Scaling up Attachment and Biobehavioral Catch-up with Latine Families: Implementation Processes and Effectiveness

Abstract

Attachment & Biobehavioral Catch-up (ABC) is an evidence-based early intervention promoting attachment security and self-regulation by enhancing parental sensitivity. When scaling up an evidence-based intervention to the community, it is essential to consider families’ preferences and cultural values to ensure cultural appropriateness, while maintaining fidelity to the intervention’s core components. In this article, we aimed to test the effectiveness of ABC in improving parental sensitivity when implemented in Spanish at a community level in the USA. The effectiveness of ABC when implemented in Spanish in the community was assessed through pre-post changes in observed parental sensitivity in 79 Latine families with children ages 6 months to 4 years. Parental sensitivity increased from pre- to post-test, with a similar effect size to previous studies (d = 0.67). Results and practice implications are discussed, highlighting the importance of the support to providers reaching families of diverse origins.

Evidence-based parenting interventions targeting modifiable parenting behaviors are important in promoting positive child outcomes, particularly when delivered early in the child’s life (Berlin et al., 2021; Dozier & Bernard, 2019, 2023). One such intervention is Attachment & Biobehavioral Catch-up (ABC). Based on attachment theory and stress neurobiology, ABC promotes early attachment security and self-regulatory skills by targeting parental sensitivity, a modifiable intervention target with powerful downstream benefits on child development (Dozier & Bernard, 2019; Humphreys et al., 2021). To ensure that early intervention services like ABC reach all families who could benefit from them at a community level, interventions should be tailored to the values and preferred language of the families and communities they are reaching, while maintaining fidelity to the intervention core components (Finno-Velazquez, 2014; Pew Research Centre, 2022).

In the United States (US), a substantial proportion of the families receiving family and child services are Latine or Hispanic (Benito-Gomez & Flores, 2022; Finno-Velazquez et al., 2022; National Home Visiting Resource Center, 2022). In many cases, however, language may pose a barrier given that services may only be available in English (Benito-Gomez & Flores, 2022; Finno-Velazquez, 2014; Zablah et al., 2022). It is important, as well, that early interventions are culturally responsive and consider families’ culture and values to promote engagement and effectiveness (Huey et al., 2023; Lau, 2006; Parra-Cardona, 2019). In this article, we describe culturally responsive efforts aimed to support service providers implementing ABC at a community level with Spanish-speaking, Latine families in the US. We also evaluate the effectiveness of ABC with Latine families in the scale-up implementation phase – meaning when the intervention is implemented in standard practice on a broad scale, therefore in less controlled and more variable conditions than in a university-led intervention study- as measured via pre- to post-intervention changes in observed parental sensitivity.

The ABC Intervention: Enhancing Early Attachment Security and Self-Regulation by Targeting Parental Sensitivity

ABC is a brief (10 sessions) home visiting intervention for parents of infants and toddlers who have experienced early adversity (Dozier & Bernard, 2019). Based on attachment theory and stress neurobiology, ABC promotes a sensitive style of parent-child interaction, a powerful predictor of early attachment security and behavioral and physiological regulation (Humphreys, 2021). Specifically, the intervention focuses on three target parental behaviors: nurturing children when they are distressed (e.g., soothing or comforting them, even when they do not clearly signal their need for nurturance), following children’s lead (i.e., a contingent and responsive style of interaction that is child-directed), and avoiding intrusive or frightening behaviors (e.g., yelling, threatening; Dozier & Bernard, 2019). Importantly for the aims of this article, parental sensitivity has been shown to buffer or protect children from maladaptive outcomes under conditions of contextual stressors or hardships (e.g., poverty; Gerlach et al., 2021), a protective effect also found in families from diverse ethnic groups (Mesman et al., 2012). In contrast, young children who are exposed to intrusive parenting are more likely to develop avoidant attachments, low self-regulation, and emotional and behavioral problems (e.g., Carlson et al., 2003; Gueron-Sela et al. 2017). However, studies including diverse ethnic groups have found inconsistent results, suggesting that the meaning and effects of intrusive behaviors may differ depending on the context and culture in which parent-child relationships are embedded (Benito-Gomez, 2022; Iruka et al., 2014).

