Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Apr 1.
Published in final edited form as: Early Child Res Q. 2023 Jan 17;63:362–369. doi: 10.1016/j.ecresq.2023.01.004

Tiered Universal and Targeted Early Childhood Interventions: Enhancing Attendance Across Families with Varying Needs

Caitlin F Canfield 1, Elizabeth Miller 2, Yudong Zhang 3, Daniel Shaw 4, Pamela Morris 5, Chardee Galan 6, Alan L Mendelsohn 7
PMCID: PMC9937541  NIHMSID: NIHMS1866479  PMID: 36818751

Abstract

This study examined whether a two-tiered parenting program, which provides universal primary prevention along with targeted secondary prevention only for families with increased needs, would have mutually beneficial impacts on attendance across two program components. A secondary analysis of the Smart Beginnings (SB) randomized controlled trial was conducted. SB takes place from birth to age 3 and combines universal delivery of the Video Interaction Project (VIP) with targeted delivery of the Family Check-Up (FCU) for families identified as having increased risks following yearly screening. The current study analyzed whether attendance in VIP in the first six months was associated with FCU attendance for eligible families at six months, and whether FCU attendance at six and 18 months was associated with subsequent VIP attendance. Analyses included logistic and mixed-effects Poisson regression, as well as group-based trajectory analysis. VIP attendance predicted later FCU attendance (AOR = 5.43, p < .01), and FCU attendance predicted later VIP attendance (IRR = 1.35, p < .01) and a high-stable VIP attendance trajectory (AOR=14.98, p < .01). Findings provide strong support for the ability of tiered models to engage parents, to promote effective and efficient service delivery to reduce disparities in school readiness, and their potential to overcome common barriers to attendance and scaling by addressing the heterogeneity of risk among low-income families.

Keywords: tiered programs, prevention, attendance, parenting

Compared to their more affluent peers, children growing up in poverty are less likely to possess the emergent literacy, numeracy, and social-emotional skills necessary for kindergarten entry (Duncan et al., 2010; Reardon, 2011). Although other factors, such as parental education, have been shown to impact children’s cognitive and social-emotional outcomes (e.g., Conway et al., 2018; Rindermann & Ceci, 2018), these are strongly correlated with income and wealth, and income itself affects both children’s outcomes and mediating factors such as parent mental health, the home environment, and parenting practices (Cooper & Stewart, 2021; Yeung et al., 2002). In fact, early parenting practices may account for as much as 25–50% of income-related disparities in children’s early development and school readiness (Brooks-Gunn & Markman, 2005).

Thus, a number of early childhood programs aim to reduce disparities in school readiness by focusing on parents as children’s key sources of self-regulatory and educational support (Cates et al., 2016; Landry et al., 2012). However, there is considerable heterogeneity in strengths and vulnerabilities among families and children living in poverty. While universal programs designed to address general parenting knowledge and skills may be beneficial for most families, others may also benefit from more targeted services addressing correlates of poverty that compromise parenting quality (e.g., mental health, social support; Heneghan et al., 1998). Meta-analytic research in the fields of intervention and prevention science has demonstrated that parent participation is critical for program impact (see Ingoldsby, 2010 and Reyno & McGrath, 2005 for reviews) regardless of whether interventions are universal or more targeted in scope. Yet, rates of participation in parenting programs are generally quite low – less than 50% of target families actually enroll (Baker et al., 2011).

Smart Beginnings (SB) is a tiered early preventive parenting program designed to reduce income-related disparities in early school readiness that couples a universal primary prevention program delivered in pediatric primary care (Video Interaction Project [VIP]; Tier I) with a secondary prevention home-visiting program (the Family Check-Up [FCU]; Tier II) for families with additional risks based on annual screening (Shaw, Mendelsohn, & Morris, 2021, additional details below). Early findings from a randomized controlled trial (RCT) of the SB model have demonstrated high rates of initial attendance in VIP (Miller et al., 2020). The present study seeks to expand these findings to examine bidirectional attendance across the two intervention programs (VIP and the FCU), both with overall initial levels of attendance as well as attendance trajectories over time to account for patterns and identify critical points in attrition.

Importance of Attendance for Program Impact

In line with recent advances in terminology (Sims & Crump 2018), the current study considers initial engagement as the activities related to beginning an intervention, including enrollment; attendance or retention as taking part in intervention sessions; and participation as taking part in a broader set of program activities. Although studies have indicated that each of these steps can affect intervention outcomes (Berkel et al. 2018), attendance is a particularly central component of intervention success (Baker et al., 2011; Heinicke et al., 2006), as parents cannot receive program content without attending sessions. This is especially true of programs for infants and toddlers, the focus of the current study. For example, Huebner and colleagues (2002) found that attendance in a clinic maltreatment prevention program was a significant predictor of improvement in parenting stress and the home environment. Previous research on VIP, one of the program components of SB, have also shown associations between program attendance and parent outcomes, with dose-response relationships (Cates et al., 2018; Weisleder et al., 2016).

Barriers to Program Attendance

Despite the centrality of program attendance, enrollment in single-program parenting interventions is relatively low, with rates on average less than 50% (Baker et al., 2011), even for well-established programs for infants and toddlers like Early Head Start (Besharov et al., 2011) and the Nurse-Family Partnership program (Holland et al., 2018). This is problematic as previous studies have also indicated that families who may benefit more from additional services, namely, those experiencing higher levels of socioeconomic (e.g., lower income/education) or psychosocial (e.g., maternal depression, marginalization/racism) adversity, are less likely to attend parenting programs (Gross et al., 2001; Ingoldsby, 2010; O’Brien et al., 2012; Reyno & McGrath, 2006).

