Comparative Effectiveness of EIBI and Adaptive ABA for Children With Autism (ISOLDE)

The Cleveland Clinic, May Institute, Nationwide Children's Hospital, Vanderbilt University Medical Center, University of Michigan, University of Virginia

Approximately 15,000 children with autism spectrum disorder (ASD) in military families currently receive applied behavior analysis (ABA) interventions through TRICARE insurance. This includes early intensive behavioral intervention (EIBI), which involves 20 or more hours per week of individualized instruction based on ABA and is often considered the standard of care for toddlers and preschoolers with ASD. More recently, research has found that less intensive, time limited ABA interventions can effectively target specific core and associated features of ASD. With these latest data, the investigators assert that an individualized approach to adapting and combining targeted interventions could be at least as effective as EIBI, yet substantially reduce expenditures of time and resources. The investigators call this approach adaptive, modular ABA (MABA), and propose to compare EIBI as usual, provided for approximately 20 hours per week, and MABA, provided for up to 10 hours per week, in a 24-week RCT of 132 children with ASD, under age 5 years, in military families. The investigators hypothesize that, at the end of intervention, MABA will be no less effective than EIBI as usual, or only slightly so, on the primary outcome measure (a standardized measure of adaptive skills). The primary investigators also hypothesize that, at follow-ups conducted 24 weeks after intervention and 90 weeks and/or when children are 5 years old, MABA will be superior to EIBI on primary and secondary child outcomes (tests of cognitive and language function, parent- and provider-rated ASD symptoms and adaptive skills) and on parent outcomes (parent stress and sense of competence).

Background: Approximately 15,000 children with autism spectrum disorder (ASD) in military families currently receive applied behavior analysis (ABA) interventions through TRICARE insurance. This includes early intensive behavioral intervention (EIBI), which involves 20 or more hours per week of individualized instruction based on ABA and is often considered the standard of care for toddlers and preschoolers with ASD. Research indicates that EIBI accelerates development of cognitive and adaptive skills in many children with ASD. However, the evidence base has important gaps, notably a dearth of randomized controlled trials (RCTs), limited data on whether EIBI reduces ASD symptoms, and few studies on outcomes of EIBI in community settings such as private agencies where most children with ASD receive services. In addition, EIBI is expensive and requires a substantial commitment of time and effort from children and families. More recently, research has found that less intensive, time limited ABA interventions can effectively target specific core and associated features of ASD. With these latest data, the investigators assert that an individualized approach to adapting and combining targeted interventions could be at least as effective as EIBI, yet substantially reduce expenditures of time and resources. The investigators call this approach adaptive, modular ABA (MABA). Objectives/Hypotheses: Investigators propose to compare EIBI as usual, provided for approximately 20 hours per week, and MABA, provided for up to 10 hours per week, in a 24-week RCT of 132 children with ASD, under age 5 years, in military families. They hypothesize that, at the end of intervention, MABA will be no less effective than EIBI as usual, or only slightly so, on the primary outcome measure (a standardized measure of adaptive skills). Investigators also hypothesize that, at follow-ups conducted 24 weeks after intervention and 90 weeks and/or when children are 5 years old, MABA will be superior to EIBI on primary and secondary child outcomes (tests of cognitive and language function, parent- and provider-rated ASD symptoms and adaptive skills) and on parent outcomes (parent stress and sense of competence). Specific Aims: The primary aim is to compare EIBI and MABA on key child and parent outcomes after 24 weeks of intervention, at a 24 week follow-up (week 48), and at age 5 years and/or 90 weeks. The investigators also intend to explore whether child functioning and family environment moderate the effects of intervention, and examine potential facilitators and barriers to future implementation of MABA (e.g., parent and provider buy-in, fidelity of intervention, cost). Design: 132 children with ASD will be randomized to receive 24 weeks of either 1) EIBI as usual for approximately 20 hours per week or 2) MABA in which children start with intervention focused on social communication 5 hours per week for 4 weeks and then, depending on clinical evaluation of their response, either continue in this intervention or receive augmented intervention for 20 weeks, up to 10 hours per week. Both EIBI and MABA use a tiered delivery system in which trained paraprofessionals provide most of the direct intervention under the supervision of licensed or credentialed professionals. Participants will be recruited from 4 sites that have longstanding relationships with nearby military bases: Vanderbilt University Medical Center, Cleveland Clinic, Nationwide Children's Hospital, and May Institute. Experts on military families will advise the study team on implementation of the study. Child outcome measures assess 1) adaptive skills, 2) ASD symptoms, 3) cognition, and 4) language. Parent outcome measures assess caregiver stress and sense of competence. Linear mixed models (LMMs) will be used to contrast the two interventions in terms of change in outcomes on each measure from baseline through the primary endpoint (end of intervention at Week 24) and at follow-up evaluations. To explore moderators, investigators will augment the LMMs with standard moderated (i.e., covariate-by-treatment) regression analyses. To examine implementation, investigators will compare EIBI and MABA on implementation variables such as buy-in, fidelity, and cost.