Home visitors, referred to as “parent coaches” in the ABC model, use video feedback, psychoeducation, and activities to practice behavior targets related to parental sensitivity, along with in-the-moment comments on parent behaviors as they occur during the session. Both the parent and the child are present during these sessions (Dozier & Bernard, 2019). The in-the-moment comments are considered the active ingredient of the intervention, as a higher number and quality of comments predict behavior change in the target behaviors (Caron et al., 2018). Mastering these in-the-moment comments is challenging because parent coaches must provide comments as the behaviors occur, and these comments should include specific components, such as a clear behavioral description (Caron et al., 2018).

Through several RCTs, ABC demonstrated effectiveness in improving parental sensitivity (Bick & Dozier, 2013; Yarger et al., 2020), children’s attachment security and organization (Bernard et al., 2012), diurnal patterns of cortisol production (Bernard, Dozier et al., 2015) and behavioral regulation (Lind et al., 2017). Importantly, some of these intervention effects were maintained for three and up to eight years after receiving the ABC intervention when children were infants (Bernard et al., 2015; Korom et al., 2020; Zajac et al., 2019). Moreover, RCTs of ABC have shown positive outcomes among different populations, including families involved in child protection services (Bernard et al., 2012), foster families (Bick & Dozier, 2013), and internationally adoptive families (Yarger et al., 2020). The RCT of ABC with families receiving child protective services yielded the largest amount of evidence on its efficacy (Bernard et al. 2012; Bernard et al., 2015; Korom et al., 2020). The families receiving the intervention in that study were predominantly urban Black families, whereas the sample of other RCTs (e.g., with internationally adoptive families) were predominantly white. As detailed below, ABC has also shown beneficial effects with low-income Latine mothers and their infants (Berlin et al., 2018, 2019), so overall there is evidence of the intervention effectiveness with diverse populations in several RCTs.

Scaling Up with Fidelity: ABC Implementation at a Community Level

Given its success in improving child outcomes through RCTs, the ABC intervention started to scale up and be implemented in the community. Implementing an evidence-based program in the community is a complex task, and typically effect sizes tend to be smaller than when implemented in RCTs by model developers, among other reasons because of inadequate fidelity to the model (Allen et al., 2018). To ensure the effectiveness of the intervention when scaling up, the ABC team used a fidelity monitoring process involving intense and continued supervision of parent coaches, including a coding system for tracking in-the-moment comments (Caron & Dozier, 2019). The coding system, based on observational methods, facilitates a quantitative fidelity monitoring process and supports the delivery of the intervention by coding parenting behaviors and parent coaches’ comments in their own sessions (Caron & Dozier, 2019).

A play interaction task assessing parental sensitivity has been administered since at least 2013 in the community implementation of ABC while parent coaches are in the training year, providing pre- and post-intervention observational data on the effectiveness of the ABC intervention. Using this outcome, previous studies have found pre- to post-intervention changes in parental sensitivity with a large effect size, comparable to that found in RCTs (Caron et al., 2016; Roben et al., 2017). This effect has been attributed to the intense fidelity monitoring system developed. Pre- to post-intervention changes in parental sensitivity were also found when implementing ABC through telehealth at a community level (Schein et al., 2023).

A further step in the translational pipeline involves evaluating the effectiveness of an intervention when implemented with different populations or when adaptations are made to meet specific implementation needs (e.g., delivery in another language). Evaluations of effectiveness in scale-up research help to establish the external validity of treatment effects initially demonstrated in high internal validity studies like RCTs, which often have limited external validity (Shadish et al., 2002).

Latine Families in the US as Recipients of Early Home Visiting

One of the largest ethnic groups served by evidence-based home visiting interventions like ABC in the US is Hispanic or Latines. In 2022, 30% of the families who received evidence-based home-visiting programs in the US were Hispanic or Latine and 15% of the total number of families received services in Spanish (National Home Visiting Resource Center, 2022). There is a need to ensure Latine families receive high-quality services, particularly considering health and service access disparities among the Latine population (Finno-Velazquez et al., 2022). Ethnic and racial health inequalities are well documented in the US. For example, Latines in the US are less likely to have access to and receive adequate services (Petersen et al., 2014) and these inequalities are exacerbated when Latine families are undocumented (Cabral & Cuevas, 2020).