Attendance may be especially challenging for families offered multiple programs. Enrollment in multiple programs may reduce attendance overall, particularly if families view previous or additional services as less relevant for their circumstances (Kazdin & Wassell, 2000), which may be a specific concern for less intensive programs, or if the parent becomes fatigued by the demands of several interventions. Logistical concerns, such as limitations on time, limited transportation, and lack of childcare, are also well-known barriers to program participation for families with low incomes (Gross et al., 2001; Ingoldsby, 2010), and requirements of multiple programs only exacerbate them. On the other hand, participation in one program could also make participation in others more likely, due to alleviation of stressors and increases in parent knowledge and self-efficacy (Feinberg et al., 2010).

Tiered Models

In this context, tiered models such as SB have the potential to address both barriers to attendance and heterogeneity in the needs of families of young children by customizing program design and implementation. In tiered programs, less intensive services, such as parenting skills, screening, and provision of learning materials, are offered to families universally. More intensive options, such as addressing family- and community-level stressors, are offered to those with greater identified risks and tailored to their specific needs. Unlike a “one-size-fits-all” approach, tiered interventions link participants to different levels and kinds of support corresponding to individual families’ strengths and challenges. Thus, tiered models have the potential to improve efficiency and effectiveness in promoting positive outcomes, while minimizing the need to provide more costly interventions to participants who may not require them (Lane et al., 2014; Phaneuf & McIntyre, 2011). Tiered interventions can also address the difficulties related to multiple programs and facilitate family attendance by offering programs in locations that families regularly visit, such as the pediatrician’s office, enabling communication between providers in different programs and between providers and families, promoting trust via a warm handoff, and providing support at different levels according to family needs. Pediatric primary care in particular represents an ideal location to connect with families due to nearly universal reach and repeated points of contact early in life (Peacock-Chambers et al., 2017).

Current Study

A recent study of SB attendance reported very high levels of attendance in the first six months of the program (Miller et al., 2020), even for high-risk families, including those with lower literacy/education and lower parenting self-efficacy. However, as that study examined attendance only prior to screening and implementation of Tier II, it could not examine potential bidirectional effects of attendance in multiple components of the tiered model, and thus it remains unclear whether these components have positive cascading effects on attendance. In addition, despite the potential strengths of tiered models noted above, little is known about attendance in such programs, and especially within the context of pediatric primary care as an initial point of contact with families. A central question, therefore, is whether tiered interventions can address the barriers related to multiple programs and amplify the positive benefits of attendance across levels of services.

The current study examines this question in the context of SB, examining bidirectional associations of attendance across intervention programs within the treatment group, independent of sociodemographic and psychosocial characteristics. In addition to examining overall initial levels of attendance, we analyzed attendance trajectories over time to account for patterns and identify critical points in attrition. We hypothesized that attendance in VIP would lead to increased attendance in FCU and further, that attendance in FCU would strengthen subsequent attendance in VIP.

Materials and Methods

Participants

Mother-infant dyads were enrolled in the RCT of SB in the postpartum units of large urban hospitals in NYC between June 2015 and February 2017 (N = 200) and Pittsburgh between June 2016 and October 2017 (N = 203). Sample size was determined based on power to detect an effect size for the primary outcomes of the study of at least d = .3 and a 20% cumulative attrition rate. Inclusion criteria included full-term, singleton, non-low-birthweight births, no significant prenatal/perinatal medical conditions, no eligibility for Early Intervention at birth (e.g., Down syndrome), and plans to receive pediatric care at the institution. Mothers also needed to be the primary caregiver, speak English or Spanish, and could not have previously participated in VIP or FCU. Of the 403 mother-child dyads enrolled, 201 were randomized into the treatment arm (Tier I: VIP 0–36m; Tier II for positive screens [see below]: FCU at 6m, 18m, and 30m), with the others receiving pediatric care-as-usual. In the current study, we report on the first two screening points for FCU eligibility at 6m and 18m, as there were only two VIP visits following the last screening at 30m. 126 of the 201 were eligible for FCU at either 6m or 18m, (see criteria below), with 90 eligible at both time points, comprising the analytic sample for the current study. No data from the analytic sample was excluded from this study. Figure 1 shows the CONSORT diagram.

Figure 1. Participant Enrollment and Assessment in the Smart Beginnings RCT (CONSORT).

Figure 1

Open in a new tab

Ethical Considerations

Informed consent was obtained from all parents in the study for both their own and their child’s participation. Institutional review board approval was obtained from University 1 (number), University 2 (number), and University 3 (number). Research approval was also obtained from the hospital. In addition, the larger RCT study was registered at www.clinicaltrials.gov (identifier).

SB Intervention

The SB tiered intervention integrates a universal primary preventive intervention (VIP) with a secondary preventive intervention for families with increased risk (the FCU) based on screenings performed at baseline and then annually beginning at 6m. Families could screen into the FCU based on either primary or secondary criteria known to compromise parenting quality (Conger & Elder, 1994). Primary criteria consisted of one of the following: positive screen for maternal depression, family violence, child welfare agency involvement, very low maternal literacy. Secondary criteria consisted of at least two of the following: child behavior challenges, maternal well-being and social support (maternal stress), caregiving (e.g., low involvement), and financial capital (e.g., food insecurity). All families eligible at 6m were considered eligible at 18m, regardless of 18m screening results. When families screened into the FCU, the VIP interventionist regularly provided a warm hand-off to the FCU interventionist in-person or via phone.

