How To Evaluate Intervention Programs

Evaluating Claims About Treatments for Autism

Testimonials alone are simply too ambiguous to be the basis for making critical decisions about which treatment program to choose. Resources and time are too scarce to be wasted on treatments that have not been shown to be effective.

Now, we have a substantial body of controlled quantitative research on programs of treatment for autism. Now there has been ample time to properly investigate currently popular treatments, but most of these programs have not been. Advocates of treatments should be asked to:

1. describe the exact purposes of the treatment – what is it intended to achieve?
2. describe exactly how the treatment is conducted – there should be no mystery or secrecy about the methods and procedures being used.
3. describe how treatment effects were measured – what numerical data were collected and how were they collected?
4. show before and after data collected by independent – unbiased – evaluators;. and
5. show follow up data – do the persons maintain gains? do they continue to improve? do they regress?

Only applied behavior analysis is able to answer those questions convincingly. Gina Green argues the case in the next chapter entitled "Early Behavioral Intervention for Autism: What Does Research Tell Us?" Then, in Chapter 4, Tristram Smith answers the question, “Are Other Treatments Effective?” His conclusions are:

“Nonbehavioral special education classes, individual therapies, and biological interventions (except major tranquilizers) have not been established as effective treatments for children with autism. Some treatments, especially Facilitated Communication and psychoanalysis, are quite harmful and definitely should be avoided. Major tranquillizers offer an alternative to behavioral treatment for managing disruptive behavior, but they can cause major side-effects and therefore are a last resort rather than a first-line intervention. Several other biological treatments (Prozac, Anafranil, naltrexone, and B6 with magnesium) may be effective but require further research.

In short, behavioral treatment has much more scientific support than any other intervention for children with autism. Consequently, if behavioral treatment is available, or if families are in a position to set up their own behavioral treatment program, the best initial course of action may be to concentrate exclusively on carrying out behavioral treatment as well as possible, rather than looking for ways to supplement it with other treatments.” (Maurice, Green & Luce, 1996, Page 56).

In a paper just published, [Smith T. (1999). Outcome of early intervention for children with autism. Clinical Psychology: Research and Practice, 6, 33-49], Tris Smith has carefully studied peer-reviewed outcome investigations of ABA programs, Project TEACCH, and Colorado Health Sciences. He found that the latter two programs have shown little improvements for most of the children, but some subgroups may have benefited.

In contrast, he found convincing evidence that ABA programs increase adaptive behavior and reduce maladaptive behavior. He also noted that these programs may substantially raise IQ and other standardized test scores, while reducing the need for special services. However, he cautioned that the quality of the research on IQ, other test scores, and school placement does not permit firm conclusions; replications of this research are needed.

For up to date evaluations of biological interventions, see www.autism-biomed.org/research.htm.

Return Home