Structural and institutional racism reinforces health inequalities through, among other ways, the cascading effects of residential segregation based on ethnicity and race. Predominantly Black or Latine neighborhoods tend to have less access to healthcare services, limited opportunities for high-quality education, and so on (Bailey et al., 2017). An additional barrier to accessing evidence-based treatments is the lack of availability of services in Spanish, particularly for those Latine families with a recent immigration background (Finno-Velazquez, 2014; Pew Research Centre, 2022). In these cases, the demands of adapting to a new culture and language may be combined with stressful experiences such as uncertain legal status, pre-migration traumatic experiences, or child-parent separation at the US-Mexico border (Benito-Gomez & Flores, 2020; Lopez-Zeron et al., 2020).

Adaptations beyond providing services in Spanish may be necessary to ensure cultural appropriateness and engagement with Latine families (Stirman et al., 2019). However, these adaptations should be balanced with maintaining fidelity to the core components of the evidence-based intervention, as modifications to these core components can jeopardize effectiveness (Stirman et al., 2019). It has been argued that cultural adaptations to evidence-based interventions should be selective and based on evidence of diminished effectiveness or engagement with certain groups (Lau, 2006). Therefore, to evaluate the suitability of an intervention like ABC for Latine families, the first step is to assess its outcomes within that population.

There is evidence of ABC’s effectiveness with Latine families, mainly from an RCT by Berlin and colleagues at the University of Maryland (US). The sample consisted of 208 mothers, 87% of them of Hispanic or Latine ethnicity (Berlin et al., 2018, 2019; Hepworth et al., 2021). ABC was delivered in Spanish or English depending on the mother’s preference by two bilingual parent coaches, as a complement to Early Head Start. Intervention manuals and other documents were translated into Spanish and the research team took particular care in coding videorecorded Spanish-speaking mother-infant interactions. There were intervention effects on intervention targets such as parental sensitivity (Berlin et al., 2018), maternal sensitivity to child distress (Hepworth et al., 2021), children’s cortisol regulation (Berlin et al., 2019), and emotion regulation (Hepworth et al., 2020). Furthermore, first-generation Latine mothers who received ABC in this study were interviewed to gather their views on ABC. They found the intervention beneficial and culturally relevant and reported no difficulties understanding ABC concepts (Aparicio et al., 2016).

Ensuring Cultural Appropriateness of ABC with Latine Families

To ensure cultural appropriateness and cultural humility when delivering ABC to families of diverse ethnic and racial backgrounds, including Latine parents, the program developers implemented several key measures. A cultural competency and humility training model was established for providers and supervisors, which may be more practical than creating multiple specific cultural adaptations for each target group receiving the evidence-based intervention (Huey et al., 2023). Parent coaches participate in weekly group supervision sessions focused on case conceptualization and discussion, where issues related to culture, race, and power dynamics are considered and explored. Supervisors are trained in anti-racism, cultural humility, and cultural sensitivity to enhance their ability to address these aspects in a personalized manner, consistent with a precision approach and other approaches as the personal relevance of psychotherapy (Hall et al., 2021; Huey et al., 2023). This personalized approach is preferred due to the diversity among Latine families and findings indicating that individual-level factors, such as acculturation, influence the need for and effectiveness of cultural adaptations (Huey et al., 2023).

Regarding parent coaches and supervisors’ recruitment, ABC encourages community recruitment by not requiring a specific diploma to be a parent coach, but rather to pass a screening process based on valuing attachment needs and the ability to deliver in-the-moment comments (Caron, Roben et al., 2018). Community agencies implementing ABC, however, may have their own requirements. As will be detailed in the Methods, in the case of parent coaches serving Latine families at a community level, the vast majority were from the same ethnic and racial background. Although the evidence on whether provider-client ethnic-racial matching is beneficial to outcomes is inconclusive (Meghani et al., 2009), in our experience familiarity with families’ cultural values and background facilitates an organic inclusion of cultural-related factors in ABC delivery. The availability of Hispanic and bilingual supervisors has been one of the major supports identified by Latine parent coaches working with Latine families, together with the availability of high-quality Spanish translations of the intervention materials, translated by graduate-level native Spanish speakers and ABC supervisors and trainers to ensure cultural appropriateness (Carrera et al., 2022a; Zablah et al., 2022).

More structured adaptations for Latine families receiving ABC have also been implemented. Following the Framework for Reporting Adaptations and Modifications-Expanded (FRAME) to evidence-based interventions (Stirman et al., 2019), these adaptations can be categorized as content modifications, particularly in terms of tailoring, tweaking, or refining aspects of the intervention content to align with values relevant to Latine families. A companion guide to the ABC manual was developed and distributed to all parent coaches working with Latine families. This guide included a glossary of Spanish translations for different Latin American dialects, reflection activities, and discussions on cultural aspects.

