Early Emotion Development Intervention Improves Mental Health Outcomes in Low-Income, High-Risk Community Children

Abstract

Children living in poverty and facing related forms of adversity are at higher risk for experiencing concurrent and later psychopathology. Although negative psychological outcomes can be improved by enhancing sensitive and responsive caregiving early in development, interventions targeting the caregiver–child dyad are not readily accessible. The present study investigated the feasibility and effectiveness of delivering a shortened eight-session form of Parent–Child Interaction Therapy-Emotion Development (PCIT-ED) in-person or remotely as an early intervention for 3–6-year-old children (N = 62) at elevated risk for psychopathology who were growing up in low-income communities. Caregiver–child dyads were randomized to eight-sessions of PCIT-ED or online parenting education. Relative to parenting education, children receiving PCIT-ED exhibited lower externalizing symptoms and functional impairment and more positive peer relationships following the intervention. Findings support the effectiveness of this shortened form of PCIT-ED, delivered in-person or remotely, as an early intervention to improve symptoms of psychopathology and functioning in high-risk children living in poverty.

Trial registration

Clinicaltrials.gov; NCT04399629.

Introduction

Children living in poverty and facing related forms of adversity are at higher risk for poor socioemotional outcomes [1, 2] as well as concurrent [3] and later psychopathology [4, 5]. Although there is increasing evidence that these negative developmental trajectories can be significantly improved by enhancing sensitive and responsive caregiving early in development (e.g., [6–9]), interventions that target the caregiver–child dyad early in life are not readily available in most communities in need within the United States. Moreover, a robust literature suggests that such early interventions have long-lasting impacts on children’s well-being and result in significant social cost savings over time [10]. Based on this, there is an urgent need to make empirically supported interventions feasible and available to families most in need but with the least access. Potential ways in which this can be accomplished include shortening existing intervention programs, offering interventions remotely, and providing interventions through school settings where communities in need can most easily access them [10].

Background

Decades of basic developmental research has established the foundational importance of a sensitive, emotionally available, nurturing, and consistent caregiver in the first five years of life for the formation of a secure attachment relationship between the child and caregiver [11–14] and to promote a wide array of adaptive emotional and cognitive developmental outcomes [15–17]. Children who do not receive this type of caregiving are more likely to experience maladaptive outcomes in a number of social, emotional, and cognitive domains and are at elevated risk for psychopathology across the lifespan [5, 16, 53, 54, 55]. Moreover, the presence of a supportive caregiver is likely necessary not only to meet the basic needs of the child but also to buffer the negative effects of environmental challenges, stressors, and adversities that the child may encounter. This may be particularly important for children growing up in under-resourced environments and/or children of racial or ethnic minority backgrounds who most likely face structural racism and discrimination, and for whom these stressors are acute and unfortunately currently inevitable.

Importantly, there is substantial empirical evidence demonstrating the modifiability of the parent–child relationship in the context of a variety of early interventions (see [10], for review), with changes that include enhancing parental sensitivity and building the attachment relationship [18]. There is also evidence that modifying the parent–child relationship can be key to the treatment of both externalizing and internalizing disorders in early childhood (e.g., [19–22]). For example, two preventative interventions that are delivered in the home and target parenting, the Nurse-Family Partnership [20] and the Family Check-Up [22] have been found to prevent socioemotional problems such as disruptive behavior [23] in young children. Moreover, additional parent–child interventions have been developed that focus on delivering ‘in the moment’ feedback to parents on their parenting behaviors and interactions with their children. One such intervention, the attachment and biobehavioral catch-up, has been found to enhance children’s attachment and regulatory abilities [19].

Perhaps the most widely used parent–child intervention that uses this ‘in the moment’ parenting feedback technique is Parent–Child Interaction Therapy (PCIT). PCIT is an empirically supported treatment for externalizing behaviors in 3-to 7-year-old children [22] that meta-analytic evidence demonstrates yields large effects on children’s disruptive behavior [24]. PCIT allows for direct modification of parenting behavior as it is occurring through use of the “bug-in-the-ear” technique in which the parent wears a microphone and is coached by the therapist through a one-way mirror during in vivo interactions with the child. PCIT consists of two treatment components: ‘child-directed interaction (CDI)’ during which the therapist coaches parents to let their child lead in play and increases positivity in the dyad, and ‘parent-directed interaction (PDI)’ which teaches parents warm yet firm limit setting. Each component is taught, modeled, and role-played until the caregiver demonstrates proficiency, with a typical course of treatment taking place over 14 to 16 weekly 1-hour sessions [25]. Compared to a large literature documenting the efficacy of PCIT in clinical settings, fewer effectiveness studies have been conducted to examine child outcomes when PCIT is administered in community settings. Moreover, although several community mental health agencies have published data indicating that PCIT effectively decreases children’s disruptive behavior [26–31], these effectiveness studies have typically not included a comparison treatment group, limiting knowledge about whether PCIT is superior to other treatments available in the community setting. One effectiveness study that compared PCIT delivered by community therapists to usual therapeutic care found that children in the PCIT group had larger reductions in disruptive behavior, but intent-to-treat analyses found no significant differences between the PCIT and treatment as usual groups [32].

Parent–Child Interaction Therapy-Emotion Development (PCIT-ED) is a modified version of PCIT designed to target early childhood depression and includes a focus on emotion development [33, 34]. These goals are achieved by helping the parent serve as a more effective external emotion regulator and emotion coach to the child. Parents are taught to help their child recognize emotions in self and others (emotion awareness) as well as to tolerate and regulate intense emotions (emotion regulation). A randomized-controlled trial (RCT) demonstrated that children who received PCIT-ED evidenced lower rates of depression, lower depression severity, and lower impairment compared to children in a waitlist control condition [35]. Children also demonstrated decreases in oppositional symptoms through standard PCIT components, and the effects of PCIT-ED on depression and oppositional symptoms were large in magnitude. The treatment also reduced stress and depression symptoms in caregivers, indicating that the positive benefits of PCIT-ED extend beyond the child without specifically targeting the caregiver. Follow-up studies with this sample have demonstrated that families in the PCIT-ED group also evidenced improved child emotional competence and parent-reported and observed improvements in parenting practices, and many of these additional benefits were sustained at a 3-month follow-up assessment [36, 37]. Importantly, PCIT-ED was administered in a modular fashion, with the novel 8-sesssion emotion development (ED) module following 12 sessions of standard PCIT. Assessments were completed at the end of each module allowing the researchers to demonstrate added efficacy of the emotion development module for the treatment of child depression [38]. The addition of emotion development to standard PCIT thus increases therapeutic application by targeting a broader range of psychopathology (i.e., internalizing and externalizing presentations). However, despite the promise and availability, many families who could benefit from PCIT do not have access [39]. Implementation within community and school settings provides an opportunity to increase access in underserved communities with high need, but limited access.

