2.4 Problems and benefits of diagnosis
The checklist format of the criteria is specifically designed to accommodate individuals who have different numbers and patterns of criterial symptoms, as well as difficulties that also occur in other disorders. Nonetheless diagnosis must involve placing an individual's profile of difficulties within a category and, inevitably, one concern is that such diagnosis could sometimes be unreliable, with serious consequences for individuals and their families. In practice, however, research suggests that clinicians are reliable in differential diagnoses between the autistic spectrum and other disorders. Box 2 describes how Klin et al . (2000) investigated this.
Box 2: Evaluating diagnostic reliability
Klin et al. studied the measure of agreement (inter-rater reliability) between different clinicians (including psychiatrists, clinical psychologists and speech and language pathologists) when diagnosing the problems of over 900 participants. A substantial number of these were thought to have autistic spectrum disorders. Experienced clinicians showed a very high measure of agreement on differential diagnoses, not only when using the DSM-IV-TR™ criteria, but also when basing their diagnoses solely on experienced clinical judgement. For the less experienced clinicians, judgements were extremely reliable when directly based on DSM-IV criteria, but less so when based exclusively on the clinicians' judgement. All clinicians agreed strongly in deciding which participants had symptom patterns within the autistic spectrum and which called for alternative diagnoses. However, the level of agreement dropped for differential diagnoses of ‘sub-types’ within this spectrum. The reasons for this uncertainty in diagnosing ‘sub-types’ will be considered further in Section 3 .
Another possible criticism of the diagnostic approach is that it involves an inappropriate application of ‘the medical model’ to autism and thus ‘pre-judges’ the causes of the condition as biological. Yet medicine includes many different ideas about cause, ranging from the notion that biological organisms cause diseases such as the common cold, to the notion that psychological stresses may contribute to high blood pressure (Roth and Kroll, 1986). Similarly the diagnostic approach to autism embraces a variety of explanatory perspectives. Most of these agree that biological influences play a role in causing autism (discussed in Section 5 ), but they vary widely in how they link these biological influences to functioning at socio-cognitive and behavioural levels.
A further concern is whether a diagnosis of autism serves as a ‘label’ from which a person cannot escape, even if his or her symptoms have ameliorated or disappeared. However, the balance of arguments for and against labelling may be different for different disorders. ASDs tend to be pervasive disorders, i.e. involving most areas of the person's psychological functioning. Arguments for highlighting these difficulties by labelling may be stronger than for other less pervasive conditions.
Specialists and parents frequently argue that the decision to diagnose a child or adult with autism represents the first step towards helping them. Parents may have experienced years of perplexity, distress and frustration because the problems of their children are misunderstood. The diagnosis provides an explanation for their children's behaviour, helps them to understand and cope with their special and distinctive difficulties, and facilitates access to special educational and therapeutic facilities.
Despite these practical benefits, there remains the ethical issue of whether labels serve to place people in ‘pigeon holes’ that deny their individuality and their humanity. There are some sharply contrasting views about this. Clare Sainsbury (2000) points out that the concern to avoid labelling is often linked to a belief in ‘normalisation’ – the belief that the best way to revalue people with disabilities is by denying or de-emphasising their difference. She points out that:
Instead of starting with the needs, choices and values of a disabled individual, [normalisation] starts with the unchallenged standards of ‘normal’ people … Far from seeming radical or positive, the philosophy of normalisation seems painfully familiar to those of us whose very disability lies in our ‘differentness’.
(Sainsbury, 2000, p. 33)
For Sainsbury, then, the label ‘autistic’ validates the right of individuals to be different.
Yet the expression of difference can be ‘normalist’ too. Diagnostic systems such as DSM-IV-TR™ are intentionally expressed in ‘normalist’ language to highlight the ways in which the behaviour and experience of people with autism departs from statistical norms. The distress that this can cause is well expressed by the writer Valerie Paradiz (2002), the mother of a child with autism:
In the DSM-IV, there are words which I cannot bring myself to say about Elijah. Words like ‘lack’, ‘deficiency’, ‘impairment’, and ‘failure’. Condescension litters the DSM-IV and betrays a burdensome psychiatric history.
(Paradiz, 2002, p. 59)
There is no easy remedy for the difficulties expressed by Paradiz. All discussions of clinical problems, including the present unit, must engage with formal and normalist language to some extent. But this needs to be balanced by a respect for people with autism as individuals with the same variability of personality and outlook as the rest of us. Some approaches to differentiating across the autistic spectrum, in order to express peoples' differences, are evaluated in the next section.