Adapting to Unprecedented Times: Community Clinician Modifications to Parent–Child Interaction Therapy During COVID-19

ABSTRACT

Parent–Child Interaction Therapy (PCIT) is an evidence-based practice that effectively prevents and treats child disruptive behaviors and child physical maltreatment and reduces parenting stress. PCIT was adapted for telehealth delivery, internet-delivered PCIT (iPCIT), before the COVID-19 pandemic but was not widely implemented until the rapid transition to telehealth during stay-at-home orders. To understand how clinicians adapted PCIT during COVID-19, we followed up on a previous study investigating community clinician adaptations of PCIT pre-COVID-19 using the Augmenting and Reducing Framework. Clinicians (N = 179) who responded to the follow-up survey and reported delivering PCIT remotely completed a quantitative measure of adaptations at both time points (Fall 2019; Summer 2020) to assess how adaptations to PCIT changed following lockdown measures. Clinicians (n = 135) also provided qualitative descriptions of adaptations made early in the COVID-19 pandemic. Clinicians in the full sample were 74.3% Non-Hispanic White and 14% Latinx. Most clinicians had a master’s degree (66.5%), were licensed (80.4%), and were PCIT-certified (70.4%). Paired sample t-tests showed that clinicians reported similar levels of augmenting (t(179) = -0.09, p = .926) and reducing adaptations (t(179) = -0.77, p = .442) at both time points. Unlike quantitative findings, qualitative findings indicated that clinicians described engaging in many types of adaptations in response to the pandemic. Clinicians discussed engaging in augmenting adaptations by extending treatment length and integrating other practices into treatment. Clinicians also discussed engaging in reducing adaptations. Implications and future directions will be discussed.

The onset of COVID-19 presented unique challenges for families across the United States. With the implementation of physical distancing measures and the downturn of the economy, caregivers responsible for the well-being of their children experienced increased levels of stress (Cameron et al., 2020; Riegler et al., 2020). In times of heavy stress and trauma, children may mirror their parents’ emotional well-being by reflecting their parents’ worries, concerns, and trauma reactions (Goodman & Gotlib, 1999; Kerns et al., 2014). Further, stress limits a caregiver’s ability to interact with and respond to their children in consistent and positive ways, making parent–child bonds susceptible to disruption (Kerns et al., 2014; Sigelman et al., 2018) and increasing the risk of child maltreatment (Teo & Griffiths, 2020). Overall, caregiver distress during a crisis has been shown to predict increased child mental health symptoms (e.g., externalizing behaviors) even after controlling for shared traumatic experiences (Kerns et al., 2014). As a result, COVID-19 increased the need for access to quality mental health evidence-based practices (EBPs) for parents and children experiencing high levels of stress (Gurwitch et al., 2020; Imran et al., 2020). To make this access possible, clinicians were tasked with the challenge of transitioning to providing EBPs for children and families remotely (i.e., telehealth). The current study seeks to understand how clinicians modified their delivery of one EBP, Parent–Child Interaction Therapy (PCIT), during the first months of the COVID-19 pandemic.

PCIT

PCIT has strong evidence demonstrating its ability to support families with young children presenting with disruptive behaviors such as tantrums and aggression (Lieneman et al., 2017; Ward et al., 2016; Webb et al., 2017). PCIT clinicians coach parents in-vivo through a 1-way mirror, working to strengthen the parent–child bond and establish consistent approaches to limit setting (Funderburk & Eyberg, 2011; Kaminski & Claussen, 2017). PCIT consists of two phases – Child-Directed Interaction, or CDI, and Parent-Directed Interaction, or PDI (Funderburk & Eyberg, 2011). The first phase, CDI, focuses on supporting the caregivers’ use of positive parenting skills (i.e., PRIDE skills, including praise, reflections, and descriptions of appropriate child behaviors) and building a strong foundation for the parent–child relationship (Funderburk & Eyberg, 2011). The second phase, PDI, provides caregivers with consistent and safe disciplinary strategies (i.e., time-out) that do not risk endangering the parent–child bond (Funderburk & Eyberg, 2011). The empirically based intervention follows a clear protocol and uses standardized measures to monitor progress and assess clinical outcomes (Funderburk & Eyberg, 2011). Not only has PCIT proven effective in preventing and reducing externalizing and internalizing symptoms (Chase & Eyberg, 2008; Niec et al., 2016) and child maltreatment (Chaffin et al., 2004), but also in lowering caregiver stress (Niec et al., 2016) and depression (Urquiza & Timmer, 2012), and in improving parental responsiveness to their children (Niec et al., 2016).

PCIT adaptations

Given the benefits of PCIT for children and families, PCIT has been adapted formally in clinical trials to ensure that clinical outcomes remain intact when changes are made to protocol to enhance its reach and acceptability for families (Thomas et al., 2017). Eyberg (2005) identified how PCIT can be tailored, adapted, and modified. By its nature, tailoring is considered inherent to PCIT to make sure core components are delivered in a manner that meets the needs of a family, for instance helping a parent praise behavior in their child that is culturally important to them (Eyberg, 2005). Modifications include universal changes to the treatment protocol (Eyberg, 2005). Adaptations to PCIT include changes to the content or structure to enhance feasibility or effectiveness with new populations (Eyberg, 2005). For instance, Guiando a Niños Activos (GANA), or Guiding Active Children is a culturally adapted version of PCIT that has shown efficacy in decreasing attrition and decreasing externalizing and internalizing symptoms for Mexican American families (K. McCabe et al., 2012; K. M. McCabe et al., 2005). GANA aimed to increase the cultural relevance of PCIT by reframing the intervention, emphasizing rapport, and including culturally appropriate materials to reduce stigma and increase parental engagement. PCIT has also been adapted for populations that present with varying clinical presentations, including children with selective mutism (PCIT-SM; Cotter et al., 2018) and callous unemotional traits (PCIT-C; Fleming & Kimonis, 2018).

