseControl and Prevention; CHW: Community health worker; EBI: Evidence-based intervention; HIV/AIDS: Human immunodeficiency virus/ acquired

immunodeficiency syndrome; NIH: National Institutes of Health; T1research: Translation to humans; T2 research: Translation to patients; T3research: Translation to practice; T4 research: Translation to populationhealth; U.S.: United States

FundingThis research was supported by the Division of Intramural Research, NationalInstitute of Minority Health and Health Disparities, National Institute onAging grant no.1 P30 AG15272, and California Breast Cancer Research GrantsProgram Office of the University of California grant no. 21OB-0135. None ofthe funding sources were involved in the design of the study and collection,analysis, and interpretation of data and in writing the manuscript.

DisclaimerThe contents and views in this manuscript are those of the authors andshould not be construed to represent the views of the National Institutes ofHealth.

Authors’ contributionsAMN and ALS contributed equally to the conception, writing, and editing ofthis manuscript. Both authors read and approved the final manuscript.

Authors’ informationAMN, Scientific Director, Intramural Research Program, National Institute onMinority Health and Health Disparities.ALS, Professor, Institute for Health & Aging, Center for Aging in DiverseCommunities, University of California San Francisco.

Ethics approval and consent to participateNot applicable.

Consent for publicationNot applicable.

Competing interestsThe authors declare that they have no competing interests.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Author details1National Institute on Minority Health and Health Disparities, 9000 RockvillePike, Building 3, Floor 5, Room E08, Bethesda, MD 20892, USA. 2University ofCalifornia San Francisco, 3333 California Street, Suite 350E, San Francisco, CA94118, USA.

Received: 23 May 2018 Accepted: 4 September 2018

References1. Frieden TR, Centers for Disease Control and Prevention. CDC health

disparities and inequalities report - United States, 2013. Foreword. MMWRSuppl. 2013;62(3):1–2.

2. Gillespie CD, Hurvitz KA, Centers for Disease Control and Prevention.Prevalence of hypertension and controlled hypertension - United States,2007-2010. MMWR Suppl. 2013;62(3):144–8.

3. Johnson AS, Beer L, Sionean C, Hu X, et al. HIV infection - United States,2008 and 2010. MMWR Suppl. 2013;62(3):112–9.

4. Kindig DA, Cheng ER. Even as mortality fell in most US counties, femalemortality nonetheless rose in 42.8 percent of counties from 1992 to 2006.Health Aff. 2013;32(3):451–8.

5. Maingi S, Bagabag AE, O’Mahony S. Current best practices for sexual andgender minorities in hospice and palliative care settings. J Pain SymptomManag. 2018;55(5):1420–7.

6. Krahn GL, Walker DK, Correa-De-Araujo R. Persons with disabilities as anunrecognized health disparity population. Am J Public Health. 2015;105(Suppl 2):S198–206.

7. National Institutes of Health. NIH health disparities strategic plan andbudget, Fiscal years 2009-2013 Washington, D.C.. 2010 https://www.nimhd.


https://www.nimhd.nih.gov/docs/2009-2013nih_health_disparities_strategic_plan_and_budget.pdf

nih.gov/docs/2009-2013nih_health_disparities_strategic_plan_and_budget.pdf. Accessed 14 Aug 2018.

8. Brownson RC, Fielding JE, Maylahn CM. Evidence-based public health: afundamental concept for public health practice. Annu Rev Public Health.2009;30:175–201.

9. Chinman M, Woodward EN, Curran GM, Hausmann LRM. Harnessingimplementation science to increase the impact of health equity research.Med Care. 2017;55(Suppl 9 Suppl 2):S16–23.

10. Purnell TS, Calhoun EA, Golden SH, Halladay JR, et al. Achieving healthequity: closing the gaps in health care disparities, interventions, andresearch. Health Aff. 2016;35(8):1410–5.

11. Onken LS, Carroll KM, Shoham V, Cuthbert BN, et al. Reenvisioning clinicalscience: unifying the discipline to improve the public health. Clin PsycholSci. 2014;2(1):22–34.

