Autism spectrum disorder (ASD) is a neurodevelopmental condition characterized by deficits in social communication and reciprocity as well as stereotypic behaviors and interests (American Psychiatric Association,
2022). Families of young children with autism often face myriad co-occurring conditions. One of the most prevalent issues is child disruptive behaviors which occur in 12–48% of the population (see Hossain et al.,
2020, for a review; Mazurek et al.,
2013). These behaviors may include anger outbursts, irritability, aggression, noncompliance, and oppositionality (Burke et al.,
2002). Autism researchers have long posited that problem behaviors in ASD function in the context of multiple factors including adaptive functioning/social communication deficits, reinforcement in the environment, and broader predisposing biological characteristics (Kanne & Mazurek,
2011; Smith et al.,
2016). Although disruptive behavior in youth with ASD may clinically differ from typically developing youth (Beauchaine et al.,
2010), research indicates that approximately one in four children with ASD meets criteria for a comorbid disruptive behavior disorder (Kaat & Lecavalier,
2013). The presence of these disruptive behaviors can predict higher levels of parental stress, reduce a child’s ability to benefit from early intervention services, and impede a child’s school functioning (Postorino et al.,
2019; Soke et al.,
2018). Furthermore, disruptive behaviors may also limit a child’s success in other ASD-specific treatment modalities meaning even if the child does have access to care, they may not benefit as much as their counterparts without such concerns (Jang et al.,
2011).
Analysis Plan
Comparisons between demographic characteristics were performed using independent samples t-tests and chi-square analyses to determine if differences arose between families with or without a child with ASD/DD. To examine potential differences in treatment processes between groups, independent samples t-tests were conducted on the number of days participants were in treatment and the number of PCIT sessions attended. Therapist-reported relationship enhancement and rates of graduation were examined between groups using chi-square analyses.
Differences in pre- to post-treatment ECBI scores between groups were compared using a mixed analysis of variance. While our demographic comparisons and descriptive statistics reflect the entire sample, families with fewer than four sessions (
N = 1,490) and/or with missing data (
NIntensity = 733,
NProblem = 745) were excluded from analyses of ECBI scores (some families fitting into both categories). A total of 804 families were included in EBCI Problem Scale analyses and 769 families were included in ECBI Intensity scale analyses. In previous work with this sample, four complete sessions were estimated to be the smallest possible dose of PCIT for meaningful behavior change and skill acquisition to occur (Lieneman et al.,
2019). It was assumed that the first three PCIT sessions typically cover intake assessments and introduction to parent training (CDI teach). Following CDI teach, families receive parent coaching. It is also assumed that only families who received at least one coaching session may experience change from participating in PCIT; therefore, change would be observed at the fourth session. Preliminary analyses explored if pre-treatment differences were present for ECBI scores between families who attended less than four sessions versus those attended four sessions or more. Outcomes indicted no significant differences between groups on either the ECBI Problem or Intensity subscales (all
p-values > 0.05).
All families were included in the present sample rather than using a matched sample comparison as it ensured full representation of families who received services within the Oregon agencies (rather than a select group) during this period. Due to limited demographic information, the research team was also concerned that matched samples still would not accurately represent samples that were similar. For example, given the large portion of participants with “unknown” racial identity, matching may have either (1) lumped this broad classification as one group therefore unintentionally creating a homogenization of racially or ethnically diverse families or (2) required that these individuals not be included in the matching process, again possibly unintentionally eliminating the experiences of racially or ethnically diverse families. Further, the intention of this project was to see how outcomes of individuals with autism or developmental delays in PCIT compared to all other families receiving PCIT (rather than just exploring the experiences of only a few families). Therefore, all eligible families were included in the analysis plan.
