Are we over-diagnosing autism? The case against

Basu and Parry (Basu and Parry, 2013) present a thought provoking argument for an ‘epidemic’ of over diagnosis of autistic spectrum disorders (ASD). What are the possible reasons for the increase in the diagnosis of ASD? Is this solely a case of over-diagnosis, leading to incorrect labeling and unnecessary intervention, or are there other factors involved?

One reason for an increase in a particular diagnosis is the availability of effective interventions. Until recently some health professionals had been sceptical about therapies for autism, and nihilistic about the outcome for severe autism in particular. There is now a growing body of evidence to support early intensive behavioural intervention (EIBI) (Thompson, 2013), in particular intensive applied behaviour analysis (ABA) (de Rivera, 2008). Other interventions now available include medication, in particular the atypical antipsychotics, selective serotonin re-uptake inhibitors and stimulants. While medication does not address the core deficits in ASD there is evidence for improvement in some of the associated behaviours (West et al., 2009), although caution is needed due to side effects.

There are also biomedical drivers for the growth of interest in ASD, including the increase in knowledge of ASD neurobiology. It is now estimated that 10% of children with ASD have an identifiable genetic disorder including Rett’s Syndrome, Fragile X Syndrome, Tuberous Sclerosis or Angelman Syndrome (Yang and Gill, 2007) and this percentage is expected to increase. ASD is thought to be the neuropsychiatric disorder most affected by genetic factors, which contribute 90% of variance when heritability is estimated from family and twin studies (Levy et al., 2009). Genes associated with autism are also associated with schizophrenia and bipolar disorder, with the syndrome of catatonia common to all three disorders, increasing interest in ASD beyond intellectual disability researchers (Chagnon, 2005).

What might be other reasons for the rates of ASD diagnosis to increase? Basu and Parry hypothesise that there is diagnostic ‘upcoding’ to allow children to access services. Government funding is available in Australia if a diagnosis of ASD is made before age six (Australian Government Department of Social Services), leading to pressure to make a rapid and potentially inaccurate diagnosis to meet the age funding cut off. With the rollout of the national disability insurance support scheme it could be expected that increased funding would become available for more age groups. Psychiatrists and paediatricians will increasingly become the gatekeepers for access to services. However, this is a two-edged sword, with the potential for families at psychosocial disadvantage being unable to navigate the system rapidly and therefore missing out on effective interventions, in contrast to those who know the rules pushing for a diagnosis to be made.

Is it possible to make an accurate diagnosis of ASD? Although this may need to be reviewed with the changes in criteria in DSM-5, it is possible to diagnose to DSM-IV-R criteria with over 80% accuracy. The gold standard method is using either a multi-disciplinary team to assess history, behaviour, intelligence and social communication (including parent interview), or using detailed semi-structured assessments, such as the Autism Diagnostic Interview-Revised (ADI-R) and Autism Diagnostic Observation Schedule (ADOS) administered in combination (Falkmer et al., 2013). This is a much higher level of diagnostic validity than seen in many other disorders. However, a single interview with a child and family may not provide a reliable diagnosis.

The author agrees with Basu and Parry that psychosocial issues and trauma should not be ignored. However, psychosocial issues are often independent of a diagnosis, and may even be secondary to the stress of living with a disabled child. There is currently little evidence for a syndrome identical to ASD being caused by trauma. However, catatonia has been linked with trauma (Dhossche et al., 2012) and children with ASD have an increased risk of trauma (Reading, 2008).

Basu and Parry caution us that one of the costs of inappropriately diagnosing ASD may be to cause an individual to be socially isolated, spending hours on the computer, on a disability pension. This scenario, while not desirable, seems rather unlikely. The main reason for diagnosis is to expedite access to treatments and services that allow ASD people to reach their full potential, not shut the door on their future.

In summary, the author believes that the increasing awareness of ASD as a disorder amenable to treatment, not only by health, education and welfare systems but also the general community is a good thing. The risk of diagnostic ‘creep’ to access services is, however, possible and relies on psychiatrists or paediatricians working with allied health professionals such as psychologists, occupational therapists and speech pathologists to ensure reliable and valid assessments of ASD.

See Viewpoint by Basu and Parry, 2013, 47(12): 1116–1118.