For example, the Latine value of “respeto,” which emphasizes children’s obedience to authority and deference, may interact with the concept of following a child’s lead discussed in ABC sessions (Lopez et al., 2022). Periodic workshops for parent coaches working with Latine families addressed this and other cultural aspects, offering strategies to balance ABC concepts with families’ values (Carrera et al., 2022a; Zablah et al., 2022). Following the FRAME framework, the workshops and the companion guide can be categorized as adaptations to the training and evaluation of providers (Stirman et al., 2019).

Other suggestions included incorporating discussions of cultural values in elements of the ABC intervention with an in-built flexibility to consider families’ characteristics, as in the “voices from the past” conversation in sessions seven and eight (Dozier & Bernard, 2019). In these sessions, parent coaches discuss with the caregiver possible influences that are getting in the way of implementing the target parenting behaviors. The “voices from the past” conversation was originally intended to consider the caregivers’ attachment experiences, but the conversation can be flexible enough to deal with the influence of cultural values in their parenting. One example is discussing gender stereotypes and gender socialization, since they may be related to dismissing the need for nurturance in boys (Raffaelli & Ontai, 2004). All the adaptations mentioned were fidelity-consistent since they retained the core components of the intervention (Dozier & Bernard, 2023; Carrera et al., 2022b).

The effectiveness of ABC in Spanish in the scale-up implementation phase (varied providers and implementation settings, etc.; Gottfredson et al., 2015), has not been empirically assessed yet. This is important to address potential concerns about the external validity of evidence-based parenting interventions when implemented in diverse ethnic groups, like Spanish-speaking, Latine families in the US (Finno-Velazquez et al., 2022; Sattler et al., 2022). Furthermore, assessing intervention outcomes when scaling up an evidence-based intervention, like parental sensitivity in the case of ABC, is key to testing whether effect sizes from RCTs are maintained upon scale-up. Consequently, the main goal of this article is to assess the pre- to post-intervention change in parental sensitivity among Latine families who received the ABC intervention in Spanish across different community sites implementing ABC. We also evaluated the parenting behaviors of responsiveness, intrusiveness, and positive regard to assess potential differences compared with previous ABC effectiveness results, considering the potential cultural meanings of intrusiveness.

Method

Participants

Families Receiving ABC

The sample consisted of 79 parent-child dyads who received the ABC intervention in Spanish in the US between June 2018 and September 2022. Dyads were included if they had completed both a pre- and a post-intervention parental sensitivity assessment and received the intervention fully or partially in Spanish. There were 47 families from the Mid-Atlantic/Northeast region of the US (59.5%), 29 from the South Central and West area (36.7%), and 3 from the Southeast (3.8%). All families received the full ABC intervention, with 26 cases (32.9%) receiving the intervention through telehealth. Fifty-nine families received the ABC-Infant version (for children aged 6 to 24 months) and 20 families the ABC-Early Childhood version (for children aged 24 to 48 months). Both ABC versions shared most of the content and focused on enhancing parental sensitivity through in-the-moment comments. Parental sensitivity was coded using the same indicators for both models.

The data were collected for program evaluation purposes. Sociodemographic information was available for only nine families from the included sample (11.39%) because that information was collected separately from the play interaction assessments and not systematically gathered until 2021. Families were referred for ABC when infants had experienced some adversity. To double-check that the families and the parent coach used Spanish during the play interaction assessment, the first author reviewed the stored play interaction videos of all the included families.

From the nine cases with available information, all the target caregivers identified as female. Age ranged from 20 to 46 years (M = 31.22, SD = 8.17). Eight were the birth mother and one was the foster caregiver of the target child. Regarding the socio-educational level, four had not completed high school, three had completed high school, and two had attended college. Five out of nine (55.6%) had a household income below $35,000 a year, the federal poverty guideline for a household of five people in the US in 2023 (US Department of Health and Human Services, 2023). All nine caregivers identified as Latine; five identified as Mexican, two as Colombian, and two as Central American (Guatemala and Honduras). Six caregivers identified as White and three selected the option “other, two or more races”.