The Present Study

The present study aimed to assess the effectiveness of a shortened version of PCIT-ED for use as an early intervention in a sample of 3-to 6-year-old community children with or at elevated risk for psychopathology. Whereas standard PCIT is effective in reducing children’s externalizing symptoms [22], the added emotion development module included in PCIT-ED effectively reduces internalizing symptoms [34], thus broadening the scope and applicability of the intervention to target a broader range of psychopathology symptoms. This randomized-controlled trial (RCT) design compared the effectiveness of a modified 8 session form of PCIT-ED with a proven effective online parent training program, Parenting Wisely (PW; [40, 41]), that provides parents with constructive problem-solving skills, but does not include individual therapist contact, coaching, or focus on emotional development, as provided in PCIT-ED. To enhance feasibility, this early intervention was implemented within a local public school system with high rates of poverty and delivered by community-based therapists who were trained and received ongoing supervision by PCIT-ED global trainers with expertise in the administration PCIT-ED.

Target outcomes included children’s mental health (i.e., internalizing and externalizing symptoms), functional impairment, and social functioning, along with caregiver’s mental health, stress, and well-being. We hypothesized that children and caregivers assigned to PCIT-ED would show positive improvements in each of these respective areas relative to those assigned to the Parenting Wisely comparison. Because the study was in progress when the COVID-19 pandemic occurred, we pivoted to remote “zoom” delivery of the PCIT-ED intervention partway through the study. This provided an additional unique opportunity to explore potential differences in the efficacy of the intervention when delivered in-person versus remotely.

By streamlining PCIT-ED to focus on key aspects deemed most appropriate for early intervention, the ultimate goal is to create effective, feasible, and sustainable preventative and early intervention programs that are accessible to families at a time in early development when the intervention may have the strongest impact. By assessing efficacy along with collecting key information about feasibility and implementation (e.g., intervention acceptability) the present study aimed to provide new insight into how PCIT-ED may be best used to accomplish this goal, both in terms of the effectiveness of this shortened 8-session version of PCIT-ED and implementation and uptake when accessed through a community setting: local public schools.

Methods

Participants

Participants were 62 3- to 6-year-old children (M = 5.08, SD = 0.92, Range = 3.35–6.98) and a primary caregiver. The sample consisted of 50% girls and 50% boys. Racial identification, obtained via caregiver report, is as follows: 98.4% Black, 1.6% more than one race (see Table 1 for additional demographic information).

Table 1.

Baseline characteristics of study participants

Demographics	All participants (N = 62)		PCIT-ED (N = 36)		PW (N = 26)		PCIT-ED vs. PW
	Mean	SD	Mean	SD	Mean	SD	t	p
Child age in years	5.08	0.92	4.99	0.95	5.20	0.89	− 0.91	.367
Income-to-needs ratio	1.07	0.82	1.13	0.73	0.99	0.94	0.66	.514
Parent age in years	33.81	7.38	33.71	7.71	33.93	7.04	− 0.12	.909
	%	N	%	N	%	N	x 2	p
Female sex	50.0	31	50.0	18	50.0	13	0.00	1.000
Race							F.E	.419
Black	98.4	61	100.0	36	96.2	25
More than 1 race	1.6	1	0.0	0	3.9	1
Hispanic ethnicity	0.0	0	0.0	0	0.0	0	–	–
Mother’s highest level of education							F.E	.706
11th grade or below	5.7	3	3.6	1	8.0	2
High school/GED	35.8	19	42.9	12	28.0	7
Some college	47.2	25	50.0	15	44.0	11
Associates degree	7.5	4	3.6	1	12.0	3
Bachelors degree	0.0	0	0.0	0	0.0	0
Graduate degree	3.8	2	0.0	0	8.0	2

F.E. Fisher’s exact test

Recruitment/Screening:

Children between 3.0 and 6.11 who lived in the Jennings School District between Sept. 2019 and Oct. 2020 who met inclusion criteria were invited to participate (see Fig. 1 for consort diagram). The Child Behavior Checklist (CBCL; [42], a measure of child psychopathology filled out by either a caregiver or teacher, was used to screen participants. We obtained N = 124 completed CBCLs. CBCL T-scores are normed by age and sex (M = 50; SD = 10) with scores > 70 considered clinically significant. Children who scored ≥ 60 on the internalizing or externalizing subscales of the CBCL (to capture those with high psychopathology risk) were deemed eligible for participation. Exclusion criteria consisted of living with primary caregiver for < 6 months, concurrently in active psychotherapy or on unstable doses of psychotropic medication, major medical or neurological disease, or Autism Spectrum Disorder or significant developmental delay.

Fig. 1.

Consort diagram of Jennings study

Study Design

Participants meeting eligibility criteria were invited to complete a comprehensive baseline assessment (N = 83; described in detail below). Those who completed the baseline assessment were randomized using a computer-generated randomization table to one of two intervention conditions (shortened 8 session version of PCIT-ED or Parenting Wisely [PW] online education). Randomization used a permuted block design, and participants were stratified by sex and age (male vs. female and age 3.0–5.5 vs. age 5.6–6.11). To achieve a distribution of 60% randomized to PCIT-ED and 40% randomized to PW, each set of five randomizations within a stratification group included 3 PCIT-ED and 2 PW assignments. Part-way through the study, it became clear that families were interested in enrolling otherwise eligible siblings. When sibling participants were enrolled, they were treated as a unit for randomization purposes and assigned to the same intervention condition (to enhance feasibility for stressed families). Three sibling pairs were randomized to PCIT-ED; three to PW. These procedures resulted in 36 participants randomized to PCIT-ED and 26 randomized to PW. Once treatment was completed, the same comprehensive assessment that had been completed at baseline was again administered. For both PCIT-ED and PW, participation was tracked throughout the trial and multiple attempts were made to contact families who failed to begin or complete the intervention and to resolve any barriers to participation. All study procedures were approved by the WUSM Institutional Review Board and written informed consent from caregivers (and age-appropriate verbal assent from children) was obtained prior to participation in research procedures.

