A Community-Responsive Adaptation to Reach and Engage Latino Families Affected by Maternal Depression

Abstract

As family researchers and practitioners seek to improve the quality and accessibility of mental health services for immigrant families, they have turned to culturally adapted interventions. Although many advancements have been made in adapting interventions for such families, we have yet to understand how the adaptation can ensure that the intervention is reaching families identified to be in greatest need within a local system of care and community. We argue that reaching, engaging, and understanding the needs of families entails a collaborative approach with multiple community partners to ensure that adaptations to intervention content and delivery are responsive to the sociocultural trajectory of families within a community. We describe a cultural adaptation framework that is responsive to the unique opportunities and challenges of identifying and recruiting vulnerable families through community partnerships, and of addressing the needs of families by incorporating multiple community perspectives. Specifically, we apply these principles to the cultural adaptation of an intervention originally developed for low-income African American and White families facing maternal depression. The new intervention, Fortalezas Familiares (Family Strengths), was targeted to Latino immigrant families whose mothers were in treatment for depression in mental health and primary care clinics. We conclude with key recommendations and directions for how family researchers and practitioners can design the cultural adaptation of interventions to be responsive to the practices, preferences, and needs of underserved communities, including families and service providers.

Context of Latino Immigrant Mental Health

For family-focused interventions to reach, engage, and address the mental health needs of Latino immigrant families in the United States, the interventions must be responsive to the families' life trajectories and contexts (Castro, Barrera, & Steiker, 2010). Although many immigrants come to this country with much hope and enthusiasm, the stressors associated with immigration, settlement, and acculturation, coupled with absence of familiar supports, have contributed to increasing vulnerability for depression (Valdez, Abegglen, & Hauser, 2013). Depression is prevalent among women in the general population, but Latinas have been shown to have an earlier and more persistent course of depressive illness than their non-Latina counterparts (Breslau, Kendler, Su, Gaxiola-Aguilar, & Kessler, 2005). This tendency is especially so among Latinas who are less educated, of immigrant status, and primarily Spanish speaking (Myers et al., 2002).

Latinas' high vulnerability to depression poses serious consequences for children because of the potential for children to experience long-standing impairment associated with their mother's depression (Goodman et al. 2011; Timko et al., 2009). Yet, the majority of evidence based family-focused interventions for maternal depression have been developed for the general population and have not been widely disseminated to Latinos (Cardemil, Kim, Pinedo, & Miller, 2005). Latinos are the largest minority group in the United States, yet they continue to be one of the most underrepresented ethnic groups in mental health services and intervention research (Stacciarini, 2009). Some reasons for Latino families' low levels of mental health service utilization include practical barriers, such as lack of insurance, as well as cultural barriers, such as stigma of mental health (Valdez, Dvorscek, Budge, & Esmond, 2011). Moreover, mental health interventions have failed to be grounded in the linguistic and cultural experience of Latinos (Bernal & Sáez-Santiago, 2006; Cardemil et al., 2005). This failure is alarming because clients' worldview of illness, client-therapist characteristics and relationship, and organizational factors, such as cost and availability of culturally grounded mental health services, appear to improve mental health outcomes (Benish, Quintana, & Wampold, 2011; Bernal & Sáez-Santiago, 2006).

As family practitioners and researchers are becoming increasingly interested in addressing the mental health of minority populations, adaptations must be developed to improve the cultural and linguistic relevance of interventions. We propose a cultural adaptation needs to be responsive to the local community context, including an understanding of which families need the intervention the most, and how and where to reach them to maximize the potential impact of the intervention. This step is particularly important for families experiencing clinical distress because they may be more difficult to reach through conventional recruitment methods. Reaching families is followed by engaging them and other community stakeholders to inform the adaptation, and addressing family needs via a culturally grounded intervention.

In this article, we describe the process of culturally adapting an intervention originally developed for low-income White and African-American women with depression and their families into an intervention that can be used with Latino immigrant families. We begin by briefly presenting the original intervention, Keeping Families Strong, by describing its theoretical foundation, format, and content to illustrate our starting point. We then provide a rationale for culturally adapting this intervention for Latino families. Next, we provide the cultural adaptation models that guided our work along with the unique processes we used to respond to our local context. We also address concerns raised in the literature about the utility and justification of cultural adaptations. Moreover, we describe in detail the systematic approach we took to adapting the intervention as a Latino family-based program, Fortalezas Familiares (Family Strengths). Finally, we offer implications for interventions with Latino families and for cultural adaptations.

Foundation for a Family Intervention

Keeping Families Strong (KFS) was developed in Baltimore by the first author and clinical researchers from a local university, in partnership with administrators and practitioners from local outpatient mental health clinics. Clinic partners described a need for family-focused services for adult women in individual treatment for depression, many of whom were low-income African American and White single mothers. These groups were identified as key targets for intervention because although adult women were in treatment, they reported high levels of distress and unmet needs in the family, and no family services were available to address these needs. Thus, clinic partners expressed an interest in building their capacity to offer family-focused services where adult and child practitioners would work together to coordinate and deliver care for the whole family. With input from practitioners, we spent 12 months developing a manual for a family-focused intervention that could be adopted by clinical settings. Practitioners referred their adult female clients and their families to KFS, and co-facilitated the program with clinical researchers in clinics to build trust in locations convenient for and familiar to the families (Valdez, Mills, Barrueco, Leis, & Riley, 2011).