Parent Coaches

The families were served by 18 parent coaches across eight community sites in seven states in the US and in Washington DC. The (participating) parent coaches were nested within community sites in the following way: two sites provided data for five and four parent coaches respectively, one site for two, and five sites for one parent coach each. Each parent coach had seen a range of one to 17 families (the median was two families per coach). We had sociodemographic information for 12 of the 18 parent coaches (66.67%). All the coaches identified as female and 11 of the 12 identified as Latine or Hispanic (91.7%). They reported diverse Latine origins (nonexclusive options): Mexico (n = 4), Colombia (n = 3), Dominican Republic (n = 2), or Cuba, Venezuela, Salvador, and Puerto Rico (n = 1 each). Six identified as white, two as Black, and three as other races. Their mean age was 42 years (SD = 10.05, range 31-61 years).

The parent coaches received 2-3 days of training and weekly and group supervision during the training year to ensure fidelity to the model. They had to implement ABC with at least three cases before certification. The one-hour small group supervision (2-3 parent coaches per group, named general clinical supervision) focused on content delivery and case conceptualization and was delivered by advanced-degree supervisors. The 30-minute individual supervision focused on in-the-moment commenting and coding of target parental behaviors and comments. All parent coaches with available data (n = 17, 94.4%) had an in-the-moment commenting supervisor who spoke Spanish, and four (22.22%) had a clinical supervisor who spoke Spanish. There was one coach with missing data about their supervisors. All the parent coaches were certified in ABC and had met fidelity criteria during their training year. Parent coaches certify in ABC if in their last ten sessions before certification they meet certain criteria on their in-the-moment comments: frequency (at least one comment per minute), accuracy (at least 80% of the comments should be on-target, meaning they should describe the parent behavior is actually happening) and quality (at least one average component per comment). Components mean either describing the parent behavior, labeling the target behavior, or naming an outcome of that parent behavior in the comment, so one average component means that the coded parent coaches’ comments included at least one of those elements on average. For more information on ABC certification criteria and fidelity monitoring procedures, see Roben et al. (2017) or Caron and Dozier (2019).

For those for whom we had fidelity monitoring data (n = 12), their in-the-moment commenting scores for their last ten ABC sessions before certification were: mean comments per minute = 1.71, average number of components per comment = 1.46, percentage of on-target comments = 94%. Fidelity monitoring data was not available for six parent coaches.

Procedure

A semi-structured play interaction task was administered to parent-child dyads at baseline (before receiving Session 1 of the ABC intervention) and at discharge (after receiving Session 10, the last session) to assess parental sensitivity. ABC sessions are typically delivered weekly. Videos of the play interaction were stored in HIPAA-compliant servers and coded by reliable coders at the University of Delaware. Given that the data were collected for program evaluation purposes, they were considered exempt by the Institutional Review Board of the University of Delaware.

Measure

Parental sensitivity assessment coding

For the play interaction task, parents were instructed to play with the target child as they normally would with a standardized set of toys (e.g., set of blocks, rattle, squeaky toy) for about nine minutes. The assessment is based on the National Institute of Child Health and Development Recording of the Caregiving Environment (ORCE; NICHD Early Child Care Research Network [ECCRN], 1996). The same protocol and instructions were used for the pre- and post-intervention parental sensitivity assessments. This measure was developed and used with the large multi-ethnic sample of the NICHD ECCRN study (1996), and it predicted positive child outcomes across ethnic groups.

Coders at the University of Delaware scored the parent-child interaction on three 5-point Likert scales related to the construct of parental sensitivity: Responsiveness to non-distress, intrusiveness, and positive regard (NICHD ECCRN, 1996). Responsiveness assessed the degree to which the caregiver responded flexibly and in a contingent way to the child (e.g., the child moves blocks, and the caregiver imitates them; Humphreys et al., 2021). Intrusiveness assessed abrupt or physically invasive parenting behaviors, unattuned to the child’s cues (e.g., shaking a rattle in the child’s face), whereas positive regard assessed the parental expression of positive affect towards the child (e.g., smiling). The scores were averaged (intrusiveness reverse scored) to create a composite score of parental sensitivity. The videos were coded by eight research assistants who were masked to the site and timing (pre- or post-intervention) of the parental sensitivity assessment and understood or spoke Spanish. All coders completed training in the coding of observed parental sensitivity using the three scales mentioned and reached an 80% or above agreement with a master coder on 10 test videos to become reliable before coding any videos.

Thirty-two play interaction videos (20.25% of the total 158) were double coded to assess inter-rater reliability. Reliability was excellent for all scales: one-way, random intraclass correlation coefficients were .80 for responsiveness, .86 for intrusiveness, and .85 for positive regard.

