6.4 Developments in early identification
Section 2 indicated that securing a diagnosis for a child can be a complex process. People with Asperger's syndrome are especially likely to remain undiagnosed and to grow up considered eccentric or socially aloof.
Therefore, improving identification techniques offers one way to address practical needs.
Is it necessarily better for individuals to receive a diagnosis that ‘labels’ them as early as possible?
For children with severe autism, the benefits of early identification almost certainly outweigh the drawbacks. They are more likely to receive treatment for any medical problems and to gain access to specialist services and appropriate educational support. For children with mild or borderline ASDs, there may be arguments for keeping the boundaries blurred. Yet Clare Sainsbury writes:
Finally getting the right label was one of the best things that has ever happenned to me. By my teens I was seriously depressed after years of being different and not knowing why, and believing that, since no-one gave a name to my problem, I must just be imagining it, or not trying hard enough (after a decade of trying very hard and failing very hard to be like everyone else). This experience was shared by many other people with Asperger's.
(Sainsbury, 2000, p. 31)
Significant progress in early identification comes from instruments such as the CHAT (Checklist for Autism in Toddlers, see Box 12, Baron-Cohen et al., 1992; 1996). This instrument seeks to shift the focus of diagnosis from outward signs towards symptoms that are part of the core socio-cognitive deficit, and this illustrates the fruitful interplay between theoretically driven research and practice.
Box 12: Developing and using the CHAT
The CHAT consists of a questionnaire for parent and health visitor about behaviours thought to serve as early ‘building blocks’ for the development of mind-reading skills, in particular gaze monitoring, proto-declarative pointing and pretend play (see Box 5). These questions are mixed with questions about other developmental milestones, for example:
‘Does your child ever use his/her index finger to indicate interest in something?’ (Proto-declarative pointing.)
‘Does your child like climbing on things, such as up stairs?’ (General motor development.)
Baron-Cohen et al.'s (1992) first CHAT study, mentioned in Box 5, established that the three key behaviours were present in typically developing children by 18 months. They were absent from the behavioural repertoire of a group of older children diagnosed with autism, and from some 18-month-old siblings identified as genetically ‘at risk’.
Baron-Cohen et al. (1996) collaborated with health visitors and GPs on the screening of 16,000 children in Southeast England. The CHAT was administered during a routine developmental check-up at 18 months. Out of all children screened, 12 failed on all three critical items from the CHAT. On further assessment, using a range of the diagnostic instruments described in Section 2, 10 of these children were diagnosed as autistic. The remaining two received a diagnosis of ‘developmental delay’. (This means that they were markedly late in achieving significant developmental milestones, particularly in the area of language and communication. Further delay in this pattern might meet criteria for a spectrum disorder.) Another 22 children failed to show either or both proto-declarative pointing and pretend play. Fifteen of this group also received a diagnosis of developmental delay.
From this study the researchers concluded that the CHAT is successful in screening for the absence of a core group of behaviours, which carries an 83.3 per cent risk of autism. The CHAT was seen as making a useful contribution to early identification of autism, though not as a diagnostic tool in itself.
We can evaluate this study in light of the criteria in Box 11.
Theoretical rationale: The approach is extensively grounded in the empirically well-supported ToM framework.
Methodological considerations: The approach includes many methodological checks – the questionnaire was ‘pre-screened’ to identify key indicators; the study included a very large population; children initially identified by the CHAT were extensively assessed using diagnostic tools such as DSM-IV. However, the test is not 100 per cent reliable. It is, in principle, possible for children to pass the test and yet later develop the symptoms of autism (known as a ‘false negative’).
Ethics: The procedure is conducted in a sensitive way, involving the child with parent and health visitor at a routine check-up. However, children who fail on the three ‘target’ behaviours automatically become part of the group further assessed for autism. If children so assessed turn out not to meet diagnostic criteria (known as a ‘false positive’) this could be a needlessly upsetting experience. In this study, 10 of the children picked up by CHAT did receive a diagnosis of autism, and the remaining two received a diagnosis of developmental delay.
A second ethical problem is that parents did not know that the checklist included with the routine 18-month check-up was actually screening for autism. Had they known, some parents might have withheld their consent.
The ‘false negative’ and ‘false positive’ problems have both been addressed in a follow-up study (Baird et al., 2000). In current use of the CHAT, the false positive rate is almost zero, and the false negative rate has been substantially reduced by a second administration of the questionnaire after an interval. The CHAT is one of a developing range of screening questionnaires aimed at improving early identification across the autistic spectrum: others have been developed for the specialised problem of identifying older individuals with ‘high-functioning’ symptoms.