The project was launched in September 2019 with service delivery occurring in-person within the district in well-equipped rooms with childcare provided. However, when the COVID-19 pandemic hit the St. Louis area in March 2020, schools closed and service delivery was quickly transitioned to remote delivery using video conferencing (44% in the PCIT-ED group completed all sessions online; Supplementary Table 1 details cases completed in-person vs. video-conferencing/Zoom).

Interventions

PCIT-ED

For the current study, a previously tested and proven effective novel treatment for preschool psychopathology, PCIT-ED (PCIT-ED; [34] was adapted to a shortened 8 session format for use as an early intervention [35]. The prior adaptation of PCIT-ED from standard PCIT was designed to target early childhood depression and includes a focus on emotion development [33, 34]. The 8 session¹ version used in the current study retained and integrated core elements of each component of PCIT-ED within sessions in order to create a more feasible intervention to target broad early-onset risk for and elevated psychopathology symptoms. The treatment focused on strengthening the parent–child relationship by teaching and in vivo coaching of positive play techniques, giving effective commands, and methods for handling child noncompliance in a firm, non-punitive manner as well as enhancing the child’s emotion development by teaching the parent to serve as a more effective emotion guide and regulator for the child. Many of the elements of each module of the PCIT-ED intervention (child directed interaction/CDI, parent directed interaction/PDI and emotional development/ED) are integrated across the 8 sessions, rather than restricted to discrete modules. In general, the intervention is about 20% CDI, 20% PDI, and 60% ED. Parents are taught to help their child recognize emotions in self and others (emotion awareness) as well as to tolerate and regulate intense emotions (emotion regulation). The shortened 8 session early intervention implemented in the present study was conducted in a dyadic format between the therapist and caregiver, with some sessions devoted first to “teaching” the caregivers, after which the therapist coaches the parent during live in vivo interactions. Session content was the same regardless of whether the sessions were conducted in person or remotely.

PCIT-ED Therapist Training and Supervision

All primary caregivers were invited to participate, although the mother was the primary participant for most (87.1%) families. Three therapists (Master’s level clinicians), two of whom were from the same community as the study population, delivered the intervention. All received four days of formal training in standard PCIT from certified trainers and two additional days of training in the novel emotional development (ED) module. Subsequently, they received formal supervision by global trainers on an ongoing basis throughout the trial. These global trainers were Master’ level clinicians who were involved in the development of the ED module and the adaptation of the shortened treatment, and have extensive experience with administering PCIT-ED as well as training and supervising therapists in the administration of PCIT-ED. Each session was subsequently rated by a supervising therapist using the Treatment Integrity Checklist, customized for the content within each session, to monitor adherence to the PCIT-ED manual. Therapists maintained a high degree of adherence across the 155 total sessions conducted (M = 89.3%, SD = 11.6; see Supplementary Table 2 for additional details). All intervention sessions were videotaped and viewed by the supervising global trainers in weekly supervision sessions, in which fidelity to the PCIT-ED manual and therapeutic issues specific to each case were discussed.

Parenting Wisely (PW)

An evidence-based parenting education course, Parenting Wisely (PW), was used as the comparison condition [40, 41]. This intervention is an online self-paced parenting training, comprised of 7 modules delivered over 3–5 h, that emphasizes constructive problem solving skills, without individual therapist contact, coaching, or focus on emotion development as provided in PCIT-ED. PW also provides a section for skills practice after the course is completed. In addition, weekly telephone consultation was available for parents to discuss the skills and application of the techniques taught in the online modules with a researcher familiar with PW.

Pre- and Post-Intervention Assessment Measures

Trained interviewers blind to the intervention condition and uninvolved in the intervention process conducted the baseline and post-intervention assessments. Participants were randomized following administration of the full baseline assessment battery, and the first intervention session typically took place within four weeks of the baseline assessment. For post-intervention assessments, attempts were made to conduct the assessment shortly after completion of the final PCIT-ED or PW session. However, there was substantial variability in the timing of the post-assessment relative to the final PCIT-ED or PW session due to COVID-19 complications and difficulties determining the last session for those who did not complete the intervention (PCIT-ED M = 61.86 days, SD = 52.46, Mdn = 51, Range = 6—162; PW M = 49 days, SD = 35.85, Mdn = 48, Range 2–122). Parents were compensated $100 and the child was given a small toy for both the pre- and post-intervention assessments. Parents were paid an additional $100 for completing approximately 50% of their assigned intervention (4 PCIT-ED sessions or 4 PW modules).²

Baseline assessments prior to mid-March 2020 (n = 22 PCIT-ED; n = 15 PW) were completed in person; the remaining baseline assessments (n = 14 PCIT-ED; n = 11 PW) were completed remotely via video conferencing/Zoom. All participants from the PCIT-ED group who completed their post-assessment did so remotely (n = 22), whereas in the PW group 7 participants completed their post-assessment in person, and 11 completed their post-assessment remotely. There were no significant differences in either internalizing or externalizing symptoms for children in the PW group at either the baseline or post-intervention assessment as a function of whether they completed the post-intervention assessment in-person or remotely, indicating that is not a meaningful source of variance in the study.

