Comparative Efficacy of Self-directed & Therapist-assisted Telehealth Parent Training Intervention for Children With ASD

Autism spectrum disorder (ASD) is a chronic and pervasive neurodevelopmental disorder
characterized by deficits in social communication and the presence of restricted and
repetitive behaviors. Individuals with ASD often require intensive and comprehensive
intervention across the life span. There has been a dramatic increase in the number of
individuals with this diagnosis over the last two decades, with current prevalence rates as
high as 1 in 68. However, there has not been corresponding growth in the availability of
evidence-based intervention services, contributing to high levels of unmet service needs for
individuals with ASD and their families. The vast majority of families of children with ASD
report receiving substantially fewer hours of services than recommended by the National
Research Council. Furthermore, levels of unmet service needs are even higher for families
residing in rural and medically underserved communities. Thus, systematic research focused on
developing and improving strategies for dissemination and implementation of evidence-based
ASD services is a high priority, particularly for chronically underserved communities.

Parent training programs are one cost-effective and ecologically valid way to increase access
to evidence-based ASD intervention. Training parents to provide intervention themselves can
increase the number of intervention hours a child receives and improve child outcomes. A
number of studies have demonstrated that parents can learn to use evidence-based intervention
strategies with a high degree of fidelity and their children experience gains in language and
social communication development, decreases in disruptive behavior, and greater
generalization and maintenance of child skill. These benefits are recognized by parents, who
report parent training to be the most effective practice for promoting their child's overall
development. Parent training also improves family quality of life by increasing parent
self-efficacy and reducing parenting stress, which is particularly important given the high
rate of stress and depression found in parents of children with ASD.

Despite these benefits, there continue to be barriers to the dissemination of training to
parents. Formal parent training programs are rare in community-based early intervention
settings for young children with ASD. For example, in a recent North Carolina survey, only 8%
of parents of children with ASD under 4 reported receiving parent training. Barriers to the
provision of parent training include a shortage of trained professionals, lengthy waitlists,
limited financial resources and transportation, lack of child care, geographic isolation, and
time limitations. These barriers may be even more pronounced for families who live in rural
areas, for whom long distances, poor roads, and climatic barriers limit the ability of
parents to receive services. Thus, it is essential to consider the adaptation of
evidence-based parent training programs to non-traditional service delivery models to
increase access to services.

Telehealth, or the provision of health services and information through the Internet and
related technologies, has the potential to replace or, at the very least, augment traditional
service models to increase access to evidence-based services. Users of telehealth programs
are able to interact directly with instructional content through video, animation, and active
learning tasks, as well as with other individuals, including expert clinicians, via email and
teleconferencing systems. The number of individuals with access to internet-based
technologies has grown considerably in recent years (File T, 2013), and it is now estimated
that 75% of U.S. households have a home computer with high-speed internet, and nearly 87% of
adults are able to access the internet from home, work, or elsewhere. Telehealth services are
becoming increasingly common with over 3,000 U.S. sites using distance-based service delivery
models to provide patient care.

Telehealth programs can reduce patient and provider costs and increase provider system
coverage relative to traditional in-person service delivery models. Such programs have been
shown to greatly improve care for patients with chronic diseases, such as diabetes, heart
disease, and asthma, and increase access to evidence-based health promotion (i.e., smoking
cessation, dietary change, physical activity), psychological (i.e., CBT for depression and
anxiety), and parenting interventions. Patients are often very satisfied with the care they
receive through telehealth services, and efficacy studies have found moderate to large
effects of telehealth interventions on participant knowledge and behavior change. Further,
several recent meta-analyses have found that CBT delivered via telehealth is as effective as
traditional therapist-delivered intervention. Taken together, these data suggest that
telehealth may serve as a promising alternative service delivery model to increase the
dissemination of ASD parent training programs.

Telehealth interventions can be either self-directed, in which the participant engages with
the interactive program at his or her own pace, or therapist-assisted, in which the
participant receives additional guidance from a trained professional as he or she completes
the program. Self-directed programs have far greater dissemination potential as they do not
require a trained professional and can typically be administered at a much reduced cost. At
the same time, research on telehealth CBT interventions for mood and anxiety have found that
therapist-assisted programs lead to better client outcomes than self-directed programs. This
finding may be particularly relevant for telehealth parent training interventions, as
research suggests that parent coaching is important for increasing parents' fidelity of
implementation, and coaching is a core component of existing evidence-based parent training
interventions for children with ASD.