VIP was conceived as an enhancement to Reach Out and Read and involves 14 one-on-one sessions with a bachelor’s-level interventionist that take place in the pediatric primary care clinic at the time of well-child visits between birth and 36m, with supplemental sessions every 3 months beginning at 21m, when pediatric visits become less frequent. During each session, the interventionist supports the parent as an active observer of their child’s development. VIP’s core component is video-recording of the parent and child for 3 to 5 minutes of play together with a provided toy or book, after which the interventionist reviews the video with the parent in real-time, reinforcing strengths and discussing ways to extend these behaviors at home. To emphasize messaging and encourage parents to discuss sessions with other caregivers, a copy of the video and a personalized pamphlet with goals for parent-child interactions are provided at the end of the session (Mendelsohn et al., 2013).

The FCU is a brief (3–4 sessions) intervention that employs a Master’s-level parent consultant who incorporates motivational interviewing (MI; Rollnick & Miller, 1995) to engage families in making changes to their caregiving practices, and if desired, support in learning new parenting skills to address problematic child behavior. The FCU consists of 1) a comprehensive, ecological assessment using normed measures, as described above, 2) a rapport-building “Get-to-Know-You” session that focuses on building a collaborative framework and incorporating MI for subsequent intervention, and 3) a Feedback session where the parent consultant continues to use MI in summarizing results of the assessment, thereby creating dissonance for the parent between the child’s current status and the parent’s aspirations for the child. The Feedback session also motivates caregivers to engage in follow-up services. Typically, follow-up sessions focus on parenting needs and factors that compromise parenting quality (e.g., depression, social support, resources), with the parent consultant using the evidence-based Everyday Parenting Curriculum to address these concerns in 1 to 4 sessions (EPC; Dishion, Stormshak, & Kavanagh, 2011).

Measures

VIP Attendance.

VIP attendance was examined in two ways. First, we assessed the number of sessions attended for three sub-periods aligned with timing of the FCU screening and delivery: 0–6m, 9–18m, 21–30m (maximum possible for each: 4). These time periods were chosen because they corresponded to the FCU screening and intervention points, which took place at 6m and 18m. Second, we generated subgroups by examining trajectories of VIP attendance from 9–36m (Sessions 5 to 14) to examine patterns of attendance following the first FCU screening point using group-based trajectory analysis in Stata (Nagin, 2005; Jones & Nagin, 2013). Models with 2 to 5 trajectory groups were examined, with larger Bayesian Information Criterion (BIC) and entropy closer to 1 indicating better model fit. Both BIC and substantive meanings of trajectories were considered in selecting the final model.

FCU Attendance.

Based on prior research in FCU trials showing child intervention outcomes most strongly related to completion of Feedback sessions (Dishion et al., 2008; Dishion et al., 2014), attendance in FCU was defined as completing both the Get-to-Know-You and Feedback sessions. Attendance was scored as a 0–1 dichotomous variable, with 1 indicating attendance at these sessions, for each screening point: 6m and 18m.

Psychosocial Characteristics.

Maternal depressive symptoms and parenting self-efficacy were assessed at the baseline interview. Maternal depression was measured using the Edinburgh Postnatal Depression Scale (EPDS; Cox, Holden, & Sagovsky, 1987; α = 0.87), which asks caregivers to rate their emotional and behavioral depression symptoms over the past seven days. Parenting self-efficacy was assessed at baseline using the 10-item Karitane Parenting Confidence Scale (KPCS; Crnec, Barnett, & Matthey, 2008; α = 0.81). Mothers were asked to rate their confidence in performing specific parenting tasks using a 1 (not at all confident) to 4 (very confident) scale; anchor points were modified to facilitate understanding for families for whom English is not their native language (D’Alonzo, 2011). Maternal stress was measured at the 6m assessment using a single item rating stress during the past month on a 6-point scale (Early Steps Multisite Study, 2004, unpublished data).

Sociodemographic Characteristics.

Parity and receipt of public assistance (e.g., TANF, SSI) were assessed at baseline and included as binary (0/1) variables. Maternal education was also assessed at baseline and included as a continuous variable reflecting level of education.

Analytic Plan

We first examined differences in psychosocial and sociodemographic characteristics between VIP and FCU attenders and non-attenders using t-test and chi square analyses. We then used logistic regression to determine whether any characteristics that were significantly different between groups predicted subsequent program attendance. These characteristics were included in subsequent analyses examining patterns of attendance in order to control for self-selection bias. Finally, based on the temporal nature and progression of VIP and FCU in the SB intervention, we conducted three complementary analyses of bidirectional associations of attendance. We.used logistic regression to examine whether VIP attendance from 0–6m predicted FCU attendance at 6m. Second, we examined associations between FCU attendance at 6m and 18m and VIP attendance in subsequent time periods (i.e., 9–18m and 21–30m). FCU attendance at 6m and 18m was operationalized as a time-varying predictor for VIP attendance for 9–18m and 21–30m, respectively. Based on the nested nature of two records of data within each family, we used a mixed-effects Poisson model in Stata (StataCorp, 2017). Third, we conducted a multinomial logistic regression to examine whether FCU attendance at 6m predicted VIP attendance subgroups across the range of subsequent VIP visits from 9 to 36 months, as identified by the group-based trajectory analysis as described above. We also evaluated the appropriateness of non-linear trends. All analyses controlled for the psychosocial and sociodemographic characteristics listed above. Materials and analysis code for this study are not available at this time.