Internet-delivered PCIT (iPCIT) is a telehealth adaptation of PCIT, which was developed to leverage technology to increase access to services prior to the COVID-19 pandemic (Comer et al., 2015). iPCIT followed the structure of standard PCIT, but instead of interacting in front of a one-way mirror, clinicians observed parent–child interactions via a webcam and provided coaching through a wireless earpiece, for example, Bluetooth earphones (Comer et al., 2015). iPCIT was shown to be effective in reducing children’s symptoms and barriers to treatment (Comer et al., 2017), but widespread dissemination efforts were yet to take place at the onset of the COVID-19 pandemic. Given its efficacy, iPCIT was positioned as an ideal intervention for clinicians to remotely support children and families during elevated levels of child behavioral problems and parenting stress during the COVID-19 pandemic (Gurwitch et al., 2020). The rapid transition to telehealth presented a unique opportunity to investigate clinicians’ in-the-field adaptations as they learned to implement the existing protocol of iPCIT.

Although formal adaptations to EBPs are typically evaluated empirically, clinicians frequently make ad-hoc adaptations in practice (Lau et al., 2017). Indeed, adaptations to evidence-based protocols are considered inevitable when interventions are implemented in real-world settings (Lau et al., 2017; Wiltsey Stirman et al., 2019). Clinicians working with diverse populations, for example, might be inclined to adapt treatment to better meet the needs of their clients (Lyon et al., 2014). Additionally, studies have highlighted that community clinicians are particularly likely to adapt EBPs in response to clinical presentation (Luis Sanchez et al., 2022), crises, or emergent life events (Barnett et al., 2018; Guan et al., 2019). Emergent life events are significant, unexpected stressors in a client or family’s life, which were especially prevalent during COVID-19, such as having a death in the family or parental job loss (Guan et al., 2019). Past studies have identified that clinicians deliver EBPS with lower levels of fidelity, covering less treatment content with lower intensity, when emergent life events are disclosed in session (Guan et al., 2019; Lind et al., 2021).

Intervention adaptations and emergent life events

Most of the research on clinician responses to emergent life events and subsequent adaptations to EBPs have focused on individual crises and life events (Barnett et al., 2018). However, COVID-19 demonstrated how widescale global events can impact the provision of services. The most evident and immediate adaptations to COVID-19 included a rapid conversion of mental health services to an online format (Green Rosas et al., 2022). However, many additional stressors increased beyond the need to stay-at-home orders, including job losses and the deaths of loved ones (Isasi et al., 2021). Adaptations became a necessary way of remaining responsive to client needs and social realities. While adaptations in times of crisis are necessary and can be an indication of clinician responsiveness to client needs, providing high fidelity EBPs that can adequately mitigate crisis-induced stress is also critical.

Current study

Whereas iPCIT demonstrated efficacy in reducing child difficult-to-manage behaviors before the pandemic (Comer et al., 2017), widespread efforts to disseminate and implement iPCIT in community settings were yet to occur (Gurwitch et al., 2020). Nevertheless, clinicians were tasked with rapidly transitioning to telehealth following social distancing measures (Green Rosas et al., 2022). This transition was supported by PCIT training organizations, which offered resources, including manuals and webinars, regarding how to deliver iPCIT via listservs and websites (e.g., http://www.pcit.org/covid-19-professional-resources.html). Understanding how clinicians modified treatments to both deliver quality evidence-based care while responding to emergent crises during COVID-19 may inform systematic approaches to EBP implementation during potential future moments of crisis.

The current study followed up on a pre-COVID-19 pandemic study investigating community clinician adaptations of PCIT (Luis Sanchez et al., 2022) following two adaptation implementation science frameworks. First, quantitative and qualitative measures were based on previous implementation research finding that community clinicians engage in two kinds of adaptations: Augmenting adaptations and Reducing/Reordering adaptations (Lau et al., 2017). Augmenting adaptations sought to supplement the intervention to meet client needs while making additions to EBPs by tailoring the presentation of strategies, integrating supplemental content, and lengthening the treatment or slowing the pacing (Lau et al., 2017). Reducing/Reordering adaptations, on the other hand, represented a separation from core elements or structure of EBPs driving an intervention’s efficacy; Reducing/Reordering adaptations include omitting components, reordering components, and shortening the treatment or increasing the pacing (Lau et al., 2017). Additionally, the Framework for Reporting Adaptations and Modifications-Enhanced (FRAME) was used to include not only process adaptations (e.g., structure and content), but also reasons for adaptations (Wiltsey Stirman et al., 2019). The current study investigated quantitative and qualitative differences in clinician-reported adaptations of PCIT from pre-pandemic in 2019 (Time 1) and during the transition to iPCIT implementation in 2020 (Time 2).

At Time 1, clinicians reported engaging in Augmenting and Reducing/Reordering adaptations to a moderate and slight extent, respectively (Luis Sanchez et al., 2022). Clinicians reported engaging in adaptations primarily in response to their client’s clinical presentation (Luis Sanchez et al., 2022). Given the contextual changes with the COVID-19 stay-at-home orders, including higher stress for families, we expected to find increases in Augmenting and Reducing/Reordering adaptations to iPCIT at Time 2. Additionally, we expected clinicians’ reasons for adapting iPCIT to be primarily driven by the emergent crises associated with COVID-19 (e.g., job loss, schooling children in the home) at Time 2.

