Prof. Michael Fitzgerald comments on the proposed new criteria for DSM V for autism and offers his considered opinion of its impact on both parents and children.


The proposed criteria for DSM V for autism is a matter of major concern for child and adolescent psychiatrists. The new criteria are inappropriately narrow and will leave many patients now meeting DSM IV criteria losing their diagnosis and their services for autism. You and Shen (2011) showed that only 60% of cases with autism DSM IV met new criteria for DSM V. The new criteria go back to Kanner’s Autism (1943) and are narrow criteria which don’t take into account current research and the broader autism phenotype, which is autism as we meet it in clinical practice. Kanner’s Autism would meet ADI criteria, the gold standard for research on autism, but not appropriate clinical practice. Adam Feinstein (2010) noted that at the International Meeting for Autism Research in London in 2008 many of the most highly regarded researchers in autism in the world ‘lambasted the tool (ADI-R) for missing many cases of autism’, and that it was ‘an expensive and ineffective instrument’.

It is extremely expensive and it is prohibitive for the developing world, and it inhibits the possibility of research in autism in developing countries if they wanted to be published in international journals. At the 2008 meeting which I attended, I heard researchers from Australia complaining about its prohibitive cost. Professor Dorothy Bishop, Professor of Developmental Neuropsychology at the University of Oxford, criticised the ADI.R for the vast amount of time it takes for ‘training’ in the use of the instrument, ‘time for administration and time for scoring, and consensus coding’ (Feinstein 2010). Professor Bishop correctly pointed out that ‘if you could be shown that there were real benefits in accuracy of diagnosis from adopting this lengthy procedure’ then she would be happy to go along with the tedious assessment procedure and instrument (Feinstein 2010).

There is absolutely no evidence for this tedious long-winded assessment procedure, particularly in clinical practice. Professor Bishop correctly concludes that ‘the originators of the instrument have never demonstrated that you actually need such a long process—it is really more an article of faith to them’ (Feinstein 2010). Professor Bishop also points out that in relation to the ADI.R.ADOS that there are ‘plenty of children who come out as meeting criteria on one instrument only, and there seem to be no sensible guidelines on how to proceed, other than to seek expert clinical opinion.’ Professor Bishop recommends ‘doing studies to see what is the minimal set of items you have to get reasonable diagnostic accuracy and I doubt that we really need a three hour interview for each case’ (Feinstein 2010).

Using narrow criteria for autism, Baird (2008) pointed out that you get a prevalence of autism of 25 per 10,000 when you use these criteria. When you use a broader autism criteria (Baird et al 2006), found the true prevalence of 116 per 10,000- Ventola et al (2006) showed that the ADI.R was significantly underdiagnosing toddlers. I am continuing to see parents of children with autism who come to me in great distress and tears because they have been told their children do not meet criteria for autism based solely on this test, when it is absolutely clear to me and parents that the child has the autism broader phenotype, that is Autism Spectrum Disorder.

It looks like now that I am going to see far more of these distressed patients if DSM V is brought in, in its present form, which appears likely—although they are still taking submissions on DSM V autism. I don’t think parents should have to suffer unnecessarily because of the above reasons. Their energy should be put into therapeutic activities for their children, not having to go from one professional to another to get a diagnosis. DSM V will have Autism Spectrum Disorder, but not under Pervasive Developmental Disorder. DSM V is creating a new category (social communication disorder) separate from autism. This will not occur in ICD II. In my view, this new category is not a separate category, but part of autism.