The CARD Position on Biomedical Treatments for Autism, January 19, 2006
A substantial percentage of children with autism currently receive “biomedical” treatments, despite a current lack of evidence to support or refute most of them. In a recent survey of parents of children with autism (Green, et al., 2005), 27% of parents reported that their children with autism currently receive treatment in the form of special diets and 43% reported using vitamin supplements.
CARD’s position on the use of biomedical treatments in clinical practice is centered around three basic points: 1) many parents of children with autism believe that various biomedical treatments have been responsible for substantial improvement in their children, 2) very little research has been conducted on the effects of biomedical treatments for autism, and 3) parents must ultimately make the decision as to which treatments are appropriate for their children, regardless of diagnosis or disorder.
CARD’s position on research on biomedical treatments and on the collaboration of behavior analysts with clinicians and researchers from other disciplines is based on the points elaborated above. At least two factors contribute to what we perceive as a grievous need for sound empirical research on the effects of biomedical treatments for autism. First, as the Green, et al., (2005) study demonstrates, biomedical treatments are being implemented on a widespread basis. This fact alone is more than ample justification for conducting research on their safety and effectiveness. Such widespread use must be tempered with sound research. Second, many parents honestly believe that their children have significantly benefited from biomedical treatments. As clinicians and applied scientists, we have an ethical responsibility to take the concerns and beliefs of our clients seriously. To dismiss the convictions of our clients would be tantamount to disrespect for those who are mostly closely affected by autism. At the same time, ample research has demonstrated that clients can be made to believe that an intervention is or is not effective, regardless of the actual effects produced, and therefore opinion alone (be it that of a client or a scientist) is never to be accepted or rejected outright. Sound empirical treatment research is the only path toward addressing such concerns in a sufficient manner.
Throughout the history of science, particular schools of thought and areas of research have risen and fallen amidst ubiquitous controversy. Ultimately, controversy may have very little effect on which approaches to a problem are borne out. When a useful solution to a problem is discovered, and the results are replicated many times over, little is left to controversy. The use of ABA for children with autism was once highly controversial but the unrelenting work of parents and the repeated and consistent replication of beneficial results in the scientific literature has moved the field of ABA closer to the mainstream. Most biomedical treatments for autism have not yet been subjected to repeated, rigorous outcome research. Thus conclusions regarding their effectiveness (either for or against) cannot be made.
The best safeguard against controversy in the evaluation of scientific issues may be skepticism. Skepticism does not refer to disbelief. It refers to the practice of withholding judgment on a given topic until such time as sufficient evidence warrants judgment one way or the other (Shermer, 2002). In the absence of conclusive evidence, then, one might be advised to be skeptical of the view that an intervention works, as well as to be skeptical of the view that it does not. It is CARD’s position that conclusive evidence for or against the effectiveness of most biomedical treatments for autism does not at this time exist. Hence the urgent need for empirical investigation and the futility of blind acceptance or denial of the validity of the biomedical treatment of autism.
A common feature of any controversy is polarization. A casual review of the major clinical, research, and advocacy factions within the field of autism today reveals that most parties ardently maintain that biomedical treatments are either extremely effective or a total sham. Many who advocate for biomedical treatments appear to believe that all biomedical treatments are equally effective and all are virtual cures for autism. Many who reject the utility of biomedical treatments, on the other hand, do so across the entire range of treatments, without regard to the particular case for or against the many divergent treatments which fall under the biomedical umbrella. Particularly given the still largely unknown etiology of autism, either position appears premature at the current time. Autism is a spectrum disorder which is comprised of millions of individuals who present with widely divergent characteristics. It is not yet even known whether all cases of autism share a common etiology. Given the widely divergent manifestations of the disorder, if any biomedical treatments are proven to be effective in the future, it seems unlikely that any particular one will prove equally effective for all persons with autism. Even less likely is the notion that all biomedical treatments will be equally effective or ineffective, if the large variability within autism is ignored.
Perhaps most disturbing is the notion that it is not possible to reliably evaluate the separate and combined effects of ABA and biomedical treatments for autism. This perspective is based on a fundamental misunderstanding of the nature of experimental scientific investigation. Virtually all disciplines of experimental science agree that experimentation consists of altering one variable at a time and observing the effects that the alteration produces on another variable. Sound experimentation depends on manipulating one variable at a time while simultaneously controlling for the influence of extraneous variables. To the extent that this is done (regardless of the particular scientific field or research topic), inferences can be made about the effects of one variable on another. There is nothing peculiar about autism, ABA, or biomedical treatments which preclude this sort of experimentation. It is common for clinicians (behavioral, medical, or other) to manipulate multiple variables simultaneously in order to bring about optimal treatment outcome and any time this is done it is likely to preclude precise analysis of which variables were responsible for improvement. In order to produce sound research on the separate and combined effects (if any) of each approach, experiments must hold one variable constant while manipulating another. This approach to autism research is largely untouched within the ABA and biomedical communities, but this fact does not preclude such research from being developed, and indeed it is currently under way at several research sites.
Toward this end, CARD is currently collaborating with medical doctors to conduct sound research on biomedical treatments for autism. The focus of this effort is to identify which, if any, biomedical treatments result in which particular improvements for particular individuals with autism, given their unique biomedical and behavioral status. All research conducted is interdisciplinary in nature and all treatment studies evaluate multiple behavioral and biomedical measures. The goal is to establish a model for interdisciplinary collaboration between behavior analysts and medical doctors in researching treatment for individuals with autism. It is our hope that the research produced will forge a path toward addressing the debate regarding biomedical treatments for autism with sound scientific data, thereby displacing the current culture of hearsay and controversy.
In summary, the CARD position on the biomedical treatment of autism is not one of belief or disbelief, it is one of uncertainty. It is our hope that the coming decade will yield the evidence which is so desperately needed to transform the current debate about biomedical treatment from one based on subjective report to one which is grounded in sophisticated analysis of sound scientific data.
Green, V. A., Pituch, K. A., Itchon, J., Choi, A., O’Reilly, M., & Sigafoos, J. (in press). Internet survey of treatments used by parents of children with autism. Research in Developmental Disabilities.
Howard, J. S., Sparkman, C. R., Cohen, H. G., Green, G., & Stanislaw, H. (2005). A comparison of intensive behavior analytic and eclectic treatments for young children with autism. Research in Developmental Disabilities, 26, 359-383.