Results

Descriptive Analyses

Descriptive statistics for the SB and FCU-eligible (i.e., analytic sample) groups are presented in Table 1. SB attendance was higher than has been documented for similar early parenting interventions (Baker, Arnold, & Meagher, 2011), consistent with prior publications of this project (Miller et al., 2020).

Table 1.

Demographic Characteristics and Descriptive Statistics.

Treatment Group
n = 201		 Analytic Sample
n = 90
Child Characteristics (%)
 Female sex 47.76 46.67
 Race/ethnicity
  Asian American 1.02 3.33
  Black/African American 49.75 53.33
  White 3.55 13.33
  Latinx 41.62 36.67
  Other 4.06 10.00
Caregiver Characteristics (% or M(SD))
 Race/ethnicity
  Asian American 1.01 1.11
  Black/African American 45.45 47.78
  White 7.07 13.33
  Latinx 42.93 38.89
  Other 3.54 6.67
 First-time Parent 33.67 29.21
 Public Assistance Receipt 96.48 96.63
 Years of School 12.05 (2.69) 11.26 (2.99)
 Maternal Depression (EPDS) 3.51(4.06) 3.63 (4.08)
 Parenting Self-Efficacy 28.53(2.53) 28.44 (2.60)
 Maternal Stress (6 mo) 2.55(1.36) 2.80 (1.54)
VIP and FCU Attendance
 VIP attendance 0–36m (M(SD)) 9.31 (4.25) 10.01 (3.70)
 VIP attendance 0–6m (M(SD)) 3.49 (0.97) 3.64 (0.72)
 VIP attendance 9–18m (M(SD)) 2.84 (1.56) 3.07 (1.40)
 VIP attendance 21–30m (M(SD)) 2.07 (1.65) 2.30 (1.55)
 FCU attendance 6m or 18m (%) 65.08 68.89
Open in a new tab

Note: Race was not included in analytic models because it is almost fully confounded by site.

Psychosocial and Sociodemographic Predictors of Attendance

As can be seen in Table 1, attendance in VIP from 0–6m was high, with 72% of participants attending all four possible sessions. Therefore, characteristics were compared between those who attended all four sessions and those who attended fewer. Compared to those who attended fewer VIP sessions, participants who attended all four had significantly lower parenting self-efficacy, M(SD) = 28.27(0.23) vs. 29.20(0.20), t = 2.34, p = .02, and completed significantly less education, M(SD) = 8.23(0.22) vs 9.05(0.28), t = 2.08, p = .04. There were no other significant differences between the groups. Parenting self-efficacy was also a marginally significant predictor of FCU attendance, β = −.20, p = .059, though maternal education was not. This was the only difference between FCU attenders and non-attenders, M(SD) = 28.07(0.38) vs. 29.17(0.28) respectively, t = 1.91, p = .059. Given the marginal significance of parenting self-efficacy, its association with VIP attendance after FCU was introduced was also examined and was a significant predictor of subsequent VIP attendance, β = −.24, p < .001.

Bidirectional Associations of Attendance in the Tiered Model

Association between Attendance in VIP before 6m on Attendance in FCU at 6m

To examine the associations between VIP attendance and later FCU attendance, analyses focused specifically on VIP attendance before 6m in relation to FCU attendance at 6m (i.e., attending at least through Feedback session). Logistic regression analysis indicated that VIP attendance from birth to 6m was a significant predictor of FCU feedback attendance at 6m in unadjusted analyses, as well as after adjusting for parenting self-efficacy and the full set of psychosocial and sociodemographic characteristics (Table 2). As can be seen in Figure 2, participants were five times more likely to attend FCU with each additional VIP session attended, with the likelihood increasing from 10% for those who attended half of the potential VIP sessions, to nearly 80% for those who attended all possible VIP sessions.. Of the other covariates in the model, lower parenting self-efficacy significantly predicted higher FCU attendance only when controlling for all potential covariates.

Table 2.

Logistic Regression Predicting FCU Attendance at 6m (n = 90).

Model 1 (unadjusted) Model 2 (adjusted for site and parenting self-efficacy) Model 3 (adjusted for all baseline characteristics)
OR (95% CI) AOR (95% CI) AOR (95% CI)
Number of VIP sessions attended before 6 months 5.12 (2.18,12.01)** 5.43 (2.10, 14.04)** 5.22 (1.92, 14.19)**
Site -- 1.63 (0.57, 4.69) 1.80 (0.54, 5.95)
Parenting self-efficacy -- 0.78 (0.59, 1.03) 0.72 (0.52, 0.99)*
Maternal education -- -- 0.84 (0.67, 1.05)
Maternal depression -- -- 0.88 (0.76, 1.01)
Maternal stress -- -- 1.49 (0.97, 2.31)
First-time motherhood -- -- 1.32 (0.39, 4.49)
Receiving public assistance -- -- 5.85 (0.29, 118.08)
Open in a new tab

OR: odds ratio; AOR: adjusted odds ratio

**

p < .01

*

p < .05

Figure 2. Bidirectional Impacts of Attendance in SB (unadjusted).