Discussion
The current study examined PCIT delivered for children with and without ASD/DD across community-based clinics in Oregon. Findings revealed significantly more males receiving PCIT than females for children with ASD/DD as compared to children without ASD/DD. This proportion is expected as ASD is diagnosed more frequently in males than females (1 male per every 4 females; CDC,
2020). Outcomes also yielded significant reductions in parent-reported frequency of disruptive behaviors (ECBI Intensity scores) and parental perceptions of problematic child behaviors (ECBI Problem scores) among all children who received at least four sessions of PCIT. Additionally, no group differences were observed for length of time in treatment, number of sessions attended, therapist-reported relationship enhancement or graduation rates, or in Intensity or Problem scores from pre- to post-treatment. This study represents one of the first wide-scale explorations of the delivery of standard PCIT for children with ASD/DD by non-specialized clinicians at the community level (see also McInnis et al.,
2020).
To determine similarities and differences in how long families were involved in PCIT, group comparisons were explored for the length of time spent in treatment as well as the number of sessions attended. No differences arose between families of children with ASD/DD and the comparison group. These outcomes indicate that treatment was not extended for families of children with ASD/DD which is reflective of several PCIT studies (see Owen et al.,
2019 for a review), although increased number of sessions has been found in previous literature (Zlomke et al.,
2017). Families of youth with and without ASD/DD also graduated from PCIT at similar rates. Given the additional challenges that can arise and adaptations that are frequently required when implementing PCIT with ASD/DD children (Owen et al.,
2019), these findings are promising for community-based clinics. Of note, clinicians in the present study were not known to have extensive experience or expertise in ASD/DD. Additionally, dissimilar to other research exploring PCIT with ASD/DD populations, clinicians in this study were not part of a university-based research team thus demonstrating PCIT’s utility within real-world settings (McInnis et al.,
2020).
Next, families with and without children with ASD/DD were compared on therapist-reported parent-child relationship enhancement. No differences were found indicating therapists perceived parents and their children with ASD/DD as demonstrating similar levels of relationship enhancement compared to the comparison group. Although the measure used in the present study was rudimentary, these findings are valuable for several reasons given that only a portion of the sample of youth was diagnosed with ASD. First, children with ASD are characterized as having deficits in the development, maintenance, and comprehension of relationships (APA,
2022). Clinical improvements in these relationships from the present study demonstrate families’ clear ability to develop and improve this important connection (Thompson & McFerran,
2013). Second, autism literature has previously portrayed parents of youth with autism as feeling burdened and as having poor parent-child relationships (Hock & Ahmedani,
2012; Picardi et al.,
2018); yet, due to PCIT’s focus on relationship-building, our results demonstrate that these positive effects are not limited to children without ASD/DD. Furthermore, older literature on attachment has previously concluded that children with autism form secure attachments less frequently than neurotypical youth, with implications for the presence and quality of the parent-child relationship (Rogers et al.,
1991; Shapiro et al.,
1987). While more modern views are changing these discussions (van IJzendoorn et al.,
2007), this continued exploration of the parent-child relationship is critical to unpacking
how parents and their children with ASD or other DDs relate rather than
if they relate (Beurkens et al.,
2012).
In addition to similarities in treatment involvement, children with ASD/DD in the present study demonstrated significant improvements in disruptive behavior (as characterized by reductions in Problem and Intensity scores on the ECBI) which mirrored the drops seen in the comparison sample. Of note, the current study’s sample of ASD/DD youth (on average) did not get to the targeted Intensity scores sought to graduate from PCIT (i.e., 114 [1.5
SD below the clinical cutoff of 131]) but the comparison sample did. The established ECBI clinical cutoff scores (131 for the Intensity Scale, 15 for the Problem Scale; Colvin et al.,
1999) were derived from studies of typically developing youth ages 2–16 years. Research by Jeter and colleagues (
2017) found that among families completing the ECBI, parents of children with ASD reported a higher rate of disruptive behaviors and more frequently perceived these behaviors as problematic compared to parents of typically developing youth. While several past case studies of PCIT efficacy among children with ASD have observed ECBI score reductions to nonclinical (i.e., 114) levels following treatment (e.g., Armstrong et al.,
2015; Masse et al., 2016), this result has not been replicated in studies with larger sample sizes (e.g., Ginn et al.,
2017; Solomon et al.,
2008). However, the present study yielded no differences in Intensity scores between the two groups even with the large sample. This lack of difference in post-scores between the two groups points to improvements of the ASD/DD group still falling to near-normative levels. Moreover, given the previous discussions of restandardization of the ECBI for ASD populations (Jeter et al.,
2017), these outcomes are even more impressive within this community-based setting.