There has been growing interest in extending telehealth interventions to parent training for
children with ASD; yet empirical evaluations of such programs are very limited. Several
uncontrolled studies of self-directed telehealth parent training programs have shown that
parents of children with ASD find such programs to be highly acceptable, and program use is
associated with gains in parent knowledge of evidence-based intervention procedures.
Preliminary empirical studies have suggested that parents can learn to implement
evidence-based intervention techniques with fidelity after using a self-directed telehealth
parent training program. A small controlled study found that parents who were given access to
a DVD-based self-directed program made greater gains in their correct use of pivotal response
training (PRT) strategies, provided their children with more language opportunities, and
displayed greater confidence in parent-child interactions, than parents in a no-treatment
control group. Using a single-subject design, demonstrated that parents improved their
correct use of reciprocal imitation training (RIT) techniques after completing a web-based,
self-directed program. However, one third of the participants required additional in-person
coaching in order to meet the pre-determined fidelity of implementation criterion. In both
studies, improvements in child behaviors targeted by the interventions were also observed.
Importantly, parents in both studies indicated that immediate feedback or coaching from an
expert clinician would have been a helpful addition to the program.

Several recent single-subject design studies have examined the efficacy of therapist-assisted
telehealth parent training programs for children with ASD. Researchers examined the effect of
both a DVD-based and web-based program in conjunction with weekly remote coaching sessions
via video-conferencing to teach parents the Early Start Denver Model (ESDM). Results
suggested that parents increased their correct usage of the intervention strategies and
altered their engagement styles to be more attentive and responsive to their children after
the therapist-assisted telehealth program. Furthermore, children in both studies demonstrated
gains in important social communicative behaviors (e.g., language, imitative behaviors). In
an attempt to evaluate the relative contributions of self-directed instruction and remote
coaching on parent learning of RIT, researchers conducted a second study that measured
parents' use of RIT techniques with their child prior to completing a self-directed,
web-based program, after completing the self-directed program but before receiving remote
coaching via video-conferencing, and after receiving remote coaching. Similar to their
previous study, all parents improved their correct use of the intervention techniques in
response to the self-directed program, but roughly a third showed additional benefit from the
provision of remote coaching. In addition, most children demonstrated gains in imitation
skills (the target of the intervention) with the onset of treatment; however, child gains
were most robust during the coaching portion of the program.

These preliminary findings suggest telehealth parent training interventions are highly
acceptable to parents of children with ASD, and can result in improvements in parent
knowledge and implementation of evidence-based intervention techniques and gains in child
social communication skills. To date, no studies have conducted a head-to-head comparison of
self-directed and therapist-assisted telehealth parent training interventions. However, data
thus far indicate that parent coaching may be necessary for some, but not all parents to
successfully implement evidence-based intervention techniques with fidelity. A more nuanced
appreciation of the contributions offered by each component, as well as which parents are
more likely to benefit from each, will make it possible to develop more cost-effective
delivery models in which services can be offered at different levels of intensity, depending
on specific needs of the family (i.e., stepped-care).

ImPACT Online is an interactive, web-based telehealth intervention that teaches parents to
promote their child's social-communication skills during play and daily routines. The
intervention content was adapted from Project ImPACT, a manualized parent training curriculum
that uses a blend of developmental and naturalistic behavioral intervention techniques,
including following the child's lead, imitating the child, using heightened animation, using
simplified language around the child's focus of attention, environmental arrangement
strategies, prompting, and natural reinforcement during child-directed activities. There is
strong empirical support for the intervention techniques for increasing social communication
in children with ASD. In addition, several studies have demonstrated the efficacy of the
original parent training curriculum for increasing parent fidelity and child language skills.