Methods

Participants

This study was part of a follow-up in the summer of 2020 to a larger study conducted in the summer of 2019 investigating clinicians’ experiences implementing PCIT in community settings. Clinicians who provided their contact information in the 2019 study (N = 309) were recruited for participation in the current study via a follow-up e-mail. Of those, 223 responded to the follow-up survey (a 72% response rate), of which 183 (82%) reported they provided PCIT since the transition to telehealth due to the COVID-19 pandemic and were eligible for inclusion in the current study. The final sample comprised clinicians (n = 179) who completed the adaptations measure. One hundred thirty-five clinicians completed open-ended questions expanding on their response to how and why they adapted PCIT. Clinicians who were trained and delivering PCIT in community settings were eligible to participate in the first study (Luis Sanchez et al., 2022). Demographics and clinician characteristics from the previous study are displayed in Table 1.

Table 1. Participant characteristics in baseline and current study sample.

	Baseline Sample (N =  314)	Current Study Sample (N = 179)
Demographics
Age, M (SD; range)	37.02 (8.57; 22–71)	36.61 (8.29; 22–68)
N (%) Female	287 (91.4%)	160 (89.4%)
Race/Ethnicity
Non-Hispanic White	232 (74.4%)	133 (74.3%)
Latinx/Hispanic	45 (14.4%)	25 (14.0%)
Other Ethnicity	35 (11.2%)	20 (11.2%)
Race		n (%) (6 missing)
White	259 (85.2%)	150 (83.8%)
Black/African American	8 (2.6%)	5 (2.8%)
Asian American/Pacific Islander	9 (3.0%)	7 (3.9%)
Multiracial	2 (.7%)	6 (3.4%)
Not Listed	26 (5.6%)	5 (2.8%)
Professional Characteristics
Language of service provision	n (%)	n (%)
English only	245 (78.0%)	145 (81.0%)
Spanish	58 (18.5)	29 (16.2)
Other	11 (3.5%)	5 (2.8%)
Professional Discipline	n (%)	n (%)
Clinical Psychology	107 (34.2%)	65 (36.3%)
Social Work	69 (22.0%)	35 (19.6%)
Counseling	64 (2.4%)	34 (19.0%)
Marriage Family Therapy	63 (2.1%)	39 (21.8%)
Not Listed	10 (3.3%)	6 (3.4%)
Highest Degree Obtained	n (%)	n (%) (1 missing)
Master’s degree	217 (69.3)	119 (66.5%)
Doctoral degree	86 (27.5)	51 (28.5%)
Other	10 (3.2%)	8 (4.5%)
N (%) who are licensed clinicians	248 (79.0%)	144 (8.4%)
N (%) PCIT Certified Therapist	222 (7.7%)	126 (7.4%)
Years Trained in PCIT M (SD; range)	5.28 (5.00; 0–37)	5.51 (5.48; 0–37)
Years as therapist M (SD; range)	8.86 (7.00; 0–45)	8.96 (7.31; 0–45)
Current Caseload Characteristics
Current PCIT caseload M (SD; range)	5.69 (4.87; 0–30)	6.06 (5.00; 1–25)
Current Total caseload M (SD; range)	2.64 (18.44; 1–100)	19.38 (15.85; 2–100)

PCIT = Parent–Child Interaction Therapy.

In the current study, clinicians (n = 179) who completed the adaptations measure were predominantly female (89.4%) and Non-Hispanic White (74.3%), with 14% self-identifying as Latinx. Regarding racial identity, the majority of clinicians self-identified as White (83.8%). Most clinicians had a master’s degree (66.5%), were licensed (80.4%), and were PCIT-certified (70.4%). They had an average of 8.96 (SD = 7.31) years of experience as clinicians. They reported seeing an average of 6 PCIT clients (SD = 5.00), and 53% saw primarily clients with Medicaid or state insurance. Seventy-nine percent of clinicians reported that they retained at least 50% of their PCIT caseload, with 29.4% reporting that they retained their entire caseload in their transition to iPCIT from in-person services. Additionally, 92% of clinicians reported providing other types of treatment via telehealth following social distancing measures - Table 1 showcases more detailed demographic information for the current sample and baseline sample. No significant differences in gender, race, ethnicity, age, or number of total cases were found between clinicians who did and did not participate in the follow-up survey.

Clinicians were asked to complete two open-ended questions regarding modifications to PCIT in response to COVID-19. One hundred and thirty-five participants provided answers regarding general modifications. Clinicians were predominantly female (89.6%) and Non-Hispanic White (77%%), with 14% self-identifying as Latinx. Regarding racial identity, the majority of clinicians self-identified as White (84.7%). Most clinicians had a master’s degree (68.9%), were licensed (80%), and were PCIT-certified (20%). They had an average of 8.76 (SD = 7.07) years of experience as clinicians and reported seeing an average of 6 PCIT clients (SD = 4.88). Qualitative analysis and results regarding PCIT adaptations focused on responses from all 135 clinicians.

Procedure

In the Fall of 2019, 324 clinicians were recruited to complete a study on PCIT implementation on two listservs managed by training and certifying organizations for PCIT. A follow-up survey about experiences with clinical care during COVID-19 was sent to the participants from the previous study who had provided their contact information (N = 309). In the previous study, clinicians provided an e-mail address, which was used to send gift codes as compensation for study participation. E-mails were connected to a participant ID number to track for follow-up, which allowed us to create unique survey links via Qualtrics and connect survey responses from Time 1 and Time 2 in a combined dataset. For the follow-up study, clinicians were asked via e-mail to complete a 15-minute self-report survey regarding their clinical practices and experiences during the COVID-19 pandemic. The e-mail specified that clinicians were eligible to participate regardless of whether they transitioned to telehealth services. This was done to gain a representative sample of clinicians. Screening questions were administered to assess whether clinicians had transitioned to remote PCIT service delivery. Qualtrics display logic tools were used to display appropriate questions based on the participants’ answers. For the current study, only participants who had delivered PCIT during COVID-19 were included in the sample. Participants were e-mailed a $20.00 Amazon gift card as compensation upon survey completion. The current study was determined exempt by the Institutional Review Board at the University of California Santa Barbara. Prior to completing the survey, participants reviewed an Information Sheet detailing the purpose of the study, their rights as participants, and reminded them of the $20 gift card compensation for their participation, which was also mentioned in the recruitment e-mail.

