3.4 Problems in evaluating autism interventions

Taking a new intervention through a rigorous series of evaluations is usually lengthy and costly – to borrow another example from medicine, think of how long a malaria vaccination has been in development. In the knowledge that, in autism, intervening early in development is best, parents, teachers and even clinicians may feel that they cannot wait for the full outcomes of clinical trials. They may therefore resort to procedures known to be broadly helpful and not actively harmful. Unfortunately, some parents may also feel driven to try extremely risky procedures.

Another problem is the heterogeneity of autism. It may genuinely be the case that an intervention works for some individuals but not for others. In a formal control study or RCT, where relatively large participant groups are compared, there is the risk that benefits experienced by a small subset of the participants receiving the intervention are ‘averaged out’, such that the intervention does not appear to be beneficial overall.

As you will see, some well-established treatments depend crucially on adapting the goals and procedures to the needs of the individual child, making it very difficult to specify standard group procedures and outcome measures as required for an RCT.