Primary Child Outcomes: Mental Health, Functional Impairment, and Social Functioning

Children’s mental health, functional impairment, and social functioning were measured with the Health and Behavior Questionnaire—Parent Version (HBQ-P; [43]), which was administered to primary caregivers pre- and post-intervention. Mental health includes dimensional scales of internalizing symptoms (depression, overanxious, separation anxiety) and externalizing symptoms (oppositional defiant, conduct problems, overt hostility, relational aggression). Each item is scored on a 3-point Likert scale (0 = Never or not true; 1 = Sometimes or somewhat true; 2 = Often or very true). The functional impairment scales measure the extent to which children’s mental health symptoms impact their own functioning (e.g., getting along with teachers, becoming withdrawn or irritable) and their broader impact on the family (e.g., parent’s inability to take their child out in public, have friends or relatives visit, cause worry about child’s chance for doing well in the future). Each item is scored on a 3-point Likert scale (0 = none, 1 = a little, 2 = a lot). Social functioning scales capture children’s peer relations (peer acceptance/rejection, bullied by peers), in addition to broader social engagement and withdrawal (asocial, social inhibition), and prosocial behaviors. Items in the peer relations subscale were scored on a 4-point Likert scale (1 = Not at all like; 2 = Very little like; 3 = Somewhat like; 4 = Very much like). Items in the social withdrawal subscale were scored on a 3-point Likert scale (0 = Never or not true; 1 = Sometimes or somewhat true; 2 = Often or very true). Items in the prosocial behavior subscale were scored on a 3-point Likert scale (0 = Rarely applies; 1 = Applies somewhat; 2 = Certainly applies). The subscales that comprise the primary scales, indicated in parentheses, were explored in secondary analyses. Higher scores are indicative of more severe pathology, more functional impairment (i.e., worse adaptive functioning), and better social functioning. Each scale of the HBQ-P was found to have moderate to high reliability (α’s ranged from 0.74–0.93; see Supplement Table 4).

Parent Outcomes: Stress, Depression, and Optimism

Three parent outcomes, stress, depression symptoms, and optimism were measured to assess the impact of PCIT-ED on parental well-being. The Parenting Stress Index (PSI; [44]) was used to assess the severity of parenting stress along with potential dysfunction in the parent–child relationship (α ≥ 0.97). The Beck Depression Inventory-II (BDI-II; Beck et al. 1990) was used to assess the existence and severity of depression in parents (α ≥ 0.91). The Life Orientation Test-Revised (LOT-R), a widely-used measure of dispositional optimism and pessimism, [45],α ≥ 0.61), was included based on prior work demonstrating that optimism (i.e., having positive expectations for the future) predicts psychological and physical well-being [46, 47].

Intervention Acceptability and Implementation

Intervention acceptability was characterized by both overall completion rates for PCIT-ED and PW, along with metrics regarding the number of sessions completed in-person or remotely. Descriptive information regarding completion rates is provided to add context to the present findings and to aid researchers in future study development.

Analysis Plan

Baseline demographic characteristics and primary measures of child psychopathology and functioning and parental well-being were compared between the PCIT-ED and PW groups using t-tests for continuous variables and Fisher’s Exact or Chi-square tests for categorical variables. The primary outcome measures of child psychopathology and functioning (HBQ-P scales: internalizing, externalizing, functional impairment-child, functional impairment-family, global peer relations, social withdrawal, prosocial behavior) and the secondary outcomes of parental well-being (PSI, BDI-II, LOT-R) were analyzed in all randomized subjects using multiple imputation to ensure no missing values at the post assessment [48]. This conservative intent-to-treat approach was implemented using the MI and MIANALYZE procedures in SAS v9.4. Twenty-five multiply imputed datasets were created based on the baseline variables corresponding to the post outcome measures of interest, sex, baseline age, and baseline income-to-needs ratio. Imputations were conducted by randomization group. General linear models were then conducted on the multiply imputed datasets, with post outcomes as the dependent variable, PCIT-ED vs. PW group as the independent variable, and covarying for the baseline characteristic corresponding to the dependent variable, child sex, and baseline age.

Secondary analyses compared PCIT-ED and PW participants who completed the post-assessment and at least 3 PCIT-ED sessions or 6 PW sessions, respectively. This approach limited the sample to participants who demonstrated substantive engagement with their assigned intervention (determined, in part, by our evaluation of the distribution of completed sessions), with analyses conducted on the primary measures and the subscales that comprised those measures to explore whether certain symptoms or characteristics might have been particularly impacted by the intervention. Additional exploratory analyses further probed the intervention effectiveness between in-person PCIT-ED, remote administration of PCIT-ED, and PW for the primary outcomes. As with the primary analyses, each of these general linear models covaried for baseline characteristic, child sex, and baseline age. The Benjamini–Hochberg procedure with the false discovery rate (FDR) set at 0.05 was used to control for multiple comparisons within each set of analyses.

Results

There were no significant baseline differences between participants in the PCIT-ED and PW groups on demographic characteristics or any outcome measure at baseline that is included in the present analyses (see Tables 1, 2 for details). Of the 36 participants randomized to PCIT-ED, 20 (56%) completed 3 or more therapy sessions, 9 (25%) completed 1–2 therapy sessions, and 7 (19%) completed 0 therapy sessions. Of the 26 participants randomized to PW, 16 (62%) completed 6–7 modules, 1 (4%) completed 1–5 modules, and 9 (35%) completed 0 modules.

Table 2.

Baseline caregiver-report measures for primary child and parent outcomes

Baseline child measures (HBQ-P subscales)	All participants (N = 62)		PCIT-ED (n = 36)		PW (n = 26)		PCIT-ED vs. PW
	Mean	SD	Mean	SD	Mean	SD	t	p	FDR p
Internalizing	0.44	0.24	0.45	0.25	0.43	0.21	0.20	0.843	0.949
Externalizing	0.47	0.34	0.42	0.31	0.53	0.37	− 1.25	0.216	0.949
Functional impairment—child	0.42	0.45	0.37	0.42	0.48	0.47	− 0.98	0.329	0.949
Functional impairment—family	0.46	0.51	0.47	0.50	0.44	0.54	0.20	0.843	0.949
Global peer relations	3.45	0.57	3.45	0.62	3.46	0.50	− 0.06	0.949	0.949
Social withdrawal	0.69	0.38	0.68	0.37	0.71	0.41	− 0.33	0.740	0.949
Prosocial behavior	1.33	0.38	1.32	0.37	1.34	0.39	− 0.23	0.819	0.949
CBCL internalizing/externalizing scores: at-risk vs. clinically significant
	%	N	%	N	%	N	χ2	p	FDR p
At-risk vs. Clinical							0.01	0.912	0.912
At-risk (T-score 60–70)	50.8	31	51.4	18	50.0	13
Clinical (T-score > 70)	49.2	30	48.6	17	50.0	13
Baseline caregiver well-being measures
	Mean	SD	Mean	SD	Mean	SD	t	p	FDR p
Stress: PSI total stress	226.45	55.05	234.50	57.31	215.31	50.72	1.36	0.178	0.315
Depression: BDI-II	11.63	9.87	12.97	10.67	9.77	8.49	1.27	0.210	0.315
Optimism: LOT-R	16.54	3.76	16.31	3.68	16.85	3.92	− 0.54	0.589	0.589