Measures

Therapist and caseload characteristics

Participants completed a modified version of the Therapist Background Questionnaire (Brookman-Frazee et al., 2012) in the original 2019 survey, which gathered demographic information about participants (e.g., age, gender, race, ethnicity, place of work, type of services provided), professional background (e.g., discipline, licensure status, level of education, PCIT certification), and caseload information (e.g., number of clients, racial and ethnic composition of caseload). Therapist demographics and professional background questions were only collected in the original survey to reduce redundancy and the time it took to complete the survey. In the follow-up survey, participants were asked if they had changed workplaces and reported again on their caseload characteristics to see if there were changes in the clients served in PCIT. These data are reported in a previous manuscript (Barnett et al., 2021).

PCIT adaptations (quantitative)

To examine the types of adaptations that clinicians engaged in, the current study used an adapted version of the Adaptations to Evidence-Based Practices Scale (AES; Lau et al., 2017). The AES assesses the extent to which clinicians made specific adaptations to practices and is composed of six items, using a 5-point Likert scale (0 = not at all, 1 = a slight extent, 2 = a moderate extent, 3 = a good extent, 4 = a very great extent). A factor analysis conducted by the developers of the measure revealed two-factor structures: Reducing/Reordering adaptations (3 items) and Augmentation adaptations (3 items). In the current study, the wording in the AES was adjusted to inquire specifically about adaptations to PCIT delivered remotely following the COVID-19 stay-at-home orders (e.g., During the COVID-19 pandemic, to what extent have you done the following?). Sample items included: “I integrate supplemental content or strategies when I deliver PCIT,” “I remove/skip components of PCIT,” and “I adjust the order of PCIT.” The original AES measure showed great reliability (ω = .95–.98), and the adapted version used in the current study showed adequate reliability for both the Augmenting adaptations subscale (ω = .75) and the Reducing/Reordering adaptations subscale (ω = .71). In the current study, McDonald’s omega was calculated using SPSS macro (Hayes & Coutts, 2020).

PCIT adaptations (qualitative)

Participants who provided iPCIT were also asked an open-ended question after they completed the AES to gain additional information about how they adapted and modified treatment for telehealth delivery. Clinicians were asked the following open-response question: “Based on your responses above, please describe how and why you made these modifications to PCIT.”

Data analytic plan

Mixed-methods design

This study used a QUAN + QUAL approach with simultaneous data collection and equal weighting of quantitative and qualitative data (Palinkas, Horwitz, et al., 2011). Quantitative data were gathered to garner a breadth of understanding about the extent to which clinicians made two types of adaptations, Augmenting adaptations and Reducing/Reordering adaptations (Lau et al., 2017), while transitioning to telehealth. Qualitative open-ended question responses were collected to expand upon and further elucidate quantitative findings. Specifically, clinician responses could provide details on the content of adaptations and reasons for adapting iPCIT.

Quantitative data analysis

To investigate changes in clinician-reported adaptations to PCIT from the summer of 2019 to telehealth delivery of PCIT in the summer of 2022 in response to COVID-19, paired sample t-tests were conducted from Time 1 (i.e., Summer 2019) to Time 2 (i.e., Summer 2020) for Augmenting and Reducing/Reordering Adaptations. All quantitative analyses were conducted with SPSS v28 software.

Qualitative data analysis

Of the 183 participants who transitioned to providing iPCIT, 135 clinicians reported on general modifications made to iPCIT. In line with recommendations for qualitative analyses in implementation research, a deductive approach to analyses was conducted, in which an existing implementation framework informed codes and their development (National Institute of Health, 2019; Palinkas, Aarons, et al., 2011; Palinkas, Horwitz, et al., 2011). Implementation research frequently is guided by existing frameworks to inform qualitative analysis with constructs from the framework informing codes, as opposed to more inductive approaches like grounded analysis (National Institute of Health, 2019). Consistent with the earlier study on community clinician adaptations to PCIT (Luis Sanchez et al., 2022) and the quantitative measure used, Lau et al.’s (2017) Augmenting and Reducing/Reordering adaptations classification informed the process codes related to the content and nature of adaptations, including tailoring, adding elements, and removing components of treatment. Codes related to the rationale behind adapting and modifying were developed using Wiltsey Stirman et al. (2019) FRAME.

After a codebook was developed, a group including one graduate coder and two undergraduate coders (1^st, 3^rd, and 4^th authors) engaged in an iterative approach to coding. Responses were downloaded into an Excel file, which allowed for code organization and tabulation. Following the recommendations by Syed and Nelson (2015), coders first read all responses and met to discuss and add emergent codes based on frequently referenced topics (i.e., technology, telehealth-specific challenges). With the updated codes, coders individually coded the first 25% of responses to determine coding reliability, which were found to be substantial (κ = .75, and κ = .70), between the advanced coder (1^st author) and each one of the other two coders. With good coding reliability established, coders individually coded the remaining responses and met at intervals to review and resolve coding discrepancies and consulted with the entire research team as needed to finalize the coding process. Consistent with thematic analysis in implementation research (National Institute of Health, 2019), themes were identified through analyzing the co-occurrence of content and reason codes; for example, adding time to session to address concerns about financial stress. Themes were finalized through consultation and collaboration among all research team members. The research team included a doctoral level clinical psychologist/faculty member, four doctoral students, and two undergraduate students. The faculty member is a certified trainer in PCIT. The graduate students involved in coding, overseeing data analysis, and/or writing this manuscript have all worked in community mental health, and provided PCIT services; the lead of this study identifies as a Mexican, English-Spanish bilingual Clinical Psychology doctoral candidate.