Data Analyses Plan

We used hierarchical multilevel modeling (HLM) to account for the nested structure of the data within parent coaches and community sites, replicating the same analyses of previous studies on ABC community effectiveness using parental sensitivity as an outcome (Caron et al., 2016; Roben et al., 2017). We calculated change scores for parent behavior subtracting the pre-intervention from the post-intervention parental sensitivity scores. Then, the changes scores were entered in an HLM equation without any predictor variables, but accounting for the site $(u_{00k})$, parent coach $(r_{0jk})$ and case-level error term $(e_{ijk})$. Therefore, the intercept $({\gamma }_{000})$ indicates the estimated average amount of change on each score. We mainly focused on the composite parental sensitivity score as an outcome but also analyzed the change in each indicator individually (responsiveness, intrusiveness, and positive regard), considering concerns about the cultural appropriateness of the intrusiveness construct. We provided Cohen d as a measure of effect size for the means comparison.

$${\text{Sensitivity Change}}_{ijk}={\gamma }_{000}+u_{00k}+r_{0jk}+e_{ijk}$$

Results

Pre- to Post-Intervention Change in the Parental Sensitivity Composite

The intercept (B₀ = 0.64, p < .001) shows a significant increase in parental sensitivity composite scores from pre- to post-intervention. This indicates that, on average, caregivers who participated in the 10-session ABC intervention in Spanish demonstrated greater observed parental sensitivity after the intervention. Figure 1 illustrates the distribution of these scores before and after the intervention. The pre-test parental sensitivity composite mean was 3.24 (SD = .93), whereas, for the post-test, it was 3.89 (SD = .78). The magnitude of the effect size comparing pre- and post-intervention parental composite sensitivity scores (not accounting for the nested structure of the data) was Cohen d = 0.67 (95% CI 0.42 – 0.91), a large effect size for this field (Schuengel et al., 2021).

Figure 1.

Violin Plots of the Distribution of Pre- and Post-Test Scores in the Observed Parental Sensitivity Composite for Caregivers Receiving ABC in Spanish

Pre- to Post-Intervention Changes in Responsiveness, Intrusiveness, and Positive Regard

The linear mixed models for each score showed that responsiveness increased by nearly one point on average from pre- to post-intervention (B₀ = 0.91, p = <.001). Intrusiveness decreased from pre- to post- (B₀ = −0.54, p = <.001), while positive regard increased (B₀ = 0.56, p = <.001), although relatively less than responsiveness. Table 1 displays the means, standard deviations, and effect sizes for the comparison of pre- and post-test scores.

Table 1.

Descriptive Data and Effect Size for the Mean Comparison of Pre- and Post-Test Scores in Parental Sensitivity Indices

	Pre-test	Post-test	Cohen d
	M (SD)	M (SD)	Point estimate	95% CI
				LL	UL
Responsiveness	2.54 (1.11)	3.37 (1.04)	.68	.43	.92
Intrusiveness	2.15 (1.33)	1.61 (0.93)	−.39	−.62	−.16
Positive regard	3.34 (1.13)	3.90 (1.11)	.43	.20	.67
Sensitivity composite	3.24 (0.93)	3.89 (0.78)	.67	.42	.91

Note. LL = Lower limit of 95% CI, UL = Upper limit of 95% CI.

Discussion

This study aimed to test the effectiveness of the evidence-based intervention, ABC, in promoting parental sensitivity when implemented in Spanish with Latine, Spanish-speaking families in the US under standard practice conditions. We used a single-group, pre- and post-intervention design to evaluate changes in observed parental sensitivity following the delivery of the ABC intervention. Post-intervention scores in observed parental sensitivity were higher than pre-intervention scores, with a large effect size. This effect size was similar to those reported in previous analyses of ABC effectiveness when scaled up (Caron et al., 2016; Roben et al., 2017; Schein et al., 2023), as well as in RCTs (Bick & Dozier, 2013; Yarger et al., 2020). Our analyses of each parental sensitivity index (responsiveness, intrusiveness, and positive regard) mirrored previous ABC effectiveness findings, with greater changes from pre- to post-intervention in responsiveness than in positive regard and intrusiveness (Schein et al., 2023).