The original curriculum was designed to be implemented over 12 weeks in either an individual
or group format by an in-person parent trainer. The investigators have adapted the program so
that the intervention content is presented in 12 interactive, web-based modules or lessons.
These lessons can be administered as a standalone intervention (self-directed model) or in
combination with remote coaching (therapist-assisted model). Program development was guided
by the technology acceptance model56, media richness theory57,58, and principles of
instructional design59. To ensure usability, The investigators used an iterative development
process with input from focus groups of parents and intervention providers, and interviews
with pilot participants. The investigators have conducted a feasibility trial of the
self-directed and therapist-assisted models with 25 families, the majority of whom (75%)
reside in rural and medically underserved areas. Our data thus far strongly support the
acceptability and feasibility of the web-based modules and the remote coaching protocol, and
suggest that such models can be used to effectively reach individuals residing in underserved
communities. Further, our data lend preliminary support for the positive effect of the
program on parent fidelity of implementation, self-efficacy, and child language. The next
step in evaluating ImPACT Online is to conduct a fully-powered RCT with a control group to
evaluate the comparative efficacy of the self-directed and therapist-assisted telehealth
intervention models on key parent and child behaviors, as well as to examine potential
mediators and moderators of treatment outcomes.

Statement of the Problem Research indicates that early, intensive intervention can lead to
significant improvements for children with ASD. There has been a dramatic increase in the
prevalence of ASD over the past two decades, but not a corresponding growth in availability
of evidence-based intervention services. This has led to high levels of unmet service needs
for individuals with ASD and their families. Parent training is one cost-effective and
ecologically valid approach to increase access to intervention. Additional benefits include
increases in parenting self-efficacy and decreases in parenting stress. Yet, there continue
to be barriers involved with the dissemination of training to parents. Telehealth
interventions have the potential to replace, or at the least augment, traditional service
models to increase access to evidence-based services, particularly in rural and medically
underserved areas. There is increasing evidence to suggest that such programs can reduce
patient and provider costs and increase provider system coverage relative to traditional
in-person service delivery models.

Although there has been growing interest in extending telehealth interventions to parent
training for children with ASD, empirical evaluations of such programs are limited, and
little is known regarding the relative benefits of self-directed and therapist-assisted
telehealth interventions for ASD. Self-directed programs have far greater dissemination
potential as they do not require a trained professional and can typically be administered at
a much reduced cost. At the same time, research on telehealth interventions for other
disorders have found that therapist-assisted programs lead to better client outcomes than
self-directed programs. A better understanding of the relative benefits of these two delivery
formats on parent and child outcomes, as well as the families for whom each format is most
and least effective, is crucial for the development of effective and efficient telehealth
parent training interventions.

The specific objectives of this project are to conduct a randomized control trial to examine
the effect of a novel, telehealth parent training intervention for children with autism
spectrum disorder (ASD), ImPACT Online, on parent and child outcomes. The investigators will
compare the benefits of the self-directed and therapist-assisted delivery formats, and
examine moderators and mediators of treatment outcomes. The investigators anticipate that
both the self-directed and therapist-assisted models of ImPACT Online will be effective
methods for teaching parents to use evidence-based intervention strategies and for increasing
parent self-efficacy compared to a web-based information control group. Parents' use of these
strategies will greatly enhance the amount of evidence-based intervention that their children
receive and improve generalization and maintenance of skills, positively impacting their
children's language development. The investigators anticipate that parents in the
therapist-assisted group will outperform parents in the self-directed group on the primary
parent outcome measures at the group level; however, our moderator analyses may indicate
parents for whom each approach is more and less effective. These data will be key to
informing a stepped-care model in which parents can be offered the intervention at different
levels of intensity (self-directed or therapist-assisted), depending on the specific needs of
the family. Further, if this approach to parent training is successful, it has the potential
to greatly increase access to evidence-based intervention for children with ASD, especially
those living in rural and underserved areas.

Ninety families of young children with ASD will be assessed at pre-treatment (T1), randomized
to the self-directed ImPACT Online, therapist-assisted ImPACT Online, or web-based
information control condition, and will receive treatment for 4 months. Outcome assessments
will occur at post-treatment (T2) and a 3-month follow-up (T3).

Aim 1: To examine the comparative efficacy of self-directed and therapist-assisted ImPACT
Online on parent outcomes at post-treatment and the 3-month follow-up. The primary goal of
ImPACT Online is to teach parents to use specific intervention techniques to promote their
child's social communication development. As parents gain the knowledge and skills to help
their child, their self-efficacy as a parent is expected to increase. Thus, The investigators
will compare the effects of the two telehealth intervention models on parent fidelity and
parent self-efficacy at post-treatment and a 3-month follow-up. It is hypothesized that
families randomized to either of the two telehealth conditions will demonstrate greater gains
in parent fidelity and parent self-efficacy at post-treatment and 3-month follow-up than
parents assigned to the control group. Given research on telehealth interventions for other
disorders suggesting an added benefit of therapist-assistance, it is hypothesized that
families randomized to the therapist-assisted telehealth condition will demonstrate greater
gains in parent outcomes relative to parents in the self-directed telehealth condition at
post and follow-up.

