Current status of intensive behavioral interventions for young children with autism and PDD-NOS
Section snippets
History
The use of applied behavior analysis in the treatment of autistic children is more than 45 years old. Initially, research involved establishing environmental conditions under which these children would respond, followed by the treatment of disruptive behavior (Ferster, 1961, Ferster and DeMyer, 1961). Wolf, Risley, and Mees (1963), for example, treated the food stealing and temper tantrums of a 3 1/2 year old boy evincing autism with edible reinforcers, shaping by successive approximations, and
Reviews
Summary papers of ASD are of fairly recent origin. Early papers were largely narrative, covering topics such as what autism is, with brief sections describing treatments (e.g., Waters, 1990). As noted, one of the first reviews that looked extensively at treatment characteristics was by Margolies (1977). However, as the literature expanded, reviews became more expansive. Matson et al. (1996), for example, evaluated 251 treatment studies and looked particularly at target behaviors and specific
Client factors effecting outcome
EIBI is not a monolith. Investigators pursuing this line of research have varied length of interventions, emphasized different behavior analysis procedures, began treatments at different ages, used parent or professional trainers, or both, and provided training in different settings. Proscriptively, there are a number of child characteristics which are likely to determine how the variables noted above are configured to produce the most effective intervention for a given child. In order of
Current status of treatment efficacy
The number of studies using EIBI is very small compared to treatment research in the ASD field as a whole. For example, Matson et al. (1996) report on 251 treatment studies on specific target behaviors and Lord (2000) reports almost 900 papers on communication and autism from 1979 to 1999. She indicates that publication rates remained relatively steady over 20 years, with about 30 papers per year and about half to two-thirds addressing interventions. Starting in 1984 we found an average of
EIBI studies
Different researchers can look at the same data and come to very different conclusions. For example, some conclude that decades of research have produced no method that can produce cure, while others have asserted that the disorder can be remediated at least in some cases (Honda & Shimizu, 2002; Lovaas, 1987). Additionally, there have been a substantial number of concerns about the methodology of most of the EIBI studies. These studies are hard to do and are very labor intensive. Thus, many of
Studies with no control group
Authors have referred to papers they have published where no control group is included as quasi-experimental designs (Stahmer & Ingersoll, 2004). We would argue that the failure to include controls constitutes no experimental design at all. Quasi-experimental refers to participant selection based on matching individuals across control and experimental groups (Campbell & Stanley, 1963; Spector, 1981). Thus, the experiment is close but not a “true” experiment because children are not randomly
Control group studies without random or matched assignment
Salt et al. (2002) have published one of the few EIBI studies with a true control group. Unfortunately, there are problems with their procedures. Most problematic is the fact that 12 children constituted the treatment group but only five children served as controls. Additionally, there were marked differences in IQ across conditions, which we have noted in our review is a major predictor of how responsive children with ASD are to treatment.
Howard, Sparkman, Cohen, Green, and Stanislaw (2005)
Types of control and matched or random assignment
An optional control group would involve no treatment. However, the field has advanced to a point where even the most conservative researchers would conclude that some early interventions do produce some benefit, at least in the short term. Thus, for ethical reasons, future studies are likely to include placebo control groups. These groups may be based on current clinical practice or some research, but these placebo control groups should diverge considerably from the experimental condition.
Multiple baseline
Random or matched assignment is obviously very difficult and therefore rare in the EIBI literature. Therefore, researchers may wish to consider the use of multiple baseline designs where large numbers of participants are not available. This approach could only be used when one intervention is tested, but each participant serves as their own control, thus eliminating the need for random assignment. Smith, Buch, and Gamby (2000) give a nice example of how to apply this methodology, using six
Most frequently used methods in practice
The primary thrust of our review has been on what the research looks like, from a development perspective. However, there is often a disconnect between what researchers say works and what is actually done in practice (Green et al., 2006). Nowhere is this disconnect more evident than in the ASD literature, where clinicians and researchers often attempt to follow evidence-based practice only to be overwhelmed by the latest gimmick therapy. This phenomenon led some researchers to write a book
Conclusions on effectiveness of EIBI
To date, enough behaviorally oriented EIBI studies have been conducted to suggest that not only is the approach effective, but as a congregate group of learning based methods, it stands alone as the only effective treatment(s) for young children with ASD. Furthermore, parents have expressed satisfaction with these methods. Kasari (2002) reported only five studies with control groups, a number that is increasing, but all too slowly. On the whole, research methodology is getting better, but is
Future directions
Much is yet to be learned about the value of EIBI. Is there a period when the child is too young or the intervention is too intense? Client “burn out” may be an issue at least for some children if intervention is overdone early-on. Anecdotally, we have worked with adolescent persons with ASD who have had intensive direct instruction, and who simply refuse to participate in this treatment approach.
Second, we believe the congregate literature presents a compelling agreement for treatment gains
References (96)
- et al.
Effects of superimposition and background fading on the sight-word reading of a boy with autism
Research in Autism Spectrum Disorders
(2007) - et al.
The development of performances in autistic children in an automatically controlled environment
Journal of Chronic Diseases
(1961) - et al.
Parental reports on the use of treatments and therapies for children with autism spectrum disorders
Research in Autism Spectrum Disorders
(2007) - et al.
Internet survey of treatments used by parents of children with autism
Research in Developmental Disabilities
(2006) - et al.
Patterns of emotional and behavioural disturbance associated with autistic traits in young children with severe intellectual disabilities and challenging behaviours
Research in Developmental Disabilities
(2006) - et al.
A comparison of intensive behavior analytic and eclectic treatments for young children with autism
Research in Developmental Disabilities
(2005) - et al.
Neurochemical correlates of autistic disorder: A review of the literature
Research in Developmental Disabilities
(2006) - et al.
Stereotypy in young children with autism and typically developing children
Research in Developmental Disabilities
(2007) Toilet training of an autistic eight-year-old through operant conditioning therapy: A case report
Behaviour Research and Therapy
(1966)Current status of differential diagnosis for children with autism spectrum disorders
Research in Developmental Disabilities
(2007)