Integration of quantitative and qualitative findings

Following recommendations for integrating quantitative and qualitative methods in implementation research (Palinkas, Aarons, et al., 2011), the current study sought to triangulate quantitative and qualitative data to determine if both sets of findings yielded similar responses regarding two types of clinician-reported adaptations, Augmenting and Reducing/Reordering adaptations. Additionally, qualitative data expanded on quantitative data by providing rich examples of the types of adaptations made and the reasons for these adaptations (Palinkas, Aarons, et al., 2011).

Results

Quantitative results

Clinician adaptations to PCIT

Separate paired sample t-tests were conducted to determine any changes in Augmenting and Reducing/Reordering adaptations on the AES from Time 1 to Time 2. Quantitative results indicated no change in rates of adaptation as measured on the AES, with both Augmenting (t (179) = −0.09, p = .926) and Reducing/Reordering (t (179) = −0.77, p = .442) adaptations remaining consistent from Time 1 to Time 2. Descriptives showed that as in Time 1, clinicians reported more Augmenting (modifying presentation, integrating supplemental content, lengthening/extending) adaptations (M = 1.34, SD = 0.83) than Reducing/Reordering (shortening/condensing, removing/skipping, adjusting order) adaptations (M = 0. 33, SD = 0.57). On the 0 to 4 scale, on average, clinicians reported engaging in Reducing/Reordering adaptations almost not at all and reported engaging in Augmenting adaptations to a moderate extent at Time 2. Table 2 provides detailed descriptive information regarding the AES items and augmenting and reducing subscales.

Table 2. Paired sample t-Test results on AES augmenting and reducing/reordering adaptation scores with item descriptive characteristics for both time points.

				Time 1			Time 2
Items	(N = 179)	t	p	M	SD	Range	M	SD	Range
Augmenting		−.09	.93	1.33	0.81	0–4	1.34	0.83	0–4
I modify how I present or discuss components of PCIT.				1.30	0.97		2.34	0.94
I integrate supplemental content or strategies when I deliver PCIT.				1.45	1.06		2.40	0.98
I lengthen/extend the pacing of PCIT.				1.22	1.01		2.28	1.13
Reducing/Reordering		−.77	.44	0.30	0.37	0–2	0.33	0.57	0–4
I shorten/condense the pacing of PCIT.				0.63	0.82		1.47	0.92
I remove/skip components of PCIT.				0.19	0.46		1.34	0.61
I adjust the order of PCIT.				0.12	0.41		1.20	0.56

PCIT = Parent–Child Interaction Therapy. Adaptations to Evidence-Based Practices Scale (AES; Lau et al., 2017). The AES assesses the degree to which therapists made specific adaptations to practices and is composed of six items, using a 5-point Likert scale (0 = not at all, 1 = a slight extent, 2 = a moderate extent, 3 = a great extent, 4 = a very great extent). Mean scores for the Augmenting and Reducing/Reordering adaptation subscales range from 0–4.

Qualitative results

Qualitative themes expanded upon and in some ways contradicted the quantitative results, with clinicians describing how COVID-19 and the subsequent stay-at-home orders impacted their delivery of PCIT. Though clinicians reported similar quantitative levels of adaptations before and after the transition to telehealth, their qualitative responses described adaptations aimed to address emergent life events and adjust to the transition to telehealth. Clinician responses referred to several kinds of adaptations falling within the two broader categories (as noted in Table 2), Augmenting and Reducing/Reordering adaptations. However, the findings presented here highlighted the most frequently mentioned adaptations. Table 3 shows the frequency in which each adaptation was coded in the clinicians’ responses and example quotes were provided.

Table 3. Definition, frequencies, and example quotes of therapist-reported content and reasons for adapting.

Codes	Definitions	Frequency	Example Quotes
Adaptations
Lengthening/ extending	Extending the pacing/timing of the intervention; for example, repeating sessions, purposeful addition of specific PCIT sessions.	52	“The pacing is lengthened because of technology issues. I feel more often that I have to make sure the parent can hear me OK, and also spend extra time explaining how everything will need to be set up for each portion of the session. We sometimes schedule a separate time for parents to complete the ECBI by phone.” “[I] Offer more check-in time for parent to report on the week, share challenges, process COVID anxieties and community unrest regarding George Floyd murder and protests.”
Integrating/ adding	Adding another treatment into PCIT, integrating other techniques into a general PCIT approach (including evidence-based adaptations such as BDI, emotion coaching, etc.), adding elements to PCIT.	18	“I’ve had to do way more parent support because they are in constant crisis. I’ve also added the cooperation chart for many kids before starting PDI or used the older child protocol to ease them into consequences. This is both because families are really stressed out and my inability to be physically present for PDI 1.” “Parents expressing difficulties with consistently implementing PCIT strategies due to being home 24/7 and therefore needing supplemental strategies as well.”
Tailoring/tweaking/refining	A change to intervention that leaves all of the major intervention principles and techniques intact (e.g., modifying language, creating somewhat different versions of handouts or homework assignments, cultural adaptations).	32	“I have added providing families with videos of CDI skills since I cannot model them with the child as well as before.”
Changes in packaging or materials	Alterations in the materials used: for example, changing the toys used.	7	“During PDI teach, instead of having the caregivers do the role play with a coach, we recorded a 5-min Mr. Bear going to time-out video and showed caregivers this video. Also, in PDI coach 1, instead of a live demonstration of Mr. Bear going to time out, we played the same video to the child.”
Substituting	Substituting one element of PCIT for another.	4	“With other clients who have a more complex case, I have made some adaptations (separate sessions for each parent, using a token economy instead of time out, etc.)”
Spreading	Breaking up session content over multiple sessions.	4	“Teaches [teaching sessions] take longer on telehealth, had to spread across two sessions.”
Loosening structure	Departing from the intervention (“drift”) followed by a return to protocol within the encounter; drift from protocol without returning; adding extra sessions (e.g., check-in with parents only, collateral sessions) that replace or delay PCIT.	14	“We were within 3 sessions of completing PCIT and the caregiver wants to wait until we are back in person to complete PCIT. We continue to review the basic ideas and skills of PCIT, and I continue to assign special playtime as the homework but have mostly transitioned to individual therapy with my client.” “Since the pandemic started, PCIT seems to be taking a bit longer, so I have delayed starting PDI, and introduced the cooperation chart.”
Removing/ skipping	Removing or skipping elements of the intervention.	10	“Due to inability to meet in person public behavior session was skipped but discussed fully with the family to complete on their own.” “We also skipped some sessions due to protests and caregiver not getting home in time for session.”
Shortening/condensing	Reducing the pacing/timing of the intervention; for example, limiting the number of sessions, coaching for less time.	6	“Sessions tend to be a bit shorter since parents are at home and are balancing caring for other children.”
Reordering	Changing the order of intervention modules or segments.	3	“Because I experience more need for limit-setting for in-home telehealth PCIT, I have provided some minor PDI content midway through CDI. Usually, it involves me prompting parents to give a clear direct command when the parent is ineffectively handling a behavior issue.”
Reasons
Crisis or emergent circumstances	Clinician references to specific events and circumstances that impact clients’ engagement in treatment.	16	“I am providing more handouts/information/psychoeducation on emerging issues (pandemic, social unrest) that may be contributing to problematic behaviors so as to increase parent awareness and understanding of child’s presenting issues.” “Probably the most significant tailoring is that due to COVID stressors, parents often need more than a 1–3-minute check in. Thus, sometimes check ins last longer and then I simply extend the session to get in at least 20 minutes of coaching with the parent.”
Challenges related to telehealth	Included therapist references to technology such as the internet, earpieces, Zoom access, laptops, phones, and toys.	45	“There is just more talk of logistics surrounding doing sessions within the home. PCIT has taken a little longer in the telehealth format (i.e., in the transition from CDI to PDI).” “One of the biggest modifications has been around PDI. I have not felt comfortable coaching a parent through a time out in their homes for the very first time because the home environment is so much less controlled than the clinic, the children run off screen sometimes during playtime, and I cannot intervene in person, if necessary, by just walking into the room the way I can in the clinic. I need more practice with this.”