Aim 2: To examine the comparative efficacy of self-directed and therapist-assisted ImPACT
Online on child outcomes at the 3-month follow-up. Parent use of the intervention techniques
over time is expected to lead to improvements in child verbal language. Thus, The
investigators will compare the effects of the two intervention models on child verbal
language at the 3-month follow-up. It is hypothesized that families randomized to either of
the two telehealth conditions will demonstrate greater gains in child verbal language at
follow-up than parents assigned to the control group. Given the stronger effects on parent
outcomes expected for the therapist-assisted intervention, it is hypothesized that families
randomized to the therapist-assisted telehealth condition will demonstrate greater gains in
child verbal language relative to parents in the self-directed telehealth condition at
follow-up.

Aim 3: To examine whether pre-treatment parenting stress moderates the effects of ImPACT
Online on parent fidelity and parent self-efficacy at post-treatment. Our goal is to identify
the families for whom the different telehealth delivery formats are most and least effective.
Prior research has indicated a negative association between pre-treatment parenting stress
levels and treatment outcomes in ASD parent training interventions. Thus, The investigators
will examine pre-treatment parenting stress as a putative moderator of parent treatment
response. It is hypothesized that the effects of ImPACT Online on parent fidelity and parent
self-efficacy will be stronger for parents with lower pre-treatment levels of parenting
stress. It is anticipated that pre-treatment parenting stress levels may also differentially
impact parent outcomes for the two telehealth delivery formats. For example, high levels of
pre-treatment parenting stress may be more likely to negatively impact parent outcomes in the
self-directed group, because parents do not have the additional support of a coach. Thus, it
is also hypothesized that the relationship between parenting stress and parent outcomes will
be stronger in the self-directed group than the therapist-assisted group.

Aim 4: To examine whether gains in parent fidelity and parent self-efficacy at post-treatment
mediate the effects of the ImPACT Online on child verbal language at the 3-month follow-up.
In ASD parent training studies it has been assumed that improvements in child outcomes are a
result of improvements in their parents' ability to implement the treatment. However, this
relationship has rarely been directly tested. While parent fidelity is a likely candidate, it
is also possible that changes in other important parent behaviors, such as parent
self-efficacy, may be important factors in promoting child change. For example, it may be
that parents who feel more self-efficacious in the parenting role may engage with their child
more often, have higher expectations for their child's behavior, or may more effectively
advocate for their child's needs, all of which might result in "non-specific" treatment
effects. Thus, The investigators will examine mediators of child changes to better understand
how our parent training model is affecting child outcomes. Specifically, The investigators
will examine whether improvements in child verbal language as a result of treatment are
mediated by changes in parent fidelity and/or parent self-efficacy. It is hypothesized that
the effects of ImPACT Online on child verbal language at the 3-month follow-up will be
greater for children whose parents demonstrate greater gains in parent fidelity and parent
self-efficacy from pre- to post-treatment.

Aim 5 (exploratory): To identify predictors of adherence to ImPACT Online. In order to fully
capitalize on the benefits of telehealth, it is important to understand factors that
influence parent adherence to ImPACT Online. The investigators have selected one
program-level (therapist assistance) and three individual-level (computer/internet fluency,
parent expectancies, treatment acceptability) factors that have been found to influence
adherence to parent training and/or telehealth-based interventions in previous work. The
investigators will also examine whether any sociodemographic variables predict treatment
adherence. It is hypothesized that therapist assistance will predict adherence, such the
therapist-assisted group will exhibit greater adherence than the self-directed group, and
that computer/internet fluency, parent expectancy, and treatment acceptability will be
positively associated with adherence to ImPACT online. Finally, The investigators hypothesize
that some sociodemographic variable will be found to predict treatment adherence.

