A systematic review of early intensive intervention for autism spectrum disorders

Warren, Z., McPheeters, M. L., Sathe, N., Foss-Feig, J. H., Glasser, A., & Veenstra-VanderWeele, J. (2012). A systematic review of early intensive intervention for autism spectrum disorders. Pediatrics,127(5), 1303-1312. doi: 10.1542/peds.2011-0426

Reviewed by: ToniAnne Giunta, M.A., BCBA, Caldwell College

Why review this topic?

The 1987 study conducted by Lovaas and his UCLA colleagues revealed that early intensive interventions, rooted in the principles of behavior analysis, can lead to improved outcomes in children with autism spectrum disorders (ASD; e.g., decreased symptoms of autism and interfering behaviors, increased cognitive and adaptive functioning, higher likelihood of transitioning to general education settings, increased independence). Since that time, additional research has been conducted in this area and expanded to include other interventions. The present study attempted to systematically review published evidence to examine the effectiveness of behavioral and developmental interventions, and to deter-mine questions for additional research.

What did the researchers do?

Databases were searched (e.g., PSYCHINFO, ER-IC) for studies published between 2000 and 2010. All studies included a minimum of 10 children with ASD, younger than 13 years old, and interventions that targeted multiple areas of functioning. Studies that focused on single target areas (e.g., imitation) were not included. For each study, the following were examined: (a) study design (e.g., comparing groups of children receiving behavioral intervention to children receiving non-behavioral intervention), (b) method of evaluating symptoms of autism, (c) characteristics of children prior to intervention, (d) description of intervention, and (e) characteristics of children after intervention implementation. Based on the analyses of these components, studies were rated based on their strength of evidence (i.e., the degree of confidence that the reported effects of the intervention are unlikely to change if future studies are conducted) as “insufficient,” “low,” “moderate,” or “high.” Thirty-eight studies were identified and classified into one of three categories for further analysis: (1) early intensive behavioral intervention (EIBI; i.e., approaches derived from Lovaas, 30+ hours per week for 1+ years of one-on-one instruction), (2) comprehensive interventions for children younger than two years old, or (3) focused interventions primarily utilizing parent training. Only a sample of studies was summarized for each category (i.e., seven EIBI studies, four studies with children younger than two years old, four parent training studies).

What did the researchers find?

In one study, characterized as Lovaas-based, children in an EIBI group were compared to children in a non-EIBI group. Though outcomes remained in the impaired range, children in the EIBI group gained an average of 15 points in IQ when compared to IQ points in the non-EIBI group. In a similar study that compared language and adaptive functioning areas in addition to IQ, no differences were found between the groups. Two studies compared EIBI intervention to eclectic and general public school interventions. Children in the EIBI groups demonstrated significant improvements, including an average of 41 points increase in IQ. A similar study showed improvements in IQ score, adaptive functioning, and school placement, but failed to show major differences in receptive and expressive language and socialization. Though another study failed to show significant differences in IQ when comparing groups of EIBI and eclectic interventions, other areas showed significant improvements in EIBI groups, including language and communication, reciprocal social interaction, and decreases in symptoms of autism. An EIBI study that compared high-intensity (i.e., 30 hours/week) versus low-intensity (i.e., 13 hours/week) intervention found that children in the former group showed significant improvements in both intellectual and educational functioning.

Additional studies evaluated comprehensive approaches for children with ASD younger than two years old. For example, a study that analyzed the Early Start Denver Model (ESDM, i.e., intervention that combines behavioral and developmental approaches) showed children in the ESDM group displaying more significant gains in IQ and adaptive functioning when compared to groups with less comprehensive intervention. No differences in symptoms of autism, however, were found between the groups, nor have independent replications been conducted thus far.

Several studies also evaluated short-term parent training for promoting children’s social communication. These treatments were focused on particular outcomes rather than comprehensive out-comes. Although these studies showed improvements in behavior observations conducted at the research site and in parent and teacher ratings, gains on standardized tests and long-term improvements were not demonstrated. Because of these limitations, conclusions of their positive effects must be interpreted with caution.

In general, for all three categories, many concerns were raised about the studies’ procedures, including: (a) lacking randomized, controlled designs, (b) variation of participant characteristics within the studies, (c) failing to describe characteristics of the specific individuals in which improvements were found, and (d) variation in how the interventions were implemented. Furthermore, not enough replications or extensions have been conducted for the interventions to be deemed effective. When rating strength of evidence, Lovaas-based EIBI interventions were rated as “low,” while the comprehensive and parent training interventions were rated “insufficient.” None of the included interventions were rated in the “moderate” to “high” range.

What do the results mean?

Of the three categories, the Lovaas-based EIBI studies were most prevalent in the current body of literature of early intensive intervention and also revealed the most positive changes in many areas for children with ASD. Though most of those studies reported more improvements for the children in EIBI groups post-intervention, including increased language and communication, higher reciprocal social interactions, decreased symptoms of autism, and higher likelihood of placement in general education settings, evidence was shown to be strongest for IQ. Even fewer studies analyzed comprehensive interventions for children younger than two years old (e.g., ESDM); the strengths of these interventions also remain unclear. Additionally, the studies of less intensive interventions that primarily utilized parent training were rated as “insufficient” in terms of analyzing outcomes, particularly due to the use of non-objective measures.

Although these reviewed interventions were rated as “low” or “insufficient” in terms of strength of evidence, it should not be concluded that the interventions are ineffective; instead, the results from this review should encourage researchers to conduct controlled replications and extensions of the studies to examine if effectiveness can be established.

What are the strengths and limitations of the study?

The study attempted to systematically review the current evidence of early intensive intervention, and overall showed that behavioral interventions can yield positive changes for children with ASD. Although the EIBI studies were most prevalent in the literature, many additional replications of the original 1987 study are needed, as well as comparisons of other interventions to those rooted in behavior analysis. With several of the studies analyzed having multiple limitations, the review posed several additional suggestions for future comparative analyses, including: (a) taking objective measures of the children’s behaviors pre- and post-intervention as opposed to just therapist and parent reports, (b) ensuring interventions are implemented correctly through additional recordings of therapist behaviors, (c) controlling that the presence of other interventions are not affecting the outcomes of the intervention(s) under study, and (d) examining more closely how interventions affect individual children differently versus summarizing outcomes as groups. Because a wide variety of measures are used to assess outcomes across studies (e.g., IQ score, adaptive assessment scores, parent reports), it is often challenging for reviewers to compare such studies. Finally, because characteristics of ASD vary across individual children, it is difficult to make conclusions that are representative of the entire population of children with ASD. This review summarizes the current literature of early intensive intervention while simultaneously encouraging additional research to be conducted.

Cite This:

Giunta, T. (2014). Research Synopsis: Warren, et al. (2012): A systematic review of early intensive intervention for autism spectrum disorders. Science in Autism Treatment, 11(2), pp. 24-26.