Comparative Effectiveness of Two Province-Wide Intervention Models for Preschoolers with Autism Spectrum Disorder

Background:  Many jurisdictions offer comprehensive early intensive behavioural intervention (EIBI) programs to families of preschoolers with ASD, with that diagnosis as the only eligibility criterion. Children in these programs receive many hours of individual intervention per week for extended periods. Many make significant gains, although there is ample evidence of variable outcomes, and high costs are associated with sustaining these intensive services in public systems. Little comparative information exists from either research or community contexts regarding alternative models of EIBI.

Objectives:  To compare pre-intervention characteristics and post-intervention outcomes for preschoolers with ASD in two demographically similar provinces with ASD service models differing in intensity, duration, and treatment methods. Distinct from the comprehensive EIBI model followed in New Brunswick (NB), the Nova Scotia (NS) program relies on Pivotal Response Treatment as its foundation, is less intensive, and of shorter duration.

Methods:  Participants were families of 311 children with ASD recruited from publicly funded EIBI in the Canadian provinces of NB (n = 134) and NS (n = 177). Parents rated their children’s adaptive behaviour (Vineland Adaptive Behavior Scales-2), ASD symptoms (Social Responsiveness Scale-2), and maladaptive behaviour (Scales of Independent Behavior-Revised) at Time 1 (T1; pre-intervention) and / or Time 2 (T2; after a year of service); 42% contributed data at both times. Intervention providers documented services delivered to children.

Results:  Compared to the NB group, NS children were older (55 vs. 48 mos) at T1, and had significantly lower levels of adaptive functioning (mean VABS-2 score of 74.2 vs. 80.6), higher levels of ASD symptoms (mean SRS Total of 72.0 vs. 66.3) and more frequent / severe maladaptive behavior (SIB-R Asocial Index of -7.79 vs. -3.58). Differences between VABS-2 scores at T1 and T2 for both provinces were examined in a multilinear mixed analysis (MLA). Independent variables were time, province, and province X time. Test age was covaried and was not significant, F (1, 469.63) = .160, p = .689. As expected, VABS-2 scores were higher at T2 than at T1, F (1, 451.375) = 5.143, p = .024. The interaction of Time X Province was not significant, F (1, 238.488) = 1.809, p= .180, suggesting no difference between provinces in mean degree of improvement from T1 to T2 (both gain approximately 5 standard score points, i.e., ½ SD). Both groups also showed improvements on some ASD symptoms and problem behaviour symptom measures (SRS-2 and SIB-R, respectively), with no significant differences in improvements between provinces.

Conclusions:  Developmental and behavioural changes over one year of preschool ASD intervention services were similar in two adjacent Canadian provinces, with mean gains of about .5 SD in adaptive behaviour in both groups. These results were observed despite different treatment and service delivery models, as well as significantly different pre-intervention characteristics of the children served. These findings contribute to literature suggesting that the common elements of ASD interventions have important effects that may outweigh differences between models. Implications for policy, including cost-effectiveness, will be discussed.