This is an abbreviated codebook from an original developed for a previous study on clinician adaptations to PCIT (Marked for Review). The codebooks were informed by the FRAME Coding Manual (Wiltsey Stirman et al., 2019).

Augmenting adaptations and reasons for adapting

Clinicians primarily reported engaging in augmenting adaptations including lengthening/extending, integrating/adding, tailoring, and changes in packaging/materials. Examples and reasons for each type of adaptation are described below.

Lengthening/extending adaptations

Participants frequently reported lengthening/extending sessions to address challenges with technology and conducting treatment in the home environment by lengthening portions of treatment. For instance, a clinician shared “Some components of PCIT seem to take longer for parents to integrate. I think part of this is managing the challenges associated with technology and overcoming the barriers that exist in a less controlled environment.” They also reported that sessions took longer to provide the Eyberg Child Behavior Inventory (ECBI), a weekly progress measure of the child’s disruptive behaviors. One clinician stated, “We sometimes schedule a separate time for parents to complete the ECBI by phone,” and another explained that the “length of PCIT sometimes is increased due to the extra time giving the ECBI verbally.”

While most clinician reported adaptations were driven by the transfer to a telehealth format, clinicians also reported engaging in extending and lengthening adaptions to address additional stressors arising during the pandemic; one clinician shared, “I have added extra time for parenting support and problem-solving during COVID and conducting sessions via telehealth because parents seem to need extra support and help with parenting issues.” Another clinician noted, “I have slowed down the pacing of coaching sessions to give time to address the additional stressors that the pandemic has caused to families, and then to try and give PCIT strategies to address these stressors.” This other clinician reported extending the length of sessions due to “families needing extra time to discuss stressors, needing extra focus on feelings and coping.”

Integrating/adding adaptations

Clinicians also reported integrating or adding various elements to treatment to support families with stressors brought up by emergent life events associated with the pandemic. As one clinician shared, “Different interventions have been added to treatment to assist clients in self-regulating work on the fear and the changes caused by the pandemic. Also, to educate families and help with processing emotions that have been triggered by financial challenges.”

Tailoring/tweaking/refining adaptations

Clinicians described several ways in which they tailored their delivery of PCIT to adjust to the challenges of telehealth. For example, one clinician noted, “The families I work with generally don’t use a headset. Coaching is generally done through their device speaker.” Other clinicians spoke on how they leveraged the use of technology to deliver PCIT components as one clinician state, “Role plays during teach sessions were slightly changed because they had to use their own toys and role play thru video.”

Changes in packaging/materials

In response to the loss of some traditional elements of clinic-based treatment, many clinicians reported changes in packaging/materials. One such change was in the delivery of progress monitoring, namely the ECBI. For instance, one clinician discussed providing the ECBI over the phone and through an online platform. Despite the widely reported challenges associated with the rapid shift to telehealth, some participants saw advantages to the remote format, and leveraged technological resources to continue serving clients. One clinician recalled that “During PDI teach, instead of having the caregivers do the role play with a coach, we recorded a 5-min Mr. Bear going to time-out video and showed caregivers this video. Also, in PDI coach 1, instead of a live demonstration of Mr. Bear going to time out, we played the same video to the child.” Other clinicians described using features such as screen-sharing to describe therapeutic techniques. For example, one clinician created Google Slides presentations in response to the difficulty of remote PDI sessions. Others utilized similar strategies to address concerns outside of treatment, especially sociopolitical stressors.

Reducing/reordering adaptations and reasons for adapting

Clinicians reported engaging in Reducing/Reordering adaptations much less frequently than Augmenting adaptations. Clinician-reported Reducing/Reordering Adaptations included loosening the structure and removing/skipping. Examples and reasons for each type of adaptation are described below.