This study used a single group pre-test-posttest design, which has some methodological limitations (Marsden & Torgerson, 2012). The inference of causality regarding the intervention effect on parental sensitivity is supported by the moderate stability of parental sensitivity over a short period (Behrens et al., 2012), helping to rule out history and maturation effects as alternative explanations. Most importantly, a causal treatment effect of the ABC intervention on parental sensitivity has been previously found in several RCTs with an active control group (Bick & Dozier, 2013; Berlin et al., 2018; Yarger et al., 2020).

We believe it is noteworthy that ABC appeared to be as effective when implemented in Spanish with Latine families as in previous ABC scale-up research (Caron et al., 2016; Roben et al., 2017; Schein et al., 2023). In earlier studies, the effectiveness of ABC in community settings has been attributed to intense fidelity monitoring supervision based on observational methods (Caron & Dozier, 2019). Providing this high-quality support to providers in another language is a challenging endeavor, particularly when the intervention and implementation procedures were originally developed in English. The large effect size, comparable to previous studies, suggests that the supervision and support provided to Spanish-speaking parent coaches were of the same quality as those provided to English-speaking parent coaches. In fact, Spanish-speaking parent coaches reported that having high-quality Spanish translations of the intervention materials and Spanish-speaking supervisors were the most helpful resources for implementing ABC in Spanish (Carrera et al., 2022a).

The results of this study help to expand the generalizability of ABC intervention effects on parental sensitivity across different intervention settings, ethnic groups in the US, and languages of delivery. It remains to be tested, however, whether the ABC intervention would be effective in enhancing parental sensitivity in other countries and cultures with substantial differences to the US, both in the implementation context (e.g., providers not accustomed to evidence-based interventions and fidelity monitoring) or in the outcome of interest. Regarding intrusiveness, it decreased following the ABC intervention similarly to earlier studies with diverse populations (Schein et al., 2023). Future studies could further investigate the meaning and impact of intrusiveness on child development in Latine families (Benito-Gomez et al., 2022).

A scale-up study including proximal outcomes such as ours would not be adequate without previous solid efficacy and effectiveness studies, since typically it lacks sufficient methodological rigor to draw causal inferences (Gottfredson et al., 2015). But when the effectiveness of an intervention has been established in designs with high internal validity (i.e., ability to draw a causal inference of the treatment effect), it can complement previous efficacy and effectiveness findings by showing whether the treatment effect is maintained in standard practice or scale-up conditions, that is, in less controlled and more variable conditions. Presenting more variability in providers, participants, settings, and so on threatens internal validity, but, by definition, it enhances the external validity of the treatment effect, since external validity means “inferences about whether the observed cause-effect relationship holds over variation in persons, settings (including different places as well as different times), treatment variables, and measurement variables” (Gottfredson et al., 2015, p. 896).

A limitation of this study is that we analyzed only the cases who completed the intervention and had pre- and post-intervention sensitivity assessments (i.e., completer analyses). Attrition and drop-out cases were not analyzed. Although previous studies have not shown differences in parental sensitivity (at baseline and in their last session) between completer cases and those who did not complete the intervention at a community level (Caron et al., 2016), future studies could assess in more detail this issue when delivering ABC in Spanish with Latine families.

We did not have complete sociodemographic information for the majority of the sample, however, we believe it can be reasonably inferred that the families had profiles similar to those for whom we did have information. The main reasons for this assumption are the continuity of intervention sites implementing ABC with Spanish-speaking, Latine families, and our review of all play interaction videos to confirm that families spoke in Spanish. The authors of this study have been bilingual, Spanish-speaking clinical supervisors of parent coaches implementing ABC in Spanish with Latine families. Based on our anecdotal, but lived, experience, the profile of the few participating families with sociodemographic information is broadly representative, or at least informative, of the predominant sociodemographic profile of Latine families receiving ABC in the community in the US. Racial-ethnic identity can’t be assumed without self-identification; however, we think it is not an extremely risky deduction assuming that the Spanish-speaking families receiving ABC in the USA of the sample are Latine. Out of 63.7 million Hispanics in the US, only 990,000 did not have a Latin American origin (Spain), that is, 1.6% of the total Hispanic population in the US (Pew Research Centre, 2023). Nevertheless, the missing sociodemographic information is a clear limitation of this study.