KFS consisted of 10-weekly plus two monthly booster meetings for parents and youth (Riley et al., 2008; Valdez, Mills, et al., 2011). Parent and youth meetings were initially separate, but concurrent, with later meetings focused on the whole family to facilitate family members' shared understanding of depression and family life, and collective identification of family goals (Riley et al., 2008). KFS met its goals, with youth and mothers participating in KFS reporting improvements in coping, behavioral and emotional adjustment, maternal warmth and acceptance, family interactions, and family support (Valdez, Mills, et al., 2011). KFS drew from family systems, narrative, and cognitive-behavioral approaches to address the mechanisms found in the literature to influence child and family well-being in the context of maternal depression (Riley et al., 2008). Two evidence-based interventions served as models: Prevention Intervention Program for families with maternal depression (Beardslee, Gladstone, Wright, & Cooper, 2003), and New Beginnings Program for families experiencing parental loss and divorce (Sandler et al., 2003). The developers of those interventions collaborated with clinical researchers in designing the clinical components of KFS.

Rationale for a Cultural Adaptation

Although KFS showed promising results with low-income families, it did not however account for the language, family, and sociocultural context of Latino immigrant families in the Midwest (Valdez, Padilla, McArdell Moore, & Magaña, 2013). For example, KFS was developed for and evaluated with English-speaking families, most of whom were lifelong residents of a large urban city. In contrast, Latino immigrant families increasingly settled in small cities that have been traditionally White (Falicov, 2014). Latinos' recent settlement means that not only are they less incorporated into mainstream society than long-term residents, but also that they may be more difficult to reach and engage due to their limited access to community and mental health services. Moreover, the majority of immigrant parents have limited English proficiency, so at a minimum, the KFS intervention materials would need to be translated linguistically to Spanish and facilitators would need to be fluent in Spanish to engage parents in the intervention.

Moreover, for an intervention to address the needs of Latino families it would need to broaden its language about depression to include culturally acceptable terms, explanations, and typical and atypical manifestation of depression that have been observed in Latino populations (D'Angelo et al., 2009). The intervention would also need to consider common experiences and trajectories of families within the local community that may be as or even more influential than ethnicity (Castro et al., 2010), such as immigration status and acculturative stress (Cardemil et al., 2005). For example, many Latina women and children in the United States experience trauma prior to, during, and post-migration (Falicov, 2014). In addition, immigration often is accompanied by family separation, transnational parenting, and for some, adjustment to parent-child reunification (Miranda, Siddique, Der-Martirosian, & Belin, 2005). Thus, we would need to address the interpersonal losses and life cycle challenges associated with shifting family structures and boundaries.

Immigrant parents experience additional challenges associated with settlement in a new community including managing acculturative stress and pressures to assimilate, dealing with experiences of discrimination, and adjusting to new social and community ties (Falicov, 2014). For example, Latino parents have difficulty understanding the bicultural world experienced by their children when moving between the home culture and the school culture (Smokowski & Bacallao, 2011). The difficulties and pressures of acculturation are experienced differently by the parent and youth, potentially leading to misunderstanding and distancing within the family. Many Latino families also live under the fear of potential deportation for undocumented family members (Dreby, 2015; Valdez, Lewis Valentine, & Padilla, 2013). Parents and children fear being separated from one another and parents face difficult decisions about returning together or leaving children behind if parents were to be deported (Dreby, 2015). These background experiences render the initial symptoms of depression as understandable to family members. However, as symptoms persist and worsen, family experiences with maternal depression heighten family stress and negative family interactions that foster confusion and resentment. An intervention for Latino families would need to interlace these sociocultural pressures and considerations into both the parent and youth programs in order to strengthen the cultural relevance of the adaptation.

Additionally, children in these families may face a challenging integration into U.S. schools, leaving them feeling like they do not belong in educational institutions (Valenzuela, 2010). Latino children often feel isolated from their peers, physically when separated into English learning classes, and socially when the rest of the student population is not predominantly Latino (Valenzuela, 2010). Many undocumented children also experience microaggressions or discriminatory remarks in schools, are not prepared to cope with or manage these experiences in the moment, and feel unsupported by teachers (Benner & Graham, 2011). These school experiences exacerbate Latino parents' stress and worry since many immigrated in part to provide access to educational resources and opportunities for their children (Valdez, Lewis Valentine, & Padilla, 2013). Thus, a new intervention would need to prepare parents and youth to explore and affirm their identities in spite of hostile and dehumanizing societal messages about Latino immigrants, so they can foster resilience and a sense of ethnic pride (Smokowski & Bacallao, 2011).

Finally, there is a high presence of fathers living in Latino immigrant households, particularly of Mexican origin (Cabrera & Bradley, 2012; Falicov, 2014). KFS did not address father and couple concerns because the program mostly served single mothers. Therefore, an adapted intervention would need to address how fathers cope with and support the family during the mother's depression (Valdez, Martinez, & Hoogasian, 2017). Furthermore, the intervention would need to be flexible to accommodate extended family members such as grandparents, godparents, adult children, and other trusted adults who are commonly part of caregiving in these families (Falicov, 2014). Including other caregivers not only would strengthen the intervention's ability to address the needs of immigrant families, it would also promise to facilitate engagement via increased family involvement and support.