The primary intent-to-treat analysis³ with the full sample demonstrated significantly better outcomes in the PCIT-ED group over the PW group as a function of the intervention for four of the seven child outcomes: externalizing symptoms (B = −0.27, SE = 0.07, t = − 3.76, p < 0.001, FDR p < 0.001, Cohen’s d = − 0.74), functional impairment–family (B = − 0.30, SE = 0.10, t = – 2.96, p = 0.005, FDR p = 0.018, Cohen’s d = − 0.57), global peer relations (B = 0.19, SE = 0.07, t = 2.71, p = 0.009, FDR p = 0.021, Cohen’s d = 0.59), and social withdrawal (B = − 0.28, SE = 0.11, t = − 2.58, p = 0.013, FDR p = 0.023, Cohen’s d = − 0.49). The PCIT-ED group also demonstrated significantly greater parental optimism compared to the PW group (B = 2.79, SE = 0.99, t = 2.82, p = 0.007, FDR p = 0.021, Cohen’s d = 0.85). All effect sizes were medium to large (see Table 3).

Table 3.

Intent-to-treat analysis of outcome measures at post assessment in PCIT-ED (N = 36) and parenting wisely (N = 26) groups covarying for corresponding baseline measure, child sex, and baseline age

	Estimate	SE	t	p	FDR p	ES
Child measures (HBQ-P Subscales)
Internalizing	− 0.108	0.067	− 1.63	0.109	0.153	− 0.51
Externalizing	− 0.266	0.071	− 3.76	0.000	0.000	− 0.74
Functional impairment – child	0.046	0.134	0.34	0.733	0.855	0.05
Functional impairment – family	− 0.298	0.101	− 2.96	0.005	0.018	− 0.57
Global peer relations	0.194	0.072	2.71	0.009	0.021	0.59
Social withdrawal	− 0.275	0.107	− 2.58	0.013	0.023	− 0.49
Prosocial behavior	0.020	0.125	0.16	0.873	0.873	0.24
Caregiver well-being measures
Stress: PSI total stress	0.44	13.96	0.03	0.975	0.975	− 0.31
Depression: BDI-II	− 2.04	2.22	− 0.92	0.362	0.543	− 0.48
Optimism: LOT-R	2.79	0.99	2.82	0.007	0.021	0.85

Bolded values indicate those that are statistically signficant (p < 0.05)

ES = effect size (Cohen’s d for the change from baseline to post-assessment averaged across 25 imputed datasets)

Secondary analyses with participants who completed either 3 or more PCIT-ED sessions or 6 or more PW sessions and the post-assessment indicated better outcomes in the PCIT-ED compared to the PW group for externalizing symptoms (t = − 2.20, p = 0.037, FDR p = 0.169, Cohen’s d = − 0.54), relational aggression (t = − 2.40, p = 0.023, FDR p = 0.169, Cohen’s d = − 0.60), functional impairment-family (t = − 2.69, p = 0.012, FDR p = 0.169, Cohen’s d = − 1.01), and the bullied by peers subscale of the global peer relations scale (t = − 2.11, p = 0.044, FDR p = 0.169, Cohen’s d = − 0.45); however none of these analyses survived corrections for multiple comparisons. In addition, the greater parental optimism in the PCIT-ED group over the PW group appears driven, at least in part, by significantly lower pessimistic expectations measured by the pessimism subscale (t = − 2.95, p = 0.007, FDR p = 0.017, Cohen’s d = − 0.90). All effect sizes were small to medium (see Table 4 for full model details and pre- and post-intervention means for all measures). Although the relatively small sample size in this study did not allow for a robust interrogation of specific factors impacted by the treatment, this examination of the subscales provides additional preliminary information regarding which symptoms might be most impacted by the intervention and/or malleable.

Table 4.

Comparisons of post characteristics in Jennings PCIT-ED (3 + Sessions) and parenting wisely (6 + sessions) groups covarying for baseline characteristic, child sex, and baseline age