Participating children will be randomized to EIBI or MABA with equal probability. All randomizations will occur within site and will be further stratified on 1) baseline Vineland Adaptive Behavior Scale-3 (VABS-3) Composite SS (>65 vs ≤65) and 2) age (<3yrs vs. >3yrs). This procedure will ensure that treatment groups are balanced for these two baseline variables, which are known to correlate highly with outcome during- and post-treatment. The project statistician will use blocked randomization with randomly selected blocks of size 4, 6 and 8 to generate each allocation sequence. All block-randomized allocation sequences will be concealed from all other study personnel.

The investigators hypothesize that MABA will be no less effective than EIBI when children are assessed using the (VABS-3). The VABS-3 assesses adaptive skills in three domains: Communication, Daily Living Skills, and Socialization. It was selected as the primary outcome because (1) it measures coping with everyday settings, which can be considered the best indicator of intervention success, (2) it was recently normed with a nationally representative sample of 2560 parents with excellent reliability and validity, (3) it is an outcome measure that is suitable for the developmental level of the participants, (4) it is deemed by the TRICARE Operations Manual to be reliable and valid, (5) investigators have empirically identified the minimal clinically important difference in VABS-3 standard scores for children with ASD. Thus, the VABS-3 is uniquely suited for testing our primary hypothesis that MABA will be no less effective than EIBI.

OACIS is an expansion of the Clinical Global Impressions scale that takes about 20 minutes to complete. Rather than giving one overall rating for severity and one overall rating for improvement (as is done with the CGI), providers rate severity and improvement on a 7-point Likert scale in 10 domains: ASD symptoms, social interaction, aberrant behaviors, repetitive behaviors, verbal communication, nonverbal communication, hyperactivity, anxiety, sensory sensitivity, and restricted, narrow interests. This scale will be used in two different ways: (1) it will be administered by the clinician for use in the tailoring aspect of the MABA intervention arm as described above, (2) it will be administered by treatment blind independent evaluators with this separate administration to be used in the analysis of secondary child outcomes.

Inclusion Criteria: Diagnosis of ASD based on expert evaluation, including both: (1) Confirmation by a research-reliable administration of the Autism Diagnostic Observation Schedule-2 (ADOS-2), OR confirmation by a routine clinical assessment and observation when a research-reliable ADOS-2 is not able to be collected due to COVID-19 restrictions, and (2) A symptom checklist based on the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition In process of being approved or already approved through TRICARE for ABA services Not yet receiving ABA services With no plans to move location for the 6 month intervention period Medications have been stable for the 6 weeks prior to enrollment in the study Family demonstrates proficiency in written and spoken English Exclusion Criteria: Diagnosis of genetic disorders known to be associated with ASD such as Fragile X, Down syndrome, or tuberous sclerosis Severe motor disabilities such as cerebral palsy that prevents walking or any diagnosis which might prevent standard implementation of the intervention Profound uncorrected vision or hearing loss