Figure 2

Open in a new tab

Association between Attendance in FCU at 6 and 18m on Subsequent VIP Attendance

Mixed effects Poisson regression indicated FCU Feedback attendance was significantly associated with later VIP sessions, in unadjusted models and after adjusting for parenting self-efficacy and the full set of psychosocial and sociodemographic characteristics (Table 3). The rate of VIP attendance among FCU attenders was 1.38 times the rate among FCU non-attenders, with FCU attenders taking part in over 70% of subsequent VIP sessions, while non-attenders took part in 45% (Figure 2). VIP attendance prior to 6m was not included in this model as there was no corresponding FCU predictor for this time period, and attendance was universally high prior to 6m, indicating this variable had little predictive value (Miller et al., 2020). In addition, time was a significant predictor, indicating that mothers were less likely to attend VIP sessions as their children aged. Finally, when controlling for all covariates, mothers with lower parenting self-efficacy attended more VIP sessions.

Table 3.

Mixed-effects Poisson Regression Examining Associations between FCU Attendance and VIP Attendance in Subsequent Period (n = 90).

Model 1 (unadjusted) Model 2 (adjusted for site and parenting self-efficacy) Model 3 (adjusted for all baseline characteristics)
IRR and 95% CI Adjusted IRR and 95% CI Adjusted IRR and 95% CI
FCU attendance 1.34 (1.16, 1.55)** 1.35 (1.17, 1.55)** 1.31 (1.15, 1.51)**
Time (0 = 9–18m; 1 = 21–30m) 0.98 (0.97, 0.99)** 0.99 (.98, .99)** 0.99 (0.98, 0.99)**
Site -- 0.89 (0.82, 0.97)** 0.91 (0.81, 1.02)
Parenting self-efficacy -- 0.99 (0.98, 1.01) 0.99 (0.98, 1.00)†
Maternal education -- -- 0.99 (0.97, 1.01)
Maternal depression -- -- 1.00 (0.99, 1.01)
First-time motherhood -- -- 1.08 (0.98, 1.20)
Receiving public assistance -- -- 0.98 (0.85, 1.13)
Open in a new tab

IRR: incidence rate ratio

**

p < .01

p < .05

†

p < .10

Association between Attendance in FCU on VIP Attendance Trajectories

Models with 2 to 5 trajectories were estimated, and model fit indices are presented in Appendix A. The 3-trajectory model of VIP attendance was chosen because it provided good fit, and each trajectory represented qualitatively different trends in VIP attendance. Quadratic terms were included due to being non-significant. Model fit indices of the 3-trajectory model further supported its fit adequacy: the average posterior probability of assignment for all three groups was above 0.7, and the odds of correct classification were above 5 (Nagin, 2005). Figure 3 shows the three groups. Participants in Group 1 (54% of the sample), the High-Stable group, were highly likely (80%) to attend each of the 10 sessions from 9–36m. Participants in this group attended between 7 and 10 sessions, M = 9.13, SD = 1.06. Participants in Group 2 (36% of the sample), the Decline group, were highly likely to attend the first two of these sessions, sessions 5 and 6 (probability ~0.8); however, they showed a sharp decrease in attendance from the 7th to 10th sessions, and then demonstrated low and stable probability in attending the 11th to 14th sessions. Participants in this group attended between 2 and 6 sessions overall, M = 3.92, SD = 1.50. Participants in Group 3 (10% of the sample), the Low-Stable group, were highly unlikely to attend the 5th session and then discontinued VIP sessions after the 6th session. Participants in this group attended between 1 or fewer sessions, M = 0.25, SD = 0.46. Multinomial logistic regression predicting VIP 9–36m trajectory and controlling for baseline covariates showed that participants who attended the FCU feedback at 6m were more likely to be in the High-Stable VIP group relative to the Decline (attenders: 66% vs. 31% and non-attenders: 29% vs. 46%; AOR=3.87, 95% CI [1.29, 11.67]). FCU attenders were also more likely to be in the High-Stable group than the Low-Stable groups, (attenders: 66% vs. 3% and non-attenders: 29% vs. 25%; AOR=14.98, 95% CI [2.19, 102.26]).

Figure 3. Attendance Trajectories for Three Group Model.

Figure 3

Open in a new tab

Notes. Child age 9 months = Session 5; 12 months = Session 6; 15 months = Session 7; 18 months = Session 8; 21 months = Session 9; 24 months = Session 10; 27 months = Session 11; 30 months = Session 12; 33 months = Session 13; and 36 months = Session 14.

Discussion

This paper examined how attendance in one component of a tiered preventive intervention program may promote attendance in the other tier and may be the first to examine whether the design of such tiered programs increases attendance across the intervention components. We found that VIP attendance predicted subsequent attendance in FCU, and that families who engaged in FCU attended more subsequent VIP sessions. There are multiple potential reasons for this mutually beneficial effect. First, by offering Tier I universally and creating tailored services to better meet families’ needs, reductions in stigmatization of both interventions might have occurred, supporting attendance throughout the program. Second, establishing relationships with a provider in one intervention and having that provider promote the secondary intervention may increase receptivity for the secondary program. Receptivity may have been further facilitated by the warm handoff between VIP and FCU that was part of the SB protocol. Although several previous studies on the effectiveness of warm handoffs for treatment uptake have found no or negative impacts (Horevitz et al., 2015; Pace et al., 2018), these were not done in the context of relationship-based programs, and few studies have been conducted with parents. Current findings may therefore provide new support for the use of warm handoffs, as cumulatively these components appeared to work together to improve attendance. Finally, alignment of program approach within the tiered model, including the culturally-sensitive and family-centered methods used in both VIP and the FCU, may serve to further enhance parent attendance. These results also suggest that tiered models may successfully address barriers to attendance by acknowledging heterogeneity across families, providing services at the appropriate level, increasing match between parents and program, and addressing logistical concerns, all major barriers to parent attendance. Indeed, the group-based trajectory analysis indicated that families who attended the FCU at 6m were more likely to be in the high-stable VIP group than lower attending groups in the subsequent period. These results suggest that the FCU provided more than a limited bump in VIP attendance and that the two programs worked in concert to promote strong attendance throughout the duration of the SB program.

