Improving Classroom Behaviors among Students with Symptoms of Autism Spectrum Disorder or Attention Deficit Hyperactivity Disorder

BACKGROUND: Autism spectrum disorder (ASD) and attention deficit hyperactivity disorder (ADHD) present increased risk of educational and vocational failure, psychiatric comorbidity, and physical illness. Executive functioning (EF) problems (including inflexibility and poor goal setting and planning) are common in children with ASD and ADHD and are related to negative outcomes. Behavioral classroom management (eg, giving rewards and enforcing consequences to improve behavior) and parent training are the current nonmedication standards of care for ADHD and ASD, but evidence for the long-term efficacy of behavior management is mixed. Reviewers have called for the expansion of cognitive behavioral therapy for ADHD and ASD—specifically, treatments that directly target neuropsychological processes such as EF—but there are very few studies of these treatments in school settings. Poverty impairs EF in all children. Evidence-based EF interventions such as Tools of the Mind have been developed for at-risk preschoolers and adopted by school districts with large numbers of students living in poverty, but no equivalent treatments exist for older children who have already developed significant EF-related developmental disorders, such as ADHD and ASD. OBJECTIVES: This study compared the effectiveness of 2 school- and home-based treatments—a cognitive behavioral treatment (Unstuck and On Target [UOT]) and an adaptation of contingency behavioral management (Parents and Teachers Supporting Students [PATSS])—for children with ASD or ADHD who were attending Title I schools that have a student base in which at least 40% come from low-income families that qualify for free or reduced-price school meals. Both interventions targeted executive functioning and behaviors that relate to success at home and at school. METHODS: We used a cluster-randomized controlled design in which 24 schools in low-income communities provided approximately 21 sessions of UOT (n = 11) or PATSS (n = 13) to 148 third- to fifth-grade students who had either ADHD (N = 98) or ASD (N = 50) and teacher-reported EF problems, particularly in flexible thinking, planning, and organization. The children were socioeconomically, ethnically, and racially diverse, and 41% spoke a language other than English at home. School staff delivered the 2 interventions during 1 school year in small group sessions, with lessons reinforced by parents and classroom teachers. Research staff trained parents (4 sessions), school staff interventionists (6 sessions), and classroom teachers (1 session). The primary outcome was student classroom behavior as measured by treatment-masked observers. We collected outcome data at the beginning of the school year, before the group sessions started (pretreatment) and again after they ended, either at the end of the school year or over the summer (posttreatment). We collected follow-up data 1 year later (long-term follow-up). Secondary outcomes included (1) performance on masked administration of neuropsychological and academic measures, (2) teacher's report on a students' functioning in the classroom, and (3) parent reports on their children's functioning at home. RESULTS: School staff administered both interventions with good fidelity. Regarding the primary outcome, more students with ADHD improved from their baseline classroom EF behavior score than got worse following both interventions (UOT p < .0001; PATSS p = .008). For those with ASD, UOT resulted in more improvement than worsening in observer-rated classroom EF behaviors (p < .0001), but PATSS did not (p = .648). A variety of school staff effectively implemented both treatments. Both parents (p < .001) and children (p = .020) rated UOT more positively than they did PATSS. The implementation (eg, fidelity, lessons completed) and family/student demographic factors (eg, income, race/ethnicity) were not significantly related to changes in classroom observation scores, suggesting that in this highly diverse sample the interventions worked for all kinds of children and families. At 1-year follow-up of the secondary outcome measures (task based with the children and parent report) but not the primary outcome (classroom behavior observations), we saw improvements in a variety of behaviors required for success in school and at home. CONCLUSIONS: UOT and PATSS—2 low-cost, high-gain EF treatments—were easy to implement, acceptable to families, effective, and long lasting. STUDY LIMITATIONS: The study had several important limitations. Our sample was underpowered to detect important subgroup-based moderators and interactions, such as gender. Also, we were not able to collect classroom observations at long-term follow-up, so those analyses were limited to secondary outcomes. Finally, because we did not have a control group, some of the detected improvements in the study could have been due to maturation or other factors. However, 3 of the 4 diagnosis-by-treatment groups showed improvement, which suggests that secular trends were not responsible. An additional important question is whether these results are positive because of the power of new, contextually valid EF interventions or merely because children who have neurobehavioral disorders and attend schools in low-income communities have access to so few treatments that they respond positively when their school or family is offered anything. This is an important scientific and ethical question.