In our sample, we combined families receiving ABC in-person and through telehealth, as well as ABC-Early Childhood and ABC-Infant cases. The rationale for this combination was that these versions do not differ substantially in the intervention active ingredient (in-the-moment commenting) and proximal outcomes (parental sensitivity), nor in their treatment effect sizes on parental sensitivity. Previous scale-up research of the ABC lab has shown that ABC delivered completely through telehealth and hybrid (mix of telehealth and in-person) effectively improved parental sensitivity (Schein et al., 2023), including both ABC-Early Childhood and ABC-Infant families. Given that our main aim was to assess ABC effectiveness with Latine families, the potential fine-grained differences in the magnitude of the treatment effect by delivery type were not evaluated, and families receiving ABC in different settings were combined to maximize sample size and statistical power.

In conclusion, this study provided evidence that a parenting intervention, ABC, is effective in enhancing parental sensitivity when delivered in Spanish, extending previous findings of ABC’s general effectiveness in scale-up efforts (Caron et al., 2016; Roben et al., 2017; Schein et al., 2023) and in an RCT with low-income Latine mothers (Berlin et al., 2018). Our study complements previous findings of ABC effectiveness on parental sensitivity with low-income Latine mothers receiving Early Head Start (Berlin et al., 2018) by replicating the main treatment effect under more variable and less controlled standard practice conditions across the US (e.g., different community sites, multiple providers, etc.), contributing to the substantial convergent and accumulated evidence on ABC effectiveness with diverse families (Caron et al., 2016; Roben et al., 2017; Berlin et al., 2018).

Therefore, it seems feasible to promote parental sensitivity when implementing a complex, evidence-based early intervention in standard practice across diverse settings and languages. In the ABC approach, fidelity monitoring based on observational methods and ongoing supervision is essential to maintain effectiveness (Caron & Dozier, 2019). In this case, supervision (and other kinds of implementation support) was provided in Spanish, which helped to support providers delivering the intervention in this language (Carrera et al., 2022a), and ultimately to benefit Latine families whose first language is Spanish and may not have been reached otherwise (Finno-Velazquez, 2014).

A major recommendation for practice is that having bilingual providers from the same or similar background as the target families is crucial, but not enough. To ensure success, program developers must provide support to providers in another language or cultural context with the same quality and intensity as that provided for the intervention in its original language and cultural setting (Zablah et al., 2022). Although this won’t always be possible to do in-house by program developers, strong partnerships with intervention champions and local agencies of the target communities, along with sustainability efforts (training local supervisors, train-the-trainer models, etc.) can be instrumental in meeting this goal (Costello et al., 2021).

It is important to consider how the intervention content and procedures may interact with the cultural or community values of the target communities and to include fidelity-consistent adaptations, if needed (Lau, 2006; Stirman et al., 2019). Rather than a one-size-fits-all approach, we have highlighted the importance of allowing a certain degree of flexibility and of considering each case’s needs, including factors like the level of acculturation or the differences between countries of origin or ancestry (e.g., Dominican, Mexican, Cuban, etc.), particularly within a broad and heterogenous ethnic group such as Latines (Huey et al., 2023).

These results are promising for the increasing number of Spanish-speaking Latine families in the US, who have the right to access parenting interventions and home visiting services (Benito-Gomez & Flores, 2020; Finno-Velazquez et al., 2022; Parra-Cardona, 2019). They also benefit families and children in diverse countries and contexts, provided that program developers and providers make a conscious effort to incorporate cultural and language diversity into the implementation process.

Key Findings.

• The ABC intervention was effective in improving parental sensitivity among Latine families in the scale-up implementation phase.

• The effect size of the pre- post change in parental sensitivity was large for the field and similar to previous studies. We attribute this to the intense fidelity monitoring process successfully provided in Spanish to bilingual providers.

• Program developers and other responsible parties should carefully consider how the cultural values and preferences of the ethnic-racial groups they are reaching may interact with the evidence-based intervention to incorporate appropriate fidelity-consistent adaptations.

Statement of the Relevance of the Work for Infant and Early Childhood Mental Health.

Early interventions often become less effective when scaled up and implemented under real-world conditions, especially with racial and ethnic groups different from those originally tested or when delivered in other languages. Therefore, it is crucial to continue testing the effectiveness of early interventions under variable conditions to increase the external validity of the treatment effects. Program developers should carefully balance fidelity with adaptation, supporting diverse families and providers with cultural humility and respecting their cultural values while maintaining the intervention’s effectiveness.

Data Availability Statement

Because of the research’s nature and ethical and legal constraints, supporting data is not available.