An alternative interpretation of the findings is that those participants who attend either VIP or FCU sessions are more likely to attend later sessions of the other program because of additional characteristics that make them more likely to be attenders or non-attenders across both tiers of SB. To determine whether observed characteristics affected VIP and FCU attendance, we compared attenders and non-attenders on a number of psychosocial and sociodemographic variables. Attenders of VIP had significantly lower parenting self-efficacy and less education (in line with previous findings [Authors, year]), and there were no significant differences between FCU attenders and non-attenders, although parenting self-efficacy was marginally significant. However, even when parenting self-efficacy was included in the prediction models, with and without the other covariates, attendance in early VIP remained a significant predictor of attendance in FCU, and attendance in FCU remained a significant predictor of attendance in subsequent VIP sessions. This indicates that attendance was not solely due to any of these endogenous observable factors, and provides potential evidence, even in a non-randomized sample, for the ability of a tiered intervention like SB to promote attendance across components.

That mothers who had lower parenting self-efficacy and lower education were more likely to attend VIP and FCU, similar to previous research examining early attendance in VIP (Miller et al., 2020), additionally indicates that the SB program was effective in supporting families at varying risk for disparities in school readiness. Further, given the critical importance of attendance for program impact (Baker et al., 2011; Heinicke et al., 2006), these findings indicate that SB, and potentially tiered programs more broadly, may have greater impact by providing a model for engaging families who have been traditionally difficult to retain in parenting programs.

Despite the strengths of this study, it is not without limitations. First, we cannot make causal inferences based on the present analyses. For example, there may be unobserved family characteristics that enable families to self-select into receiving more or fewer services. However, convergence of findings across analyses examining patterns of session attendance support the robustness of the bidirectional associations between attendance in VIP and FCU. Second, the focus on attendance alone in the SB program may limit its applicability. Program engagement is a multidimensional concept, and other aspects of engagement that are important for program impacts (Berkel et al., 2018; Sims & Crump, 2018) are not captured in the outcome variables in this study, which include only program attendance, and are limited to count and dichotomous variables. Future research should examine broader participation to determine whether predictors differ between these aspects of engagement and their interrelatedness. Third, the focus of the SB intervention and RCT on families with limited resources and primarily from racial/ethnic minority communities may limit the generalizability of these findings. However, given that families from marginalized backgrounds are often the most difficult to engage in programs (e.g., Gross et al., 2001), findings may provide important knowledge for program design and implementation. Finally, although the current study is one of the only to examine effects of attendance across components of a tiered program, the present analyses could not answer how or why attendance in VIP may impact attendance in FCU or vice versa. Further mixed methods research to provide a deeper understanding of parents’ reasons for continued engagement is needed to answer these critical questions for effective implementation of tiered parenting programs.

Conclusion

The findings reported here provide evidence for the ability of tiered intervention models to engage parents, particularly among mothers who have traditionally been difficult to engage and whose children may be at increased risk for disparities in school readiness. Attendance in each of the tiered components also predicted subsequent attendance in the other program. These findings have important clinical implications for the implementation and dissemination of preventive parenting programs. By integrating programs that are mutually reinforcing in impact and delivery, SB offers the potential to overcome common barriers to attendance and scaling, using pediatric primary care as an initial point of contact (with near universal access to families with young children) to implement and efficiently tailor two evidence-based programs to address the heterogeneity of risk within low-income families. That SB is also able to maintain high levels of attendance throughout the course of the program further indicates the potential for increased impacts on children’s school readiness.

Highlights.

  • Tiered intervention models may address barriers to parent participation.

  • Integrated programs may be beneficial for attendance across components.

  • Support for tiered implementation to increase impacts for children and families.

Acknowledgments

This work was supported by the National Institutes of Health, Eunice Kennedy Shriver National Institute for Child Health and Human Development [R01HD076390 01–05; 06–10 [Competing]; 3R01HD076390-08S1]

Appendix

Table A.1.

Model Fit Indices of Group Based Trajectory Models

Number of groups BIC (N=885) BIC (N=90) AIC Entropy
2 −428.93 −423.22 −416.97 0.90
3 −423.07 −413.93 −403.93 0.93
4 −427.47 −414.90 −401.15 0.82
5 −434.30 −418.30 −400.80 0.89
Open in a new tab

Notes. N=885 refers the total number of assessments used in model estimation across persons and time, which overstates the theoretically correct N. N=90 refers to the number of individuals in the estimation sample, which understates the true N. The theoretically correct BIC falls between these two BICs. For detailed explanation, please refer to Nagin, 2015, p.68.