Loosening the structure

Clinicians reported delaying the start of the second phase of treatment, PDI, which reflected the challenges of transitioning from CDI to PDI within the home environment, due largely to technological issues (i.e., lack of a bug-in-the-ear) and concerns about practicing the time-out procedure in the less controlled environment of the home. In response, many clinicians reported intentional delays to the PDI phase of treatment to solidify clients’ understanding of CDI, which was accomplished through additional sessions and supplemental content. One clinician stated, “At the start of the pandemic, our agency held off on starting PDI,but have since started implementing it as per usual with clients who meet CDI mastery.” Other clinicians also reported on the impact that additional stressors faced by families had on delays to the second phase of treatment. For instance, one clinician shared, ““I’ve had to do way more parent support because they are in constant crisis. I’ve also added the cooperation chart for many kids before starting PDI or used the older child protocol to ease them into consequences. This is both because families are really stressed out and my inability to be physically present for PDI 1.” Another clinician reported, “extended CDI [is needed] to accommodate for heightened stressors and lack of family readiness for PDI.”

Removing/skipping

Occasionally, clinicians reported they stopped delivering progress measures (i.e., the ECBI) due to challenges with technology and the time it took to administer the measure orally. One clinician stated, “Sometimes parents don’t remember to fill out the ECBI before the session with Telehealth and I don’t want to take the time during the session. It was easier to remind them when it was an in-person copy for them to fill out.”

Some clinicians also spoke about having to skip certain practices and sessions due to the increased challenges and risks of meeting families via telehealth. For example, one clinician shared, “I have found that I skip some of the role play that I did in office because doing it via telehealth is more difficult.” Another clinician stated, “I shortened the program slightly by skipping the ‘throughout the day’ PDI homework (they were doing this along with cleanup in that assignment), and skipped the actual public outing and sibling sessions but did talk about these and still have the public outings handout.”

Discussion

The importance of maintaining the core features of PCIT has been pointed to as critical to maintain its effectiveness (Ward et al., 2016). While tailoring to the individual case is considered to be part of PCIT delivery, adaptations to the structure or content are ideally tested in efficacy trials (Eyberg, 2005). The field of implementation science has recognized even though treatment developers may caution against adaptations, these often occur in response to unique community contexts (Allen et al., 2012; Park et al., 2018; Wiltsey Stirman et al., 2019). It is unclear how the balance between fidelity and adaptations plays out in the context of sudden global emerging life events. The social upheaval in wake of the onset of COVID-19 illuminated the need for EBPs to be flexible and responsive to stressors faced at an individual and societal level. The current study highlights how clinicians were able to deliver PCIT during these tumultuous times, including adopting a previous adaptation to the model – iPCIT. By having a previously adapted model for telehealth, clinicians described making minimal adaptations, with the majority related to adjusting to telehealth delivery. Mixed methods research prior to the COVID-19 pandemic found that clinician adaptations to telehealth delivery of EBPs were frequently driven by technological barriers, suggesting that these adaptations would likely occur with any transition to telehealth (Parisi et al., 2021). Additionally, studies during COVID-19 stay-at-home orders found that clinicians predominately adapted other EBPs, including Dialectical Behavior Therapy and Trauma Focused Cognitive Behavioral Therapy, due to challenges with technology and the home environment (Landes et al., 2022; Schriger et al., 2022). Therefore, the current study emphasizes the importance of recognizing adaptations that can occur as EBPS are transitioned to telehealth, while also highlighting how to remain flexible to stressors that arise during times of crises. These findings could help EBP developers and trainers plan for ways to help support high fidelity treatment delivery that can be responsive to future crises that emerge and continue to expand telehealth options to promote access in care.

Contrary to our hypotheses, there were no quantitative differences in the extent of adaptations participants reported following the onset of COVID-19 in the current study. Similarly to the first study on community clinician adaptations to PCIT, the majority of adaptations intended to augment the treatment for clients and few adaptations reduced or removed components (Luis Sanchez et al., 2022). However, clinicians’ qualitative responses to open-ended questions expanded upon this finding, highlighting that though the adaptations reported were minimal, they had different reasons for occurring. Adaptations were primarily in response to the clinicians and families adjusting to telehealth, along with addressing an increased number of stressors families experienced during the first months of the COVID-19 pandemic. Many of the modifications described resulted in prolonged treatment, including delaying teaching the second phase of treatment due to challenges controlling the home environment and concerns about causing additional stress for families, were consistent with recommendations disseminated to PCIT clinicians through articles and guidelines made available over PCIT listservs (Gurwitch et al., 2020). However, this study did not specifically capture whether participants made modifications based on the recommendations in these resources or whether such recommendations reached all participants.

Given the types of modifications that clinicians reported qualitatively, including adding more information or strategies, extending the delivery of treatment, and reducing the use of parent-report measures designed to monitor progress in treatment, it was surprising that clinician quantitative reports of adaptations did not show significant differences across time points. On the one hand, it is promising that clinicians reported that they were delivering PCIT with minimal adaptations, suggesting high fidelity implementation even during COVID-19 stay-at-home orders. The minimal adaptations reported by clinicians were not surprising given that iPCIT was shown to be effective in reducing child difficult-to-manage behaviors during pre-pandemic (Comer et al., 2017). It is also possible that self-report quantitative measures of adaptations do not fully capture adaptations and modifications that clinicians make. For example, clinicians may hesitate to report on or discuss adaptations to EBPs given the strong emphasis on treatment fidelity frequently discussed during training and supervision (Lengnick-Hall et al., 2019). Alternatively, it may be that the adaptations captured with existing self-report measures (developed outside of a global health crisis) fall short in their ability to capture modifications required given the magnitude and severity of mental health needs occurring during COVID-19. As such, further research that validates the types of adaptations clinicians make with behavioral observation is needed.