HBQ-P	PCIT-ED (N = 17)				Parenting Wisely (N = 15)				PCIT-ED vs. parenting wisely
	Baseline		Post		Baseline		Post
	Mean	SD	Mean	SD	Mean	SD	Mean	SD	t	p	FDR p	ES
Internalizing	0.43	0.22	0.27	0.16	0.40	0.21	0.35	0.24	− 1.81	0.082	0.189	− 0.62
Depression	0.22	0.20	0.13	0.17	0.20	0.18	0.19	0.24	− 0.94	0.358	0.379	− 0.33
Overanxious	0.45	0.20	0.29	0.20	0.43	0.24	0.39	0.26	− 1.50	0.146	0.239	− 0.56
Separation anxiety	0.63	0.35	0.38	0.23	0.57	0.35	0.46	0.31	− 1.56	0.130	0.235	− 0.55
Externalizing	0.42	0.30	0.24	0.23	0.57	0.41	0.52	0.41	− 2.20	0.037	0.169	− 0.54
Oppositional defiant	0.67	0.36	0.44	0.34	0.74	0.53	0.64	0.46	− 1.30	0.205	0.284	− 0.39
Conduct problems	0.35	0.30	0.15	0.24	0.46	0.33	0.34	0.35	− 1.79	0.084	0.189	− 0.39
Overt hostility	0.43	0.41	0.28	0.23	0.55	0.46	0.48	0.45	− 1.62	0.118	0.235	− 0.25
Relational aggression	0.25	0.31	0.10	0.18	0.54	0.45	0.60	0.59	− 2.40	0.023	0.169	− 0.60
Functional impairment—child	0.41	0.39	0.12	0.20	0.48	0.37	0.28	0.25	− 1.99	0.056	0.169	− 0.25
Functional impairment—family	0.43	0.39	0.18	0.26	0.38	0.33	0.41	0.44	− 2.69	0.012	0.169	− 1.01
Global peer relations	3.40	0.63	3.71	0.35	3.39	0.52	3.55	0.32	2.00	0.056	0.169	0.39
Peer acceptance/rejection	3.43	0.67	3.65	0.41	3.43	0.54	3.56	0.37	0.97	0.340	0.379	0.21
Bullied by peers	1.63	0.69	1.24	0.33	1.64	0.56	1.47	0.41	− 2.11	0.044	0.169	− 0.45
Social withdrawal	0.72	0.38	0.57	0.35	0.64	0.50	0.64	0.42	− 1.24	0.226	0.291	− 0.52
Asocial	0.43	0.42	0.41	0.36	0.41	0.46	0.43	0.39	− 0.29	0.774	0.774	− 0.12
Social inhibition	1.00	0.57	0.73	0.46	0.87	0.60	0.84	0.53	− 1.35	0.187	0.281	− 0.60
Prosocial behavior	1.38	0.34	1.53	0.30	1.32	0.38	1.40	0.32	1.10	0.281	0.338	0.26
	PCIT-ED (N = 18)				Parenting wisely (N = 15)				PCIT-ED vs. parenting wisely
	Baseline		Post		Baseline		Post
	Mean	SD	Mean	SD	Mean	SD	Mean	SD	t	p	FDR p	ES
PSI
Total stress	235.2	55.58	207.9	50.70	220.0	56.44	212.0	56.79	− 1.23	0.230	0.288	− 0.61
BDI-II
Maternal depression	9.78	9.30	6.22	5.17	10.20	10.75	8.93	9.60	− 1.00	0.325	0.325	− 0.28
	PCIT-ED (N = 17)				Parenting wisely (N = 15)				PCIT-ED vs. parenting wisely
	Baseline		Post		Baseline		Post
	Mean	SD	Mean	SD	Mean	SD	Mean	SD	t	p	FDR p	ES
LOT-R Total	17.12	3.66	18.82	2.56	17.13	3.85	15.13	4.17	3.30	0.003	0.014	1.02
Optimism	10.00	1.73	10.53	1.74	9.93	1.91	9.47	1.88	1.46	0.155	0.258	0.45
Pessimism	4.88	2.52	3.71	2.37	4.80	2.83	6.33	2.92	− 2.95	0.007	0.017	− 0.90

Bolded values indicate those that are statistically signficant (p < 0.05)

Subscales in italics

Exploratory analyses probing the effectiveness of PCIT-ED delivered in-person versus remotely compared participants with a post assessment who completed 75–100% of their PCIT-ED sessions remotely (n = 12), participants who completed 0–44% of their PCIT-ED sessions remotely (n = 6), and participants in the PW group who completed 6 or more PW sessions (n = 15). Relative to the PW group, participants who completed most of their PCIT-ED therapy sessions remotely demonstrated significantly better outcomes for externalizing symptoms (F = 6.76, p = 0.015, FDR p = 0.039), internalizing symptoms (F = 6.65, p = 0.016, FDR p = 0.039), functional impairment-child (F = 5.90, p = 0.022, FDR p = 0.039), functional impairment-family (F = 6.36, p = 0.018, FDR p = 0.039), and global peer relations (F = 5.03, p = 0.034, FDR p = 0.048). There were no significant differences in outcome measures in analyses that compared in-person PCIT-ED to PW or in-person PCIT-ED to remote administration of PCIT-ED (see Table 5 for full model details). However, as the number of participants in the in-person PCIT-ED group is particularly small, these null findings should be interpreted with caution due to limited power.

Table 5.

Comparisons of post characteristics in Jennings in-person PCIT-ED (3 + sessions), remote PCIT-ED (3 + sessions), and parenting wisely (6 + Sessions) groups covarying for baseline characteristic, child sex, and baseline age

HBQ-P	Change from baseline to post						In-person vs parenting wisely			In-person vs. remote			Remote vs. parenting wisely
	In-person PCIT-ED (N = 6)		Remote PCIT-ED (N = 11)		Parenting wisely (N = 15)
	Mean	SD	Mean	SD	Mean	SD	F	p	FDR p	F	p	FDR p	F	p	FDR p
Externalizing	− 0.11	0.30	− 0.22	0.25	− 0.06	0.22	0.45	0.507	0.821	1.80	0.192	0.448	6.76	0.015	0.039
Internalizing	− 0.03	0.13	− 0.24	0.20	− 0.05	0.17	0.01	0.933	0.933	3.98	0.057	0.399	6.65	0.016	0.039
Func. impair.—child	− 0.25	0.39	− 0.32	0.35	− 0.21	0.35	0.20	0.660	0.821	1.86	0.184	0.448	5.90	0.022	0.039
Func. impair.—family	− 0.27	0.23	− 0.24	0.31	0.03	0.28	2.31	0.140	0.821	0.12	0.734	0.856	6.36	0.018	0.039
Global peer relations	0.20	0.37	0.36	0.41	0.16	0.41	0.46	0.503	0.821	1.02	0.322	0.563	5.03	0.034	0.048
Social withdrawal	− 0.13	0.35	− 0.16	0.18	− 0.00	0.34	0.15	0.704	0.821	0.39	0.537	0.752	1.89	0.181	0.211
Prosocial	0.26	0.20	0.10	0.28	0.08	0.28	0.41	0.529	0.821	0.02	0.877	0.877	1.04	0.317	0.317
	Change from baseline to post						In-person vs. parenting wisely			In-person vs remote			Remote vs. parenting wisely
	In-person PCIT-ED (N = 6)		Remote PCIT-ED (N = 12)		Parenting wisely (N = 15)
	Mean	SD	Mean	SD	Mean	SD	F	p	FDR p	F	p	FDR p	F	p	FDR p
Stress: PSI total stress	− 38.5	49.0	− 21.7	36.5	− 8.00	20.1	1.14	0.295	0.542	0.13	0.719	0.910	0.95	0.339	0.339
Depression: BDI-II	− 4.50	14.0	− 3.08	9.20	− 1.27	4.30	0.05	0.825	0.825	1.61	0.216	0.648	2.14	0.155	0.290
Optimism: LOT-R	1.00	2.19	0.27	2.20	− 0.47	2.39	0.86	0.361	0.542	0.01	0.910	0.910	1.78	0.193	0.290