Footnotes

Credit Author statement

Caitlin F. Canfield: conceptualization; manuscript writing; Elizabeth Miller: conceptualization, manuscript writing; Yudong Zhang: manuscript writing, statistical analysis; Daniel Shaw: study leadership, manuscript editing and review; Pamela Morris: study leadership, manuscript editing and review; Chardee Galan: consultation, manuscript review; Alan L. Mendelsohn: study leadership, manuscript editing and review

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Contributor Information

Caitlin F. Canfield, NYU Grossman School of Medicine

Elizabeth Miller, NYU Grossman School of Medicine

Yudong Zhang, Northwestern University

Daniel Shaw, University of Pittsburgh

Pamela Morris, New York University, Steinhardt School of Culture, Education, and Human Development

Chardee Galan, University of Southern California

Alan L. Mendelsohn, NYU Grossman School of Medicine

References

  1. Baker CN, Arnold DH, & Meagher S (2011). Enrollment and attendance in a parent training prevention program for conduct problems. Prevention Science: The Official Journal of the Society for Prevention Research, 12 (2), 126–138. [DOI] [PubMed] [Google Scholar]
  2. Berkel C, Sandler IN, & Wolchik SA (2018). Home practice is the program”: Parents’ practice of program skills as predictors of outcomes in the New Beginnings Program effectiveness trial. Prevention Science, 19 (5). [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Besharov DJ, Germanis P, Higney CA, & Start CDMEH (2011). Research and evaluation project: Assessments of twenty-six early childhood evaluations. Welfare Reform Academy. [Google Scholar]
  4. Brooks-Gunn J, & Markman LB (2005). The contribution of parenting to ethnic and racial gaps in school readiness. The Future of Children, 15 (1), 139–168. [DOI] [PubMed] [Google Scholar]
  5. Cates CB, Weisleder A, Dreyer BP, Johnson SB, Vlahovicova K, Ledesma J, & Mendelsohn AL (2016). Leveraging healthcare to promote responsive parenting: Impacts of the Video Interaction Project on parenting stress. Journal of Child and Family Studies, 25 (3), 827–835. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Cates CB, Weisleder A, & Mendelsohn AL (2018). Mitigating the effects of family poverty on early child development through parenting interventions in primary care. Academic Pediatrics, 16 (3 Suppl), S112–S120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Conger RD, Ge X, Elder GH Jr, Lorenz FO, & Simons RL (1994). Economic stress, coercive family process, and developmental problems of adolescents. Child development, 65 (2), 541–561. [PubMed] [Google Scholar]
  8. Conway A, Waldfogel J, & Wang Y (2018). Parent education and income gradients in children’s executive functions at kindergarten entry. Children and Youth Services Review, 91, 329–337. [Google Scholar]
  9. Cooper K, & Stewart K (2021). Does Household Income affect children’s outcomes? A systematic review of the evidence. Child Ind Res, 14, 981–1005. 10.1007/s12187-020-09782-0. [DOI] [Google Scholar]
  10. Cox JL, Holden JM, & Sagovsky R (1987). Detection of Postnatal Depression: Development of the 10-item Edinburgh Postnatal Depression scale. British Journal of Psychiatry, (6), 150. [DOI] [PubMed] [Google Scholar]
  11. Crncec R, Barnett B, & Matthey S (2008). Development of an instrument to assess perceived self-efficacy in the parents of infants: Development OF KPCS. Research in Nursing & Health, 31 (5), 442–453. [DOI] [PubMed] [Google Scholar]
  12. D’Alonzo KT (2011). Evaluation and revision of questionnaires for use among low-literacy immigrant Latinos. Revista Latino-Americana de Enfermagem, 19 (5), 1255–1264. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Dishion TJ, Stormshak EA, & Kavanagh K (2011). Everyday parenting: A therapist’s guide for supporting family management practices. Champaign, IL: Research Press. [Google Scholar]
  14. Dishion Thomas J., Brennan LM, Shaw DS, McEachern AD, Wilson MN, & Jo B (2014). Prevention of problem behavior through annual family check-ups in early childhood: intervention effects from home to early elementary school. Journal of Abnormal Child Psychology, 42 (3), 343–354. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Dishion Thomas J., Shaw D, Connell A, Gardner F, Weaver C, & Wilson M (2008). The family check-up with high-risk indigent families: preventing problem behavior by increasing parents’ positive behavior support in early childhood. Child Development, 79 (5), 1395–1414. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Duncan GJ, Ziol-Guest KM, & Kalil A (2010). Early-childhood poverty and adult attainment, behavior, and health. Child Development, 81 (1), 306–325. [DOI] [PubMed] [Google Scholar]
  17. Feinberg ME, Jones DE, Kan ML, & Goslin MC (2010). Effects of family foundations on parents and children: 3. 5 years after baseline. Journal of Family Psychology, (5), 24. [DOI] [PubMed] [Google Scholar]
  18. Gross D, Julion W, & Fogg L (2001). What motivates participation and dropout among low-income urban families of color in a prevention intervention? Family Relations, 50 (3), 246–254. [Google Scholar]
  19. Heinicke CM, Goorsky M, & Levine M (2006). Pre-and postnatal antecedents of a home-visiting intervention and family developmental outcome. Infant Mental Health Journal, 27 (1). [DOI] [PubMed] [Google Scholar]
  20. Heneghan AM, Silver EJ, Bauman LJ, Westbrook LE, & Stein RE (1998). Depressive symptoms in inner-city mothers of young children: who is at risk? Pediatrics, 102 (6), 1394–1400. [DOI] [PubMed] [Google Scholar]