In summary, the sociocultural challenges facing Latino immigrant families warranted adaptations from KFS. Our ultimate goal in conducting an adaptation was to reach vulnerable Latino immigrant families and enhance the intervention's effectiveness and engagement and acceptability to families (Valdez, Padilla, et al., 2013). Research shows that adapted interventions improve outcomes for Latino families when they address relevant cultural and contextual influences on family life (Benish et al., 2011; Cardemil et al., 2005; D'Angelo et al., 2009; Hall, Ibaraki, Huang, Marti & Stice, 2016; Parra Cardona et al., 2012).

Cultural Adaptation Models

Notable efforts in previous decades to account for Latino individuals' experience of culture within the therapeutic context set the foundation for contemporary cultural adaptation frameworks (see Szapocznik, Santisteban, Kurtines, Perez-Vidal, & Hervis, 1984). One of the earliest and, perhaps, most widely accepted frameworks is the ecological validity model (Bernal, Bonilla, & Bellido, 1995), which proposes eight dimensions that must be incorporated into any culturally sensitive treatment: language, persons, metaphors, content, concepts, goals, methods, and context. These dimensions highlight the need for surface-level and deep-level adaptations (Resnicow, Baranowski, Ahluwalia, & Braithwaite, 1999). A surface-level adaptation includes changes in intervention delivery to fit Latino families, such as using clients' language or ethnic symbols in intervention materials. Meanwhile, a deeper adaptation reflects content changes to fit Latino family values, experiences, and needs more fully, such as addressing acculturative stress (Castro et al., 2010). This model has proven efficacious for interventions with Puerto Ricans (Bernal et al., 1995) and has served as a catalyst for the adaptation of interventions with other Latino groups (Cardemil et al., 2005; Domenech Rodríguez & Wieling, 2004).

Recent adaptation efforts have expanded the ecological validity model to include community engagement and stakeholder input in the development of measures and materials. For example, in the cultural adaptation process model (CAPM; Domenech Rodríguez & Wieling, 2004), the adaptation involves three phases: In Phase 1, the community needs and scientific integrity are weighed to decide who will be involved in the adaptation. Phase 2 involves selecting and adapting the evaluation measures, gathering data, and revising and adapting the intervention. Phase 3 integrates the data gathered to establish a “new” intervention that can subsequently be tested at a larger scale for the specific adaptation population (Domenech Rodríguez & Wieling, 2004; Parra Cardona et al., 2012, 2015). These recent models recognize the need for multiple stakeholder input and engagement in the adaptation.

A Community-Responsive Cultural Adaptation

In this section, we describe our adaptation efforts of an intervention for Latino families affected by maternal depression in a medium sized city in the Midwest. When we began our adaptation we turned to models such as CAPM to guide our process. Many of our activities were consistent with and influenced by CAPM, such as engaging various stakeholders, revising intervention content and delivery, and piloting and refining the new intervention through a series of iterations. CAPM was also a good starting point for our project given its focus on Latinos. However, we departed from CAPM in that it was based on a parenting program for non-clinical populations. Meanwhile, our clinical intervention was designed for high-risk families, and specifically for families of women who were in treatment for depression. Thus, our adaptation required us to give a prominent role to reaching a multi-stressed clinical population within healthcare settings, which involved identifying and recruiting vulnerable families, as well as earning the trust of the professional community, efforts that were less prominent in the original dissemination of the CAPM.

In our goal to be responsive to our local community, our process begins with planning the adaptation, with consideration to reaching, engaging, and addressing needs of families and clinics, and finally, implementing the intervention. These tasks are focused on understanding the needs, practices, and preferences of families so that researchers and practitioners are (a) reaching families who need the intervention, (b) addressing the family and sociocultural processes influencing well-being, (c) accounting for health practices and preferences for participating in interventions, (d) and building the capacity of local clinics to meet the needs of their clients and patients. An evaluation of this adaptation process and outcomes then leads to refinements in any of these tasks, reflecting an iterative process. We now describe these tasks in greater detail, which led to the development of Fortalezas Familiares.

Planning for the Adaptation

Careful planning and preparation laid a solid foundation for the adaptation, and entailed the formation of a research team, a steering committee, and a partnership with community collaborators, all with distinct roles in the adaptation. Planning also entailed training and resource allocation.

Formation of a research team

It was critical to select research team members who deeply understood the context and culture of Latino immigrant families. Team members were mainly graduate students in a counseling psychology department with a training emphasis in social justice and diversity. The majority of team members were of Latino background and spoke Spanish as their native language, and many came from immigrant families. In addition, these team members were dedicated to and experienced in research and practice with Latino families. The cultural representativeness of the team was an important surface level modification that aimed to facilitate trust with the families, who often lack access to bilingual and bicultural mental health providers (Valdez, Dvorscek, et al., 2011). Moreover, a few team members came from related fields of study (i.e., social work, school psychology, public health, ethnic studies), which expanded team conceptualization of families. To anchor the work of the research team in the local context, our project coordinator was a social work professional who was concurrently employed part-time as a mental health clinician at a community health center. Her clinical experience enriched the expertise of the team and facilitated the credibility of our project among our clinic collaborators. She was also Latina and a native Spanish speaker which enhanced trust in our initial contact with families. The coordinator participated in weekly research team meetings and was integral to the planning of the project.