General Discussion

The purpose of the present study was to examine the effectiveness of a shortened 8-session form of PCIT-ED delivered as an early intervention to children with or at elevated risk for a broad range of psychopathology living in poverty. Following the intervention that families accessed through their local public schools, we observed significantly better outcomes in children randomized to the PCIT-ED group relative to the PW active control group in each of three target domains: mental health, functional impairment, and social functioning. Specific outcomes included lower externalizing symptoms, lower functional impairment in family contexts, and greater positive relationships with peers. Caregiver optimism was also greater in those who participated in PCIT-ED. Similar outcomes were found when restricting the sample to a subset of parent–child dyads who completed the majority of their PCIT-ED sessions remotely, providing evidence that PCIT-ED can be effectively administered via video-conferencing/Zoom. Together, these findings provide evidence for the effectiveness of this shortened 8-session form of PCIT-ED, delivered in-person or remotely, as an early intervention to improve symptoms of psychopathology and functioning in low-income community children at high risk for experiencing psychopathology.

The primary findings of lower rates of externalizing behaviors and impairment align with promising outcomes of other early interventions, many of them relatively brief (e.g., [20, 21]), suggesting that interventions such as PCIT-ED are a highly feasible pathway to enact change. The current study extends findings of the effectiveness of PCIT-ED as one such early intervention that targets both externalizing as well as depressive symptoms, delivered through schools to families for whom access to high-quality early intervention is urgently needed and yet often not readily available. The overall better outcomes observed for functioning and peer relations, in addition to internalizing and externalizing behaviors, further support the use of PCIT-ED to target a broad range of psychopathology and offer multiple—potentially synergistic—mechanisms that underlie the efficacy of PCIT-ED. These are significant strengths of the intervention and, taken in conjunction with the established history and widespread use of standard PCIT in clinical contexts and therefore the availability of trained clinicians both nationally and worldwide [22, 24], offer strong support in favor of using PCIT-ED as an early intervention capturing families through the school setting on a broad scale.

Although the intervention did not significantly affect outcomes related to caregiver depressive symptoms or stress, greater caregiver optimism was observed in those who participated in PCIT-ED. Optimism has been shown to have protective effects on parental well-being and mental health (e.g., [49]) and increasing caregiver optimism thus has the potential to have positive downstream effects on parental mental health and parenting. A closer examination of caregiver optimism from baseline to post-intervention indicates that whereas caregivers in the PCIT-ED group exhibited greater optimism following the intervention, caregivers in the PW group exhibited lower optimism. Moreover, these differences occurred primarily for pessimistic or negative expectations about the future, rather than positive expectations. For example, at the post-intervention assessment, caregivers in the PW group were more likely than those in the PCIT-ED group to endorse statements indicating they expected bad things to happen to them in the future and that they hardly ever expected things to go their way. One explanation for these findings is that PCIT-ED might have provided caregivers with direct feedback on improvement in their parenting skills and/or experiences of their child’s improved behavior that fostered hope for further positive outcomes in the future. In contrast, caregivers in the PW group were unlikely to receive any feedback about their parenting, and if they failed to notice any improvements from the intervention that might have reinforced notions that their parenting and/or child’s behavior is unmalleable, and thus future improvements were unlikely. This might suggest that the experience of improvement through direct interactions with a therapist and with the child in therapy is an essential ingredient to parental positivity.

PCIT-ED was chosen as the intervention in the present study based on prior studies indicating that it targets internalizing symptoms more directly than standard PCIT [38]. We also choose it because it focuses on emotion development, a key component of adaptive functioning that is not addressed in standard PCIT. Although we observed more outcomes related to lower externalizing than internalizing symptoms in the present analyses this may be because these symptoms were more common in the study population and more easily addressed by parents and teachers who rated the behaviors. It is well known that internalizing symptoms are often overlooked due to their non-disruptive nature, therefore there may have been improvements in these domains that were not detected using the brief survey measures we employed.”

Whereas the timing of the COVID-19 pandemic and associated stay-home orders that occurred partway through this study makes some findings difficult to interpret, it also provided a unique opportunity to examine the feasibility and effectiveness of administering PCIT-ED remotely via zoom video conferencing. Indeed, exploratory analyses indicated that not only did children who completed the majority of their PCIT-ED sessions remotely show the same positive outcomes with regards to externalizing symptoms, functional impairment in family contexts, and peer relations as the larger sample, but this subset also showed positive outcomes regarding internalizing symptoms and functioning beyond family contexts. These findings build on pilot work demonstrating increased efficacy of remote-delivery of standard PCIT over in-person PCIT [50] and are of critical importance in our rapidly changing digital world where healthcare and telemedicine are becoming an accepted modality of care. Furthermore, the ability to access effective early interventions remotely might be of particular benefit to low-income families who have less time, limited access to transportation, and fewer financial resources to enable clinic appointments. For example, some benefits of participating in PCIT-ED remotely include the potential for more flexible appointment times, no transportation costs or need to factor in travel time to and from the appointment, potentially not needing additional childcare for siblings during the session (especially when the siblings are older), and the ability for the therapist to see the family in their natural home environment. In contrast, some drawbacks to remote administration of PCIT-ED include distractions from siblings, pets, or others in the home during the sessions, difficulties forming a therapeutic relationship with the therapist over a computer screen, difficulties keeping a young child engaged and present on screen throughout the session, and difficulties ensuring an appropriate electronic device for video-conferencing and reliable internet connection. More work is clearly needed to evaluate the relative costs and benefits of each of these factors and to develop methods to overcome these obstacles as well as to determine which method is best for which families. However, in our (limited) experience, we found that the switch from in-person to remote administration of PCIT-ED went smoothly, and anecdotally, families in our study found the remote sessions more logistically feasible and enjoyed the flexibility they provided. Our experiences largely mirror those described by other therapists who transitioned to remote delivery of PCIT during the pandemic (e.g., [51, 52]).