These findings should be considered in light of the present study’s limitations. Firstly, this study was conducted with secondary data meaning study researchers did not manipulate which treatment was provided or when, how clinicians were trained, if alternative treatments would be more effective, or what other services youth with ASD were receiving within the state of Oregon. Therefore, we are unable to state the effectiveness of PCIT as compared to other treatment modalities for youth with ASD. However, for families receiving PCIT in the state of Oregon, we are able to say youth with ASD/DD appear to have similar outcomes on the measured variables as compared to their peers without ASD/DD.
Another limitation of the present study was that families were excluded from analyses if they had missing data. It is possible that participants with missing data or those who did not attend at least four sessions represent a meaningful subset of families participating in PCIT. Our sample also had only a small number of girls with ASD/DD. While child sex has not been found to significantly relate to treatment outcomes in the general PCIT literature, girls tend to pursue treatment for disruptive behaviors at lower rates (Bussing et al.,
2003). PCIT treatment efficacy among young girls with ASD should be emphasized in future research.
It is important to note that just under 20% of families in our sample graduated from treatment, with over 80% failing to meet graduation criteria (see Table
3). This proportion was comparable for the comparison and ASD/DD groups. Community-based PCIT research has consistently reported high rates of treatment attrition (e.g., Lyon & Budd,
2010), with many families engaging in PCIT only receiving part of the treatment. Although substantial research demonstrates PCIT’s positive outcomes for families who complete it, more recent studies have revealed significant improvements with medium-to-large effect sizes even in families who terminate treatment early (Lieneman et al.,
2019; Stokes et al.,
2018). Past research has also implicated increased parental stress and treatment barriers (Fernandez & Eyberg,
2009) as rationale for the high attrition rates in community-implemented PCIT. Although parents of youth with ASD experience higher rates of stress and increased barriers to treatment compared to caregivers of neurotypical children (Chiri & Warfield,
2012; Hayes & Watson,
2013), we observed no significant differences in graduation rates between these groups. Indeed, factors affecting treatment completion were likely present among families completing PCIT; however, these results are promising as they indicate families of youth with ASD and/or DD in the community can graduate and benefit from PCIT at similar rates to their neurotypical peers. Nonetheless, future efforts should investigate unique treatment barriers and perspectives among families of children with ASD/DD receiving PCIT.
We also recognize that outcomes from this sample cannot be solely attributed to differences between autism and non-ASD populations. Due to uncertainty of diagnostic procedures, we decided to include all children diagnosed with a DD to be included in the ASD group in case the system of identification was unintentionally excluding children with ASD and classifying them as a more conservative DD diagnosis. Alternative explanations for DD diagnoses of these children include that it served as a catch-all for ASD-related behaviors because families did not have a proper evaluation conducted in a specialty clinic; providers knew the child was not neurotypical but did not have the expertise to narrow down a specific neurodevelopmental diagnosis; the diagnosis was provided by a primary care provider to mark a concern but did not necessarily serve as a proper diagnosis; or the diagnosis indicated concerns without additional follow-through, given that PCIT providers were not allowed to bill for these diagnostic codes. By doing so, conclusions from this study cannot be fully generalized to ASD populations. Additionally, it is possible that our current diagnostic data underestimate the actual number of children with ASD/DD in our sample. For medical insurance and reimbursement, PCIT is classified as an evidence-based treatment for disruptive behaviors, and noncompliance and disruptive behaviors are the primary conditions eligible for coverage. In Oregon, PCIT is not classified as an approved treatment for youth with ASD (Oregon Health Authority,
2020). Thus, community providers may not bill for PCIT services if a child only has a diagnosis of ASD or another neurodevelopmental condition, despite high rates of comorbid disruptive behavior and PCIT’s evidence of effectiveness with youth with ASD. Indeed, the current study contributes to the growing body of work demonstrating PCIT’s promising outcomes for children with ASD, particularly at the community level. Based on this evidence, it is recommended that governing health agencies should inform their billing policies to reflect PCIT’s benefit for youth with ASD and disruptive behavior. It is also advised that future researchers studying the dissemination of PCIT for youth with ASD form strong partnerships with community providers and oversee data reporting procedures to enhance study fidelity.