The role of the research team was to carry out the day-to-day activities and operations related to planning the intervention. Initially, the research team conducted a review of the literature on Latino mental health, maternal depression, immigrant families, and sociocultural processes. This literature pointed to critical areas of intervention content and delivery. For example, we learned about the importance of identity development of immigrant youth in our target age range. Next, the team searched for published interventions with Latino families, specifically those targeting depression. This literature search allowed us to consider not only how other interventions were addressing the needs of Latino immigrant families, but also how our own intervention filled important areas not already present in the literature. For example, the interventions we found in the literature were largely focused on the parent with depression, and even the interventions that included children focused on one child only with sporadic meetings relative to the parent component. Thus, ours remains the only intervention to have an equal number of parent and youth meetings and to include all children ages 9-17. In addition, the literature on cultural adaptation frameworks and on dissemination and implementation methods provided a roadmap to organize our own adaptation. We learned that our cultural adaptation should not only modify intervention content and delivery, but also consider strategies to engage and build mental and physical health practice in the community (Cabassa & Baumann, 2013; Castro et al., 2010).

Formation of a steering committee

To supplement the knowledge gained from the team's literature review, we also assembled a steering committee that included a local senior faculty collaborator, as well as national consultants in Latino intervention research. We met with the senior faculty collaborator monthly, and with consultants as needed over the phone and twice in person. The role of the steering committee was to advise on the adaptation, that is, to suggest additional literature, review our intervention materials, and illustrate strategies from their own work that we, the research team, could incorporate into our intervention.

From the start, we consulted with our steering committee as to whether the intervention needed to be adapted at all. We considered whether there was sufficient evidence documenting the necessity for such an adaptation, and whether the adaptation could be made with rigor (Castro et al., 2010). We followed published guidelines which suggest that interventions should be adapted judiciously when (a) risk or resilience factors are unique to the target population, (b) symptom presentation may differ for a certain group, and (c) family engagement in the intervention would be unlikely unless accommodations were made to the language and delivery of the intervention (Lau, 2006). We discussed our target population's risk and resilience factors: culturally specific experiences and contextual stressors associated with immigration, poverty and associated trauma, acculturative stress, and discrimination. Resilience factors included cultural values of familismo and colectivismo, which bind families together and strengthen families' affiliations with other families, and a particularly high father involvement in our local immigrant families. These factors could have positive implications for maternal depression and family coping strategies (Valdez, Abbeglen, & Hauser, 2013). We also agreed the frequent physical manifestation and interpretation of depression in Latinos that warrants attention in the intervention (Pincay & Guarnaccia, 2007). Finally, we determined that at a minimum, the intervention needed to be conducted in Spanish for parents, and in English and Spanish for youth, for families to participate in and benefit from the program. Thus, we concluded that a cultural adaptation of KFS was justified.

With respect to rigor of the adaptation, we looked to existing frameworks for guidance and engaged in systematic development and monitoring of our adaptation. For example, we documented any and all modifications to the intervention content and delivery, and closely tracked unplanned deviations from the major components of the manual. As noted, we drew from multiple perspectives in our decision-making and evaluated each implementation to ensure that the intervention was meeting its clinical benchmarks and participants were benefitting from the intervention (Castro et al., 2010).

Engagement of Collaborators

Rounding out the contributions of the research team and the steering committee were our clinic collaborators. We identified mental health clinics and primary care clinics with strong presence in the Latino community and contacted leaders within those settings. Depending on the site, leaders were administrators (e.g., medical directors) and more commonly, clinicians. These were individuals who had both clinical experience and interest with Latinos and who had enough authority and seniority within the organization to commit to a partnership. These collaborators were well-known within the lay community and were respected by other professionals. In the end, our primary partners were a large mental health agency, a federally qualified health clinic with a behavioral health unit, and a faith-based social service agency providing counseling among other services. As described later, these collaborators guided our understanding of community needs and connected us to participants.

Training

To ensure the cultural adaptation was systematic and judicious, team members underwent training at different levels. First, to retain KFS's clinical components, team members became versed in the KFS manuals, and the theoretical and empirical literature on maternal depression that guided the intervention. Content adaptation was carefully decided through a process of team consensus. Second, the project leader, who is a licensed professional psychologist, and who co-developed and implemented the original KFS intervention, trained team members by reviewing with them the manuals and role-playing intervention components. Third, prior to working with families, team members engaged in a process to enhance cultural competence by discussing case scenarios as a group, applying family systems concepts to these scenarios, role-playing program components, and engaging in intentional self-reflection via group discussion and journaling. During retreats, team members would reflect upon and share why this work was meaningful to them, and what personal values and experiences it reflected. Fourth, the project leader and an experienced clinician from the community co-facilitated the first iterations of the program with less experienced team members. Matching facilitators in this way allowed experienced facilitators to meet the complex needs of families while supporting the development of skills of the less experienced facilitators. In later iterations, these experienced facilitators moved to a supervisory role, where they watched the recordings of meetings led by other facilitators, and provided feedback and fidelity checks between meetings. Finally, the project leader oversaw these activities under the guidance of the steering committee and of a cadre of experts on cultural adaptation who mentored her as part of a multi-year conference on developing and adapting interventions for Latino children, youth and families.

