Placebo Effect in Autism

 

Placebo effect in autism - Parent/Child Matrix      

A big problem in autism research is the placebo effect.  It could be because the child found the therapy fun and liked all the attention and so showed improved behaviors; or it could be that the parents so desperately wanted to see an improvement, that they imagined it.

In good research, half of the kids receives the trial drug and the other half receive a placebo.  But what happens when both groups show an improvement?  Well if both groups show equal improvement then the therapy has no value.  In almost all the research I have seen, the placebo group shows an improvement.  In one study the placebo group improved 70% on the behaviour rating scale.

We need to conclude several things:-

• Good studies rely on assessment by clinicians, not parents
• If the therapy was fun and included lots of 1:1 attention, then the kid's behavior will improve, regardless of the medical value of the therapy.
• Do not reject a study because the placebo group improve moderately
• Always focus on the relative improvement of the group on the trial therapy vs the control group on the placebo.

Behavioural rating scales

In autism there are many different behavioural rating scales, including :-

• CGI (clinical global impression) Physician and Parental version 
• ABC (abherrant behaviours checklist)
• CARS (Childhood Autism Rating Scale)
• ADOS (Autism Diagnostic Observation Schedule)

Researchers can pick and choose which scale to use and which scale to emphasize.  All these scales are highly subjective.  Different people assessing the same child will get a slightly different answer.  The same person assessing the same child a week later will also get a slightly different answer. 

Tiny Studies and Not so Objective Researchers

Many studies in autism have a tiny number of subjects, sometimes fewer than ten. Often researchers have a vested interest in the research, this is not always a bad thing. As a result it is best to focus on research that has been frequently cited by other researchers, this should mean that they buy into it.

Example - Secretin Research

In a 2003 study of the hormone secretin, 62 children participated. At that time there were stories that secretin was a "wonder cure" for autism. Half received a one-off injection of secretin, the other half received a placebo. The clinicians' tests showed that there was no behavioral improvement in either group; the parents however saw things differently. The parents of 48 children saw an improvement. When asked to guess whether their child had received the placebo or the secretin, 27 guessed correctly and 27 guessed wrong. Six families would not guess and two families dropped out.