Limitations and Future Directions

There are several limitations of the present study that are important to consider. First, likely due to the COVID-19 pandemic, sample attrition was higher and more variable than anticipated. Although the intent-to-treat analysis was used to account for noncompliance and missing outcomes, it is unclear how representative the present compliance and completion rates are and whether they can be used to inform future studies. Second, although the study employed PW as an active control, there were some fundamental differences between PW and PCIT-ED worth noting. For instance, unlike PCIT-ED, PW does not target emotion regulation or include therapist contact and coaching, both of which might have provided key benefits to children and families enrolled in PCIT-ED. Moreover, it became clear throughout the study that participation in PCIT-ED constituted a substantially larger burden for families in terms of time, scheduling, and physical and mental resources. This is also evident in the completion rates for each condition—whereas families assigned to PW tended to complete either all or none of their sessions (i.e., once they started they tended to finish), the distribution of sessions completed by families assigned PCIT-ED was much more variable. The findings from secondary analyses that restricted the sample to participants who completed 3 or more PCIT-ED or 6 or more PW sessions largely mirrored those from the intent-to-treat analysis, indicating that variability in number of sessions completed did not drive the main study findings. However, the optimal number of sessions needed to obtain significant beneficial outcomes remains unknown from these data. Third, there was substantial variability across participants with regard to the time between the end of the intervention and the post-intervention assessment, largely due to the COVID-19 pandemic. This is a strength of the study because it suggests that improvements due to PCIT-ED may have lasting effects; however, it also may underestimate the immediate benefits of PCIT-ED and obscures our ability to accurately discern the time course of improvements following the intervention. A more systematic investigation of the immediate and longer-lasting impact of PCIT-ED on psychopathology and functioning should be examined in future studies. Fourth, two of the six key findings (functional impairment in family and social withdrawal) did not remain significant when we limited the analysis to one child per family, raising the possibility that the initial positive outcomes detected in these domains in the main analysis was due to dependency from the sibling pairs. Future studies with increased power are needed to address that question. Finally, study findings would be strengthened by investigating whether change in child behavior as a function of treatment can be detected using observational measures. Although an observational parent–child interaction task was administered, the sample size was quite reduced due to COVID-19 related study delays and attrition, and as a result these exploratory data will be reported elsewhere.

To facilitate the delivery and uptake of the intervention, in the present study the research team worked closely with school administrators and other community stakeholders to provide access to the intervention through the local public schools. The study also employed therapists that were from the community in which the intervention was being tested. Demonstrating efficacy of PCIT-ED with this approach was an important first step; however, further adapting these methods to train school-based counselors or social workers who are already embedded in the school system will be an important next step toward developing an intervention that is not only efficacious, but also more pragmatic and sustainable. Similarly, assessing the cost effectiveness of different aspects of the intervention, such as varying the level of therapist supervision from global trainers, will be important to calibrating an effective, feasible, and sustainable early intervention program that is accessible to families at a time in early development when the intervention may have the strongest impact.

Another important task for future research will be identifying predictors of whether in-person or remote treatment is likely to be most beneficial to children and families. It is likely that accessibility to resources plays a large role in which modality is most beneficial to families; for instance, in-person treatment requires access to transportation and childcare, whereas remote treatment requires access to technology in the home. However, there are likely other individual difference factors that might also predict which modality should be pursued; for instance, future research should examine possible child factors such as ability to attend in a remote therapy setting, and possible parent factors such as comfort in the home-environment versus an inperson therapy clinic. Such identified predictors could be used by clinicians alongside family preference to help aid treatment recommendations. Remote treatment also has the unique benefit of providing the clinician with a window into the home environment and providing an ecologically more valid context, which may be more useful in some therapeutic contexts than others. Moreover, the small sample size of the in-person group and null findings regarding differences between the remote versus in-person groups limit the conclusions that can be drawn directly from the present findings regarding the effectiveness of short-term PCIT-ED administered in-person in school settings. However, if access is not a limiting issue, we suggest that the value of an in person therapeutic relationship cannot be overemphasized.

Conclusion

In sum, the present study provides compelling evidence for the benefits of PCIT-ED for young children with or at risk of psychopathology, and the feasibility of administering this shortened eight-session form of PCIT-ED as an early intervention delivered through the school setting. Moreover, the evidence that remote administration of PCIT-ED is at least as effective as in-person administration provides an exciting new direction for early intervention work, potentially increasing access to families for whom in-person therapy is either not available or feasible. Taken together, findings demonstrate the effectiveness and feasibility of a shortened PCIT-ED implemented in a community setting targeting low-income and minority families at high-risk for experiencing psychopathology; where mental health care is urgently needed but rarely available. Future studies are now needed to further inform the optimal implementation methods and number of sessions needed for to be effective. Advances in this domain may lead to brief and feasible preventive interventions that can be accessed easily by populations in need to fortify critical child developmental outcomes for children living in poverty.

Summary

The present study investigated the feasibility and efficacy of delivering a shortened form of an empirically supported parent-child psychotherapy, Parent Child Interaction Therapy-Emotional Development (PCIT-ED), as an early intervention for 3- to 6-year-old children with or at elevated risk for a broad range of psychopathology who were growing up in low-income communities. Caregiver–child dyads were randomized to participate in either a shortened eight-session form of PCIT-ED or an online parenting education of comparable length. PCIT-ED was initially administered in-person but shifted to remote administration partway through the study due to the COVID-19 pandemic, providing a unique opportunity to explore the efficacy of PCIT-ED delivered remotely. Target outcomes included children’s mental health, functional impairment, and social functioning, along with caregiver’s mental health, stress, and optimism. Relative to the parent education group, children in the PCIT-ED group exhibited lower externalizing symptoms and functional impairment and more positive peer relationships following the intervention. Caregiver optimism was also greater in those who participated in PCIT-ED. These findings were similar when restricting the sample to a subset of parent-child dyads who completed the majority of their PCIT-ED sessions remotely, demonstrating efficacy of remote administration of PCIT-ED. Together, these findings support the effectiveness of this shortened form of PCIT-ED, delivered in-person or remotely, as an early intervention to improve symptoms of psychopathology and functioning in high-risk children living in poverty.

Data Availability

Data and materials are available upon request.