Resource allocation

While these preparatory activities took place, we secured funding to translate the KFS manuals and supplemental materials to Spanish. Not only did the manuals have to be translated word for word, they had to capture the meanings, experiences, and regional language styles and literacy levels of our target population, as recommended in the literature (see Peña, 2007). Thus, we secured funding to hire a professional bilingual and bicultural translator.

Reaching Families

Our ultimate goal was to reach families, and to do so, we needed to establish which were most vulnerable, which gaps in the healthcare system our intervention would fill, and how to best reach these families given the community needs and resources. Because many existing adaptation models like CAPM were based on non-clinical populations, we knew that reaching families would be more difficult with our clinical intervention, and thus more time would have to be devoted upfront to stakeholder engagement. To begin, we established relationships with key community leaders, including leaders of nonprofit organizations, mental health providers, and primary care physicians. Many of these leaders were collaborators from our partner clinics.

Relationship building

Acknowledging the historical distrust between researchers and underrepresented communities (Curran, Mukherjee, Allee, & Owen, 2008), we spent the first two years engaging multiple community stakeholders in the conceptualization and planning of the adaptation. The project leader sought guidance from academic colleagues established in the Latino community. In addition, she partnered with key local health and mental health providers to make community connections. Multiple meetings with these professionals helped the leader and the research team understand local Latino families' need and readiness for an intervention.

Ultimately, as these professionals were gatekeepers for families, these meetings took place in the community and over an extended period of time so that our commitment could become evident. Collaborating on smaller projects, offering trainings to clinics, and engaging in family outreach via radio and in-person contact at health fairs, were all examples of how the team earned the trust of the community.

Determining targets and needs

To understand the needs of families, we first had to determine who our target families would be. In meetings with mental health providers, physicians, community leaders, and academic colleagues, we learned that a sizeable proportion of Latino clients in mental health clinics were adult women with depression, and that most of these were mothers with children living in the home. In spite of this need, there was a scarcity of culturally grounded and family-focused mental health services. In particular, a shortage of bilingual and bicultural family and child therapists often means parents would be seen in one clinic and youth referred to another. This division of services diminishes a family's ability to work through their concerns jointly, and contributes to uncoordinated and culturally incongruent care for youth. Moreover, we learned that many families seek care through primary care providers rather than traditional outpatient mental health settings. Thus, it became clear that the families we needed to target were those whose mothers were seeking depression services in mental health or primary care settings, but for whom the mental health system was failing to address their needs as a family. Because of mental health stigma, and potential difficulty in reaching the families of these women, we also decided the best way to reach a family was through these clinics that already had a relationship with the mother.

Engaging Families

An important task in adapting an intervention is to consider families' motivations for and preferences in participating in interventions. In contrast to the recruitment practices in the original CAPM, advertising Fortalezas Familiares through flyers or even the radio would not be sufficient to reach out to highly stressed families like ours. We thus partnered with our mental health and primary care clinic collaborators serving families to conduct recruitment. Together, we designed the recruitment protocol so Latina mothers would first hear about the Fortalezas Familiares program from their mental health or physical health providers, whom they knew well and trusted. A challenge in this collaboration was involving providers in our university's institutional review procedures, which required all individuals involved in recruitment to complete human subjects training. We explained the value of this training for protecting families from potential unethical practices, which providers appreciated, and offered a summary with simplified explanations and tips for approaching each training module. Easing the burden of this requirement allowed us to conform to ethical guidelines and acknowledge the challenges faced by providers to participate in research, a common balance in community-engaged cultural adaptations (Baumann, Domenech Rodríguez, & Parra-Cardona, 2011).

Next, research team members called mothers that provided written permission to their provider, and explained the purpose and logistics of the program. During this phone call, we conducted a brief screening to ensure clinical eligibility of the mothers. After eligibility was met, we scheduled a home visit to meet the whole family, provide an opportunity for families to get to know us better and feel comfortable with us, and to answer questions about participation in the program. We were particularly interested in engaging fathers during these visits, as they play an important gatekeeping role within the family (Behnke, Taylor, & Parra-Cardona, 2008; Falicov, 2014). Only at the end of a home visit did we ascertain each family's commitment to participate and to complete informed consent. We tailored this procedure to the home in response to focus groups with families and our earlier interviews with healthcare providers, where stigma arose as a potential barrier to participation. For this reason, we also chose not to complete our pre-intervention assessment during this visit to avoid burdening families before their first program meeting. Thus, we asked them to arrive an hour earlier to the first program meeting to complete these assessments.

Addressing the Needs of Families

Community and stakeholder perspectives

As we prepared to adapt KFS, we considered the voices and experiences of local community members. We conducted separate focus groups with 14 Latino parents and 10 youth at a community organization, to understand the challenges they faced within their families. Although these persons were not recruited from a clinical setting, many parents described feeling isolated, stressed, and even depressed, and they expressed concern about their children's well-being and adjustment in the United States. Parents specifically reported, “I want to know how to reach my children,” and “I feel more and more disconnected from [my children's] lives.” Parents and youth identified language as a barrier to family wellbeing, as in this quote, “I want to tell them what I'm going through but I can't find the words to express them in a way [they] can understand.” As these quotes illustrate, parents and youth were particularly concerned with the acculturative gap that they increasingly experienced. Youth also shared their concerns about fitting in at school and managing family demands. Focus groups with families allowed us to hear first-person accounts of the stressors affecting families, and the strategies parents and youth seek to build resilience. We also sought input from families as part of community workshops and presentations we led on the topic of family communication and family stress. Nearly 60 Latino parents participated, echoing similar concerns about acculturative stress in the home. It became apparent that acculturative stress needed to be incorporated throughout the sessions. A literature review suggested strategies for addressing cultural content, such as psychodrama, for facilitating families' tangible and experiential construction and meaning making of their varied cultural experiences.