Additionally, all outcome measures in this study were based on therapist and caregiver reports. Therapist report of relationship enhancement may be biased towards expected positive change in the parent-child relationship following treatment, and caregiver reports of disruptive behavior improvement may be inflated due to the investment of time and effort in therapy (Arkes & Blumer,
1985). It is also possible that outcomes from PCIT are meaningfully impacted by agency or clinician-level variables. Extensive literature in the field of dissemination and implementation science exists (e.g., Exploration, Preparation, Implementation, Sustainment [EPIS] framework) which recognizes how inner context factors (e.g., leadership, organizational characteristics, individual clinician characteristics) may impact the successful implementation of evidence-based treatments such as PCIT (Aarons et al.,
2011). However, little work has been done to recognize how these factors impact treatment implementation for youth with ASD. Future research efforts should include multiple informant outcome measures and standardized behavioral measures as well as agency- or clinician-level variables to better assess PCIT outcomes for children with ASD/DD.
Future directions in this field should investigate a wide range of treatment outcomes important for families of youth with ASD and/or DD in the community and include follow-up research for maintenance of treatment effects. Several studies have provided preliminary evidence for PCIT’s promise in improving not only child disruptive behavior, but also language production, adaptive skills, and prosocial behavior among youth with ASD (e.g., Ginn et al., 2015; Hansen & Shillingsburg,
2016; Zlomke et al.,
2017). Although few studies have utilized follow-up measures to explore treatment outcomes among youth with ASD or DD following PCIT, the preliminary evidence is promising. In their study of 30 mother-child dyads, Ginn and colleagues (2015) found that improvements in child disruptive behavior and social awareness were maintained after 6 weeks following PCIT. Bagner and colleagues (
2010) reported statistically significant decreases in child disruptive behavior and increases in child compliance for 28 children at risk for a DD that maintained 4 months post-PCIT. Several smaller case studies have reported maintained improvements in child behavior at 3- and 5-month follow-ups (e.g., Armstrong & Kimonis,
2013;
2015; Masse et al., 2016). Research utilizing a range of relevant outcome and follow-up measures is crucial in assessing the maintenance of treatment gains over time. This work is of particular importance in the exploration of PCIT delivered for youth with ASD and/or DD at the community level.
Further, ASD is widely regarded as a heterogenous condition and patterns of cognitive, behavioral, and social functioning may manifest differently across children (APA, 2013). Youth with ASD often have comorbid intellectual disability/global DD and language delays (Matson & Shoemaker,
2009; Munson et al.,
2008). In the past, researchers have noted that PCIT may not be suitable for youth with ASD with limited receptive language capability, given the emphasis on parent-child communication (Masse et al.,
2007). However, small-scale studies have begun to demonstrate the benefits of PCIT among youth with ASD and comorbid DD, reporting that positive outcomes include increased child vocalizations (Hansen & Shillingsburg,
2016). As the youth in our sample were identified by clinicians as having ASD and/or DD, it is possible that children with limited receptive language skills displayed significant improvements in disruptive behavior following PCIT. However, without adequate assessment of language skills, intellectual functioning, and/or ASD symptom severity, little is known about the clinical characteristics of the children in this group. Given the diversity of clinical presentations and child characteristics in youth with ASD, research of within-group differences for children receiving PCIT is needed.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.