Research Design and Methods Participants. 90 children with ASD between the ages of 18 and 72
months will be recruited for this study. This sample size was determined to be adequate for
maintaining adequate power for the proposed statistical analyses, while allowing for some
attrition. To ensure that the sample includes a sufficient number of participants who are
traditionally underserved, The investigators will oversample for participants living in a
federally designated rural and/or medically-underserved area, and those with additional risk
factors associated with decreased access to care, including low parental education,
single-parent household, racial or ethnic minority, and immigrant status. To be eligible,
children must receive a DSM-V-informed clinical diagnosis of Autism Spectrum Disorder based
on the Autism Diagnostic Observation Schedule (ADOS). In light of research suggesting that
diagnostic stability is lower in children under 30 months. Results of diagnostic testing for
the younger children in our sample will be interpreted cautiously and discussed with parents
in a sensitive and transparent manner. Children with a history of significant brain injury,
known neurological or genetic condition, significant sensory or motor impairment, or major
medical problems will be excluded as they might be expected to respond differently than
children with ASD only or require program modifications.

Participants will be recruited through our ongoing contacts with community agencies and
professionals serving young children with ASD in Michigan. The P.I. has extensive experience
recruiting families of young children with ASD for intervention studies. Previous research
efforts, including a pilot RCT of ImPACT Online, strongly suggest that The investigators will
be able to recruit the proposed number of participants. The MSU Autism Lab uses a number of
strategies to ensure retention, including maintaining responsive relationships with
participating families, providing clinical consultation and referral, and maintaining contact
through newsletters and periodic family outings. Retention rates for previous studies,
including the pilot RCT of this intervention, have all been above 90%.

Group assignment. It is possible that group differences in child age and/or developmental
quotient (DQ) at intake could affect the internal validity of the study. Thus, participants
will be stratified by age (18-35 months; 36-72 months) and developmental quotient (DQ<50;
DQ≥50) prior to being randomized. After stratification by age and DQ, participants will be
randomly assigned to the one of the three groups using random permuted blocks of 6. This
approach has been shown to yield groups that are closely balanced with regard to size and
relevant pre-treatment characteristics. All pre-treatment assessments will be collected prior
to randomization to prevent bias during testing based on knowledge of future group placement.

Children in all three groups will continue to receive their existing educational programming.
It is not expected that non-treatment intervention participation will differ systematically
across groups; however, information on the amount and type of outside services and changes to
services will be collected for each child on a monthly basis throughout their participation.
All families who express a need for a computer or high speed internet will be provided with a
loaner laptop and a high speed internet service plan for the duration of their participation.

Training plan and coaching fidelity for Therapist-Assisted Group. Therapists who provide the
remote coaching will be masters' level project staff trained to fidelity in the coaching
procedure by the PI. Training includes a combination of didactic training, observation of
remote coaching sessions, role play, and hands-on coaching during practice with children with
ASD and their parents. All therapists will be required to establish fidelity with non-study
participants prior to beginning the study, as determined by >90% correct implementation on
the Coaching Fidelity Checklist (Ingersoll & Dvortcsak, 2010). This training approach has
been highly successful for ensuring fidelity of therapists. Coaching fidelity data will be
collected for a randomly selected sample of 20% of sessions and analyzed to ensure correct
implementation of the remote coaching procedure. If a therapist falls below 90% correct
implementation, the PI will provide additional training and feedback until the therapist
re-establishes fidelity.

Assessment protocol. Participants in each group will receive assessments at 3 time points:
pre-treatment (prior to randomization) (T1), post-treatment (T2), and 3-month follow-up (T3).
A 3-month follow-up was chosen to allow time for the children to benefit from their parents'
use of the intervention strategies, while ensuring that follow-up assessments for all
children can be completed within the project period. Each assessment period will consist of
three to four hours of testing, involving direct child assessment, home observation of a
parent-child interaction, and parent-report measures of parent and child functioning which
will be completed in the family home.

Examiners who are blind to the study objectives and treatment group assignment will conduct
all assessments. Assessments requiring behavioral coding will be videotaped and scored by
trained observers who are blind to treatment condition. One-third of the assessments will be
coded by a second, independent observer to determine interobserver agreement. Families will
be compensated $25 per assessment period ($75 per family).

Inclusion Criteria:

- a DSM-V-informed clinical diagnosis of Autism Spectrum Disorder based on the Autism
Diagnostic Observation Schedule (ADOS)

Exclusion Criteria:

- history of significant brain injury, known neurological or genetic condition

- significant sensory or motor impairment

- major medical problems