Separate interviews with mental health and primary care providers corroborated that depression is common in the Latino community, and that many Latinas in treatment come with longstanding and unresolved trauma histories. Informally, we had multiple conversations with our three clinic collaborators, all of whom were mental health counselors, psychologists, or social workers. Formally, we also interviewed 12 primary care physicians from our health clinic site, which had a Latino patient population of at least 40%, about the pressing needs of their Latino patients (Valdez, Dvorscek, et al., 2011). Combined, providers perceived that acculturative stress in Latino families compounded their client and patients' depression, particularly as adolescents seek more autonomy and parents attempt to remain connected to their children. Moreover, they described how immigration climate increased fear, anxiety, and isolation that constrained Latinos' utilization of treatment and access to other relevant social services. These multiple perspectives were invaluable not only in suggesting surface-level modifications, such as conveying family strengths in the name of the intervention (Fortalezas Familiares), but also deep-level content for the intervention, as suggested by other cultural adaptation models (Bernal & Sáez-Santiago, 2006; Castro et al., 2010) . We also sought these perspectives from program family graduates.

Intervention modifications

Cultural adaptations must attend to the balance between intervention fidelity and cultural fit (Castro et al., 2010; Parra-Cardona et al., 2015). With respect to fidelity, we preserved the core clinical elements of the original intervention, KFS, which were designed to address the pathways through which maternal depression affects family functioning and personal well-being. Theoretically, both interventions were grounded in a developmental psychopathology framework (Cummings, Davies, & Campbell, 2000) as well as family systems theory (Carr, 2015). The developmental psychopathology framework targeted mechanisms of risk and resilience that interact during critical periods of development, such as adolescence. Family systems theory targeted transactional family processes that increase vulnerability to and are exacerbated by maternal depression. Therapeutically, the original and the new interventions used interpersonal and group processes, cognitive-behavioral, and narrative models to integrate psychoeducation with meaning-making, self-reflection, and life stories (Riley et al., 2008; Valdez, Abegglen, & Hauser, 2013). Both interventions followed three major themes and phases: (a) understanding depression and its impact on family, (b) resilience and skill-building around family interactions and communication, and (c) integration of a shared family narrative.

While ensuring fidelity, our preparation and engagement of community stakeholders suggested necessary adaptation to the content (Domenech Rodríguez & Bernal, 2012). Many dimensions from Bernal and Sáez-Santiago's (2006) culturally centered adaptation model informed this task. In particular, we revised the intervention to (a) acknowledge content related to Latino immigrant experiences, such as family separation and loss, (b) promote concepts of parenting consistent with Latino culture, (c) address the community context of families, (d) target family-defined goals and assets, and (e) enhance learning through familiar symbols and metaphors. Moreover, Fortalezas Familiares included a deeper exploration of the ecological roots of depression, such as cultural beliefs and explanations, sociocultural experiences. In doing so, the intervention would both normalize and externalize the mother's symptoms and allow family members to identify how they could support each other's needs. Moreover, because of the central role of fathers in Latino immigrant families (Behnke et al., 2008; Cabrera & Bradley, 2012), we included content addressing couple relationship dynamics such as communication, problem-solving, conflict resolution, and co-parenting. This focus on the marital relationship and on co-parenting also illuminated the emotional burden many fathers experienced as they supported their partner with depression and their children (Valdez et al., 2017). After three trials of the intervention and repeated interviews with fathers participating in the intervention, this need became so apparent to us that we revised the program even more to dedicate one separate session to the marital, parenting, and emotional needs and experiences of fathers within these families.

In addition to these modifications to the internal family experience, we were guided by social ecological theory (Bronfenbrenner & Morris, 2006) to more fully capture the proximal and distal contexts shaping family life in the United States and the experience of family stress and depression. We focused on these contexts by addressing: (a) how family members understand and manage acculturative and immigration stressors, (b) parental involvement of children's activities outside the home, (c) transnational caregiving and loss, and (d) cultural assets, such as traditions, extended family support, and ethnic socialization (Valdez, Abegglen, & Hauser, 2013). We sought expertise in these areas from our steering committee, who suggested specific sociocultural content and delivery approaches. For example, to enhance communication between parents and youth about acculturative stress, we asked participants to craft scenarios within their group showing what it is like to be Latino in the United States from the youth and parent perspectives. These scenarios were then acted out in a larger group with parents and youth combined, followed by a conversation about what youth and parents learned from each other's experience. These action-oriented activities, or methods, the term used by Bernal and Sáez-Santiago, are aimed at enhancing perspective-taking and expression of meaning, particularly when these experiences are difficult to process via group discussion alone (Smokowski & Bacallao, 2011). These activities have been linked with collective efficacy (Nieto, 2010).