Together, the current study results emphasize the importance of supporting clinicians in how to address individual and societal emergent crises when delivering EBPs to be responsive to family stressors while maintaining fidelity to interventions. Implementation supports for iPCIT, including online videos and an implementation guide on how to address common technology challenges, were widely disseminated following stay-at-home orders via PCIT listservs (Green Rosas et al., 2022). It is possible that these supports guided the types and extent of modifications that clinicians made, promoting high fidelity delivery. This would be consistent with findings from Green Rosas et al. (2022), which found that clinicians who received iPCIT training were more likely to transition to the model following stay-at-home orders; whereas those who were unable to transition to iPCIT described challenges regarding lack of training and organizational delays. Additionally, iPCIT-specific trainings were associated with enhanced parent and child outcomes for families that moved from in-person to telehealth PCIT at the beginning of the pandemic suggesting that these trainings helped to maintain the effectiveness of treatment delivery (Garcia et al., 2021). More research is needed on the types of training and adaptations that enhance clinical outcomes.

Strengths and limitations

This study had several strengths regarding increasing our understanding of how clinicians modified one EBP in response to COVID-19. The current study was uniquely positioned to examine differences in adaptations to PCIT in community settings after the universal transition to iPCIT compared to baseline. Additionally, the current study had a high response rate of 72%, which is greater than the average response rate of 44–50% response rates in online surveys (Van Horn et al., 2009; Wu et al., 2022). Lastly, the breadth of clinician responses to how they had to modify iPCIT could help shed light on how to support appropriate adaptations to EBPs for children and families during times of crises.

Results should be interpreted within the context of certain limitations. First, it is important to recognize the limitations of self-reports of adaptations to EBPs, especially when maintaining fidelity can be seen as socially desirable (Lengnick-Hall et al., 2019). Additionally, while the Augmenting and Reducing/Reordering subscales in the current study had acceptable internal consistency (ω = .71–.75), this was lower than the original AES measure (ω = .95–.98). It is possible that the adapted AES in the present study did not fully capture therapist adaptations to iPCIT in the complex environment of the COVID-19 pandemic. Further, open-ended responses made it challenging to understand if these adaptations were in response to general barriers related to telehealth, as has been found in research prior to the COVID-19 pandemic (Parisi et al., 2021), or the stressors families faced during this time (e.g., job loss, illness, school closures). Additional research with behavior observations of sessions or in-depth qualitative interviews may better illuminate the types of adaptations made and reasons for them. Importantly, the findings described here do not reflect clinical outcomes. Therefore, we were unable to make conclusions about the impact of adaptations to iPCIT during the rapid transition to telehealth recognizing that this all occurred in a unique context. Findings that linked specific iPCIT implementation supports with improved clinical outcomes could indicate the benefits of recommending certain types of modifications (e.g., having extra sessions to address technology concerns) while maintaining fidelity (Garcia et al., 2021). Furthermore, while iPCIT-specific supports, including training and mentoring (e.g., opportunities to provide co-therapy with seasoned iPCIT providers), predicted a successful transition to iPCIT following stay-at-home orders (Green Rosas et al., 2022), the current study did not assess for clinicians length of iPCIT training if any; thus, we were unable to determine the impact that existing iPCIT experience may have had on adaptations made during COVID-19. Lastly, the current study did not evaluate the impact that regional differences in access to mental health services and social distancing measures may have had on therapist adaptations to PCIT during the pandemic.

Implications and future directions

The COVID-19 pandemic and associated stay-at-home orders had both collective and individual impacts on mental health service delivery. A major collective impact was the universal forced transition to telehealth services. Nevertheless, the degree to which individual families were impacted by stay-at-home orders was complex and likely a mix of several factors (e.g., socioeconomic status, race and ethnicity). The rapid shift to telehealth exposed many clinicians to the potential benefits of iPCIT, including opportunities to support families in their home environments, which allowed for greater generalization of skills, and addressing individual family needs. In fact, 83% of PCIT clinicians endorsed that they would like to continue delivering iPCIT after the COVID-19 pandemic (Barnett et al., 2021). Further research is needed to see how the delivery of iPCIT was sustained by these clinicians after the lift of stay-at-home orders.

The findings presented in the current study regarding the challenges of implementing an internet-delivered intervention suggest the need for community agencies and EBP developers to establish adequate training and support (e.g., technology) for clinicians especially given clinicians' interest in supporting families in the context of heightened stress. Subsequently, a shift to greater availability of telehealth could allow for greater access to EBPs, including PCIT, in regions with limited access to these treatments. However, caution needs to be taken when balancing the evident benefits that iPCIT can offer and the risk of reinforcing disparity gaps primarily impacting marginalized communities. Though changes in the racial and ethnic demographics of clients served by iPCIT were not found in previous studies (Garcia et al., 2021; Green Rosas et al., 2022), qualitative results regarding challenges with internet speed and adequate space for treatment indicate that telehealth may be more challenging for rural and low-income families. For example, qualitative answers pointed to the need for highspeed internet and adequate space to be able to deliver iPCIT with fidelity-consistent adaptations. In sum, this study pointed to ways that clinicians may make adaptations that augment treatment to support client needs during times of stress, which have the potential to enhance equity and access to care. Moving forward, it is important to continue to understand how to support clinicians in making effective adaptations to support families during high stress times, especially for families most impacted by crises.

Disclosure statement

The study protocol received an exemption from the Institutional Review Board at the University of California Santa Barbara. Participants received and reviewed an Information Sheet before partaking in the study.

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

Funding was provided by grants awarded to MLB from the UCSB Academic Senate. The time and effort to prepare this manuscript were supported by the UCSB Multidisciplinary Research on the Coronavirus and its Impacts (MRCI) Collaborative grant awarded to BELS and [K01MH110608] awarded to MLB.