12. Hill JN, Locatelli SM, Bokhour BG, Fix GM, et al. Evaluating broad-scalesystem change using the consolidated framework for implementationresearch: challenges and strategies to overcome them. BMC Res Notes.2018;11(1):560.

13. Highfield L, Hartman MA, Mullen PD, Rodriguez SA, et al. Interventionmapping to adapt evidence-based interventions for use in practice:increasing mammography among African American women. Biomed ResInt. 2015;2015:160103.

14. Kilbourne AM, Neumann MS, Pincus HA, Bauer MS, et al. Implementingevidence-based interventions in health care: application of the replicatingeffective programs framework. Implement Sci. 2007;2:42.

15. McKleroy VS, Galbraith JS, Cummings B, Jones P, et al. Adapting evidence-based behavioral interventions for new settings and target populations.AIDS Educ Prev. 2006;18(4 Suppl A):59–73.

16. Becan JE, Bartkowski JP, Knight DK, Wiley TRA, et al. A model for rigorouslyapplying the exploration, preparation, implementation, sustainment (EPIS)framework in the design and measurement of a large scale collaborativemulti-site study. Health Justice. 2018;6(1):9.

17. Layde PM, Christiansen AL, Peterson DJ, Guse CE, et al. A model to translateevidence-based interventions into community practice. Am J Public Health.2012;102(4):617–24.

18. Mueller M, Purnell TS, Mensah GA, Cooper LA. Reducing racial and ethnicdisparities in hypertension prevention and control: what will it take totranslate research into practice and policy? Am J Hypertens. 2015;28(6):699–716.

19. Lopez-Class M, Peprah E, Zhang X, Kaufmann PG, et al. A strategicframework for utilizing late-stage (T4) translation research to address healthinequities. Ethn Dis. 2016;26(3):387–94.

20. Graham PW, Kim MM, Clinton-Sherrod AM, Yaros A, et al. What is the role ofculture, diversity, and community engagement in transdisciplinarytranslational science? Transl Behav Med. 2016;6(1):115–24.

21. Ray R, Kelly N. Reaching new markets through transcreation: whentranslation just isn’t enough. Lowell: Common Sense Advisory, Inc; 2010.

22. Rivera YM, Velez H, Canales J, Jimenez JC, et al. When a common languageis not enough: transcreating cancer 101 for communities in Puerto Rico. JCancer Educ. 2016;31(4):776–83.

23. Nápoles AM, Santoyo-Olsson J, Stewart AL. Methods for translatingevidence-based behavioral interventions for health-disparity communities.Prev Chronic Dis. 2013;10:E193.

24. Nápoles AM, Santoyo-Olsson J, Stewart AL, Ortiz C, et al. Evaluating theimplementation of a translational peer-delivered stress managementprogram for Spanish-speaking Latina breast cancer survivors. J Cancer Educ.2017;

25. Kanaya AM, Santoyo-Olsson J, Gregorich S, Grossman M, et al. The live well,be well study: a community-based, translational lifestyle program to lowerdiabetes risk factors in ethnic minority and lower-socioeconomic statusadults. Am J Public Health. 2012;102(8):1551–8.

26. Stewart AL, Gillis D, Grossman M, Castrillo M, et al. Diffusing a research-based physical activity promotion program for seniors into diversecommunities: CHAMPS III. Prev Chronic Dis. 2006;3(2):A51.

27. Nápoles AM, Santoyo-Olsson J, Ortiz C, Gregorich S, et al. Randomizedcontrolled trial of Nuevo Amanecer: a peer-delivered stress managementintervention for Spanish-speaking Latinas with breast cancer. Clin Trials.2014;11(2):230–8.

28. Dickerson D, Baldwin JA, Belcourt A, Belone L, et al. Encompassing culturalcontexts within scientific research methodologies in the development ofhealth promotion interventions. Prev Sci. 2018;

29. Chambers DA, Norton WE. The Adaptome: advancing the science ofintervention adaptation. Am J Prev Med. 2016;51(4 Suppl 2):S124–31.