Finally, bilingual mental health providers delivered Fortalezas Familiares in Spanish to Latino parents. The youth groups were facilitated in English and/or Spanish per the youth's preference, and often participants used both languages to explain complex emotions and family experiences (D'Angelo et al., 2009). The program occurred during the evenings so parents and youth could be present. Every meeting started with a meal to model the importance of stable family routines and familismo (family unity and orientation), facilitate personalismo (personal connection among families), and culturally grounded relationship-building among families and between families and program facilitators. Providing a meal was part of the original KFS intervention, but care was taken with Fortalezas Familiares that the meal reflect parent and youth cultural preferences. Although this meal was a surface-level adaptation, it was critical to family receptivity and engagement (Resnicow et al., 1999).

Implementing the Intervention

In preparation for a large scale effectiveness trial of the intervention, we gathered implementation information (Curran, Bauer, Mittman, Pyne, & Stetler, 2012). We assessed the feasibility of recruiting and retaining families and their perceived value of the intervention. We conducted focus groups and administered pre- and post-satisfaction measures to examine the intervention components that were deemed the most and least helpful. In KFS, all trials were conducted in outpatient mental health clinics and co-delivered by local clinicians. This allowed us to refine delivery and training components to fit within the culture and practice of these settings (Valdez, Mills, et al., 2011). We intended to similarly implement Fortalezas Familiares in clinical settings but input from local mental health providers and families suggested stigma associated with mental health settings among Latino families might deter their participation. Therefore, we offered Fortalezas Familiares in community centers within the Latino community, while maintaining clinics' involvement in the recruitment of families.

Evaluating and Refining

Evaluations of Fortalezas Familiares have been published and described in greater detail elsewhere (see Valdez, Padilla, et al., 2013) and show it to be a promising intervention on multiple levels. In terms of recruitment and retention, a measure of feasibility, we received clinic referrals of 25 women, of whom 17 were eligible and invited to participate in a pilot evaluation. Of these, 13 attended more than 90% of the meetings (Valdez, Padilla, et al., 2013). Notably, almost 70% of fathers completed the program. We retained all clinical components of the original KFS intervention and followed strict fidelity checks of every meeting (Valdez, Padilla, et al., 2013).

We assessed participating families' perceived value of Fortalezas Familiares through focus groups and questionnaires. Families valued the intervention, which they attributed to helping them feel more united, better understand each other's needs and feelings, communicate more effectively, and enjoy family interactions. Participants attributed the group process to normalizing their experience and they reported continued contact between families after the intervention. Families requested the intervention be longer to better address adolescent risk (e.g. drugs, early sex initiation) and father struggles (e.g. depression, substance use; Valdez, Padilla, et al., 2013). From this feedback we have added two more meetings, one of which is dedicated to the struggles that fathers face as they cope with their partner's depression and support the family.

Lastly, a preliminary clinical evaluation found that mothers reported large improvements in overall functioning with decreases in depression, anxiety, and somatization (Valdez, Padilla, et al., 2013). Mothers also reported improvements in parental involvement, acceptance, warmth, and consistency. Fathers reported various levels of improvement in their own health and functioning, reporting a large decrease in somatization, anxiety, a small decrease in depression, and a large increase in global functioning. Mothers and fathers also reported increased support from family and friends. Children reported many improvements in their support-seeking behaviors, conduct problems, emotional symptoms, prosocial behavior, and a small decrease in total conflict behavior with their mothers. Moderate improvements were observed in children's perception of family functioning and parental rejection (Valdez, Padilla, et al., 2013).

An area of continued concern was school functioning. Although the intervention did not specifically address school life in our initial pilot trials, we anticipated that improvements in family life would be associated with improvements in school life. We conducted focus groups with former intervention youth participants, who discussed the difficulties they experienced at school related to racism. This feedback prompted us to address racism through racial/ethnic socialization and identity development of youth.

Discussion

A major contemporary challenge in family interventions has been to develop evidence-based interventions that are culturally responsive to ethnic minority communities in the United States (Bernal & Sáez-Santiago, 2006; Lau, 2006). Consistent with Castro and colleagues (2010), we argue that culturally adapting an intervention based on ethnicity alone is not enough in responding to the varying needs and characteristics of local communities. In this article, we describe how we culturally adapted an intervention incorporating families' trajectories and experiences within their local contexts, in reaching, engaging, and addressing the needs of Latino families affected by maternal depression. We first partnered with local service providers to understand the local landscape for families—the composition and pressing needs of the Latino community in our area, as well as the types of services needed and not available to families. Second, through meetings, focus groups, and interviews with these providers as well as with members from the lay community, including families, we learned about the stressors and pressures families were managing, including immigration and generational stress, settlement challenges in a predominantly White community, and a lack of family-focused resources within the professional healthcare sector to serve these families. Specifically, our adaptation also responded to what local providers and community leaders identified to be the greatest need in their practice. Moreover, feedback from providers and community members suggested we reach women with depression through our partner clinics, but then engage families in less stigmatizing settings, such as their homes, and later in community organizations. Consistent with this feedback, we modified our recruitment and implementation procedures so that we made initial contact through the clinic, but subsequent recruitment of the family happened through home visits, and implementation of the intervention took place in a general community setting. A participatory approach with families and local service providers has been found to enhance local ownership and commitment to the program, which in turn has been found to increase adoption and sustainability (Curran et al., 2008). In line with this practice, we are developing an implementation package for local providers to deliver the intervention in their settings to augment its adoption and sustainability (Lachman et al., 2016).