  21. Holland ML, Olds DL, Dozier AM, & Kitzman HJ (2018). Visit attendance patterns in nurse-family partnership community sites. Prevention Science, 19 (4), 516–527. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Horevitz E, Organista KC, & Arean PA (2015). Depression treatment uptake in integrated primary care: How a “warm handoff” and other factors affect decision making by Latinos. Psychiatric Services, 66 (8), 824–830. [DOI] [PubMed] [Google Scholar]
  23. Huebner CE (2002). Evaluation of a clinic-based parent education program to reduce the risk of infant and toddler maltreatment. Public Health Nursing, 19 (5), 377–389. [DOI] [PubMed] [Google Scholar]
  24. Ingoldsby EM (2010). Review of interventions to improve family engagement and retention in parent and child mental health programs. Journal of Child and Family Studies, 19 (5), 629–645. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Jones BL, & Nagin DS (2013). A note on a Stata plugin for estimating group-based trajectory models. Sociological Methods & Research, 42 (4), 608–613. [Google Scholar]
  26. Kazdin AE, & Wassell G (2000). Predictors of barriers to treatment and therapeutic change in outpatient therapy for antisocial children and their families. Mental Health Services Research, 2 (1), 27–40. [DOI] [PubMed] [Google Scholar]
  27. Landry SH, Smith KE, Swank PR, Zucker T, Crawford AD, & Solari EF (2012). The effects of a responsive parenting intervention on parent-child interactions during shared book reading. Developmental Psychology, 48 (4), 969–986. [DOI] [PubMed] [Google Scholar]
  28. Lane KL, Oakes WP, & Menzies HM (2014). Comprehensive, integrated, three-tiered models of prevention: Why does my school —and district —need an integrated approach to meet students’ academic, behavioral, and social needs? Preventing School Failure: Alternative Education for Children and Youth, 58 (3), 121–128. [Google Scholar]
  29. Mendelsohn AL, Cates CB, Weisleder A, Berkule SB, & Dreyer BP (2013). Promotion of early school readiness using pediatric primary care as an innovative platform. Zero to Three, 34 (1). [Google Scholar]
  30. Miller EB, Canfield CF, Morris PA, Shaw DS, Cates CB, & Mendelsohn AL (2020). Sociodemographic and psychosocial predictors of VIP attendance in smart beginnings through 6 months: Effectively targeting at-risk mothers in early visits. Prevention Science, 21 (1), 120–130. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Nagin DS (2005). Group-based modeling of development. London, England: Harvard University Press. [Google Scholar]
  32. O’Brien RA, Moritz P, Luckey DW, McClatchey MW, Ingoldsby EM, & Olds DL (2012). Mixed methods analysis of participant attrition in the nurse-family partnership. Prevention Science, 13 (3), 219–228. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Pace CA, Gergen-Barnett K, Veidis A, D’Afflitti J, Worcester J, Fernandez P, & Lasser KE (2018). Warm handoffs and attendance at initial integrated behavioral health appointments. The Annals of Family Medicine, 16 (4), 346–348. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Peacock-Chambers E, Ivy K, & Bair-Merritt M (2017). Primary care interventions for early childhood development: A systematic review. Pediatrics, 140 (6), Article e20171661. 10.1542/peds.2017-1661. [DOI] [PubMed] [Google Scholar]
  35. Phaneuf L, & McIntyre LL (2011). The application of a three-tier model of intervention to parent training. Journal of Positive Behavior Interventions, 13 (4), 198–207. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Reardon SF (2011). The widening academic achievement gap between the rich and the poor: New evidence and possible explanations. In Duncan GJ, & Murnane RJ (Eds.), Whither Opportunity (pp. 91–116). Russell Sage Foundation. [Google Scholar]
  37. Reyno SM, & McGrath PJ (2006). Predictors of parent training efficacy for child externalizing behavior problems–a meta-analytic review. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 47 (1), 99–111. [DOI] [PubMed] [Google Scholar]
  38. Rindermann H, & Ceci SJ (2018). Parents’ education is more important than their wealth in shaping their children’s intelligence: results of 19 samples in seven countries at different developmental levels. Journal for the Education of the Gifted, 41 (4), 298–326. 10.1177/0162353218799481. [DOI] [Google Scholar]
  39. Rollnick S, & Miller WR (1995). What is motivational interviewing? Behavioural and Cognitive Psychotherapy, 23 (4), 325–334. [DOI] [PubMed] [Google Scholar]
  40. Shaw DS, Mendelsohn AL, & Morris PA (2021). Reducing poverty-related disparities in child development and school readiness: The Smart Beginnings tiered prevention strategy that combines pediatric primary care with home visiting. Clinical Child and Family Psychology Review, 24 (4), 669–683. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Sims BE, & Crump AD (2018). Commentary on participation in preventive interventions. Prevention Science, 19 (5), 685–688. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. StataCorp. (2017). Stata statistical software: release 15. College Station, TX: StataCorp LLC. [Google Scholar]
  43. Yeung WJ, Linver MR, & Brooks–Gunn J (2002). How money matters for young children’s development: Parental investment and family processes. Child development, 73 (6), 1861–1879. [DOI] [PubMed] [Google Scholar]

RESOURCES