30. Ockene JK, Edgerton EA, Teutsch SM, Marion LN, et al. Integrating evidence-based clinical and community strategies to improve health. Am J Prev Med.2007;32(3):244–52.

31. Hill CV, Perez-Stable EJ, Anderson NA, Bernard MA. The National Institute onAging health disparities research framework. Ethn Dis. 2015;25(3):245–54.

32. Bandura A. Social foundations of thought and action: a social cognitivetheory. Englewood Cliffs: Prentice-Hall; 1986.

33. Delgadillo AT, Grossman M, Santoyo-Olsson J, Gallegos-Jackson E, et al.Description of an academic community partnership lifestyle program forlower income minority adults at risk for diabetes. Diabetes Educ. 2010;36(4):640–50.

34. Nápoles AM, Ortiz C, Santoyo-Olsson J, Stewart AL, et al. Nuevo Amanecer:results of a randomized controlled trial of a community-based, peer-delivered stress management intervention to improve quality of life inLatinas with breast cancer. Am J Public Health. 2015;105(Suppl 3):e55–63.

35. Kok G, Gottlieb NH, Peters GJ, Mullen PD, et al. A taxonomy of behaviourchange methods: an intervention mapping approach. Health Psychol Rev.2016;10(3):297–312.

36. Marks R, Allegrante JP, Lorig K. A review and synthesis of research evidencefor self-efficacy-enhancing interventions for reducing chronic disability:implications for health education practice (part II). Health Promot Pract.2005;6(2):148–56.

37. Barnett ML, Gonzalez A, Miranda J, Chavira DA, et al. Mobilizing communityhealth workers to address mental health disparities for underservedpopulations: a systematic review. Admin Pol Ment Health. 2018;45(2):195–211.

38. Ramchand R, Ahluwalia SC, Xenakis L, Apaydin E, et al. A systematic reviewof peer-supported interventions for health promotion and diseaseprevention. Prev Med. 2017;101:156–70.

39. Nápoles-Springer AM, Ortiz C, O'Brien H, Diaz-Mendez M. Developing aculturally competent peer support intervention for Spanish-speaking Latinaswith breast cancer. J Immigr Minor Health. 2009;11(4):268–80.

40. Lorig K, Laurent DD, Plant K, Krishnan E, et al. The components of actionplanning and their associations with behavior and health outcomes.Chronic Illn. 2014;10(1):50–9.

41. Fredriksen-Goldsen KI, Kim HJ, McKenzie GL, Krinsky L, et al. Plan of actionfor real-world translation of LGBTQ health and aging research. LGBT Health.2017;4(6):384–8.

42. Williams AS, Moore SM. Universal design of research: inclusion of personswith disabilities in mainstream biomedical studies. Sci Transl Med. 2011;3(82):82cm12.

43. Seligman HK, Wallace AS, DA DW, Schillinger D, et al. Facilitating behaviorchange with low-literacy patient education materials. Am J Health Behav.2007;31(Suppl 1):S69–78.

44. Stewart AL, Grossman M, Bera N, Gillis DE, et al. Multilevel perspectives ondiffusing a physical activity promotion program to reach diverse olderadults. J Aging Phys Act. 2006;14(3):270–87.

45. Burgio LD, Collins IB, Schmid B, Wharton T, et al. Translating the REACHcaregiver intervention for use by area agency on aging personnel: theREACH OUT program. Gerontologist. 2009;49(1):103–16.

46. Hawe P, Shiell A, Riley T. Complex interventions: how “out of control” can arandomised controlled trial be? BMJ. 2004;328(7455):1561–3.

47. Mejia A, Leijten P, Lachman JM, Parra-Cardona JR. Different strokes fordifferent folks? Contrasting approaches to cultural adaptation of parentinginterventions. Prev Sci. 2017;18(6):630–9.

48. Gonzales NA. Expanding the cultural adaptation framework for population-level impact. Prev Sci. 2017;18(6):689–93.

49. Nápoles-Springer AM, Stewart AL. Overview of qualitative methods inresearch with diverse populations. Making research reflect the population.Med Care. 2006;44(11 Suppl 3):S5–9.