Concurrently, as part of our goal to engage and address the needs of families, we tailored the intervention content and delivery while retaining the core clinical elements of the intervention. Our adaptation responded to the large percentage of immigrant families with a father in the home by interweaving content specific and relevant to fathers, such as couples' processes of communication and conflict resolution, caregiver burden and coping, and co-parenting support. Including fathers in the intervention was important, not only for their support of their partner's recovery from depression, but for the stability they can provide to their children during the mother's recovery (Fletcher, 2009). Moreover, because mothers may be less attuned to their children during depression, fathers can provide a unique and complementary perspective on children's psychological and family needs. Interventions not focused on maternal depression have demonstrated differential impacts depending on the gender of the parent informant, which may be linked to cultural and gender socialization (Parra-Cardona et al., 2017).

Importantly, in our adaptation we recognize that maternal depression can be a significant yet unacknowledged burden for fathers. We break down the stigma of this burden by first providing fathers with psychoeducation about stress and depression and, second, offering practical tools to increase fathers' efficacy in their family roles. Our focus on parenting efficacy is intended to support fathers' involvement with children who are acculturating to U.S. norms at a faster pace than fathers (Cabrera & Bradley, 2012). In addition, fathers may respond to their partner's diminished parenting role while living with depression by increasing their own parenting of children, yet they may feel unfamiliar or uncomfortable in this role given their primary socialization as breadwinners. Third, and only after fathers express comfort in discussing their partner and children's needs, do we shift attention to their own emotional needs. To further reduce stigma associated with men sharing emotions, we provide fathers a separate space to process their emotions of sadness, anxiety, frustration and helplessness associated with supporting a partner with depression (Valdez et al., 2017).

Our adaptation addressed local concerns about the challenges youth faced in school by building in content in the parent and youth components about minority stress and school belonging, and parent engagement in ethnic/racial socialization, including instilling in youth a sense of ethnic pride and preparation for discrimination. Similarly, we addressed concerns about acculturative stress in the family by aligning perspectives and experiences of parents and youth with respect to their connection to one another and their traditions, and to their migration and settlement trajectories. We drew upon active learning strategies often found in psychodrama (Nieto, 2010), which are widely used with immigrant groups (Smokowski & Bacallao, 2011), to help families find shared meaning in their sociocultural experiences, and reveal agency and strength in their shared trajectory. These experiences help family members feel united, and became part of the family's narrative and mother's recovery from depression.

Adapted programs can vary in the degree of cultural grounding included in the intervention—from surface refinements to deep program modifications tailored to the target audience (Castro et al., 2010). Our surface-level adaptation included modifying the language or cultural symbols to enhance families' acceptance of the program, while our deep-level adaptation considered, recognized, acknowledged, and addressed the role of immigration climate on families' lived experiences, and hence, on their parenting. In our adapted intervention we explore family concerns about immigration climate and its impact on youth identity development and family stability. Through group discussion, role-playing, and psychodrama, youth learn to process their emotions and reactions to bias and to generate possible responses when incidents of bias occur. We promote coping strategies that allow youth to manage and reject negative social messages about immigrants. Parents learn how to discuss heightened threats of deportation with their children using specific and realistic language. Deep-level adaptations acknowledge families' social pressures and enhance intervention gains (Parra-Cardona et al., 2017).

Although several cultural adaptations of interventions have been made, particularly with Latino populations, less scholarly work has focused on identifying how they have been culturally adapted. Previously we reported the outcomes of our intervention (see Valdez, Padilla, et al., 2013), but here we illustrate our adaptation, and the specific tasks and activities we undertook to adapt an intervention for a group of families with a distinct sociocultural context of risk and resilience. Our program has undergone multiple iterations to be linguistically and culturally appropriate for local Latino immigrant families by interlacing the cultural worldview and sociocultural processes present in these families, facilitating recruitment, engagement, and program completion (Bernal & Sáez-Santiago, 2006).

The ecological validity model (i.e. Bernal et al., 1995) and CAPM (Domenech Rodríguez & Wieling, 2004) were influential in our cultural adaptation. However, our adaptation became distinctive in its attention to understanding which families were at greatest risk and need for the intervention in the community, what gaps in healthcare services our intervention would fill for these families, and how to best reach these vulnerable and multi-stressed families. This task influenced recruitment, engagement, community collaboration, and feedback processes. We drew from community-based participatory methods and from implementation research to design, implement, and evaluate our adaptation. This approach is increasingly common in and outside the U.S., particularly with parenting interventions (see Cabassa & Baumann, 2013; Lachman et al., 2016; Parra Cardona et al., 2012). We hope our approach will be useful to others who want to adapt interventions for clinical populations in ways that respond to local contexts. We do not prescribe these tasks nor assume they will fit every project, community, or intervention. Rather, we suggest opportunities for considering adaptations as researchers balance the need for fidelity with an original intervention and the cultural relevance to their local communities (Domenech Rodríguez & Bernal, 2012). We believe the responsiveness to the community happens throughout the process, but particularly in the tasks of reaching, engaging, addressing needs and implementing the intervention.

With respect to future directions, we are planning an evaluation of Fortalezas Familiares that includes a systematic assessment of adaptation and implementation outcomes. Understanding the impact of our adapted intervention on families, and on the capacity-building of local clinicians, who will soon deliver the intervention, will be invaluable in refining the intervention and in ensuring adoption and sustainability of the intervention in clinical settings.