50. Mercer SL, DeVinney BJ, Fine LJ, Green LW, et al. Study designs foreffectiveness and translation research: identifying trade-offs. Am J Prev Med.2007;33(2):139–54.

51. Henry D, Tolan P, Gorman-Smith D, Schoeny M. Alternatives to randomizedcontrol trial designs for community-based prevention evaluation. Prev Sci.2017;18(6):671–80.

52. Nápoles A, Cook E, Ginossar T, Knight KD, et al. Applying a conceptualframework to maximize the participation of diverse populations in cancerclinical trials. Adv Cancer Res. 2017;133:77–94.




53. Santoyo-Olsson J, Cabrera J, Freyre R, Grossman M, et al. An innovativemultiphased strategy to recruit underserved adults into a randomized trialof a community-based diabetes risk reduction program. Gerontologist. 2011;51(Suppl 1):S82–93.

54. Stewart AL, Nápoles-Springer AM. Advancing health disparities research: canwe afford to ignore measurement issues? Med Care. 2003;41(11):1207–20.

55. Krok-Schoen JL, Weier RC, Hohl SD, Thompson B, et al. Involvingcommunity health workers in the centers for population health and healthdisparities research projects: benefits and challenges. J Health Care PoorUnderserved. 2016;27(3):1252–66.

56. Krukowski RA, Lensing S, Love S, Prewitt TE, et al. Training of lay healtheducators to implement an evidence-based behavioral weight lossintervention in rural senior centers. Gerontologist. 2013;53(1):162–71.

57. Breitenstein SM, Gross D, Garvey CA, Hill C, et al. Implementation fidelity incommunity-based interventions. Res Nurs Health. 2010;33(2):164–73.

58. Bellg AJ, Borrelli B, Resnick B, Hecht J, et al. Enhancing treatment fidelity inhealth behavior change studies: best practices and recommendations fromthe NIH behavior change consortium. Health Psychol. 2004;23(5):443–51.

59. Allicock M, Campbell MK, Valle CG, Carr C, et al. Evaluating thedissemination of Body & Soul, an evidence-based fruit and vegetable intakeintervention: challenges for dissemination and implementation research. JNutr Educ Behav. 2012;44(6):530–8.

60. Ard JD, Carson TL, Shikany JM, Li Y, et al. Weight loss and improvedmetabolic outcomes amongst rural African American women in the deepsouth: six-month outcomes from a community-based randomized trial. JIntern Med. 2017;282(1):102–13.

61. Prezio EA, Pagan JA, Shuval K, Culica D. The community diabetes education(CoDE) program: cost-effectiveness and health outcomes. Am J Prev Med.2014;47(6):771–9.

62. Perez-Escamilla R, Damio G, Chhabra J, Fernandez ML, et al. Impact of acommunity health workers-led structured program on blood glucosecontrol among Latinos with type 2 diabetes: the DIALBEST trial. DiabetesCare. 2015;38(2):197–205.

63. Lopez L, Grant RW. Closing the gap: eliminating health care disparitiesamong Latinos with diabetes using health information technology toolsand patient navigators. J Diabetes Sci Technol. 2012;6(1):169–76.

64. Albritton E. Health equity and health system transformation: how states canfund community health workers through Medicaid to improve people’shealth, decrease costs, and reduce disparities. Families USA. 2016; http://familiesusa.org/sites/default/files/product_documents/HE_HST_Community_Health_Workers_Brief_v4.pdf. Accessed 14 Aug 2018.

65. Kent EE, Mitchell SA, Castro KM, DeWalt DA, et al. Cancer care deliveryresearch: building the evidence base to support practice change incommunity oncology. J Clin Oncol. 2015;33(24):2705–11.


http://familiesusa.org/sites/default/files/product_documents/HE_HST_Community_Health_Workers_Brief_v4.pdf



transcreation: an implementation science framework for

Documents