On the DSM-5 proposal for the definition of Specific learning disorder

16.06.2012 par Franck Ramus, dans dys, psychiatrie

NB: il s'agit ci-dessous de commentaires envoyés à la task-force du DSM-5, le comité chargé de rédiger la prochaine version de la classification des troubles mentaux pour l'association américaine de psychiatrie (actuellement DSM-IV-TR). Mes commentaires concernent la proposition de définition des troubles spécifiques des apprentissages.

Dear colleagues,
Please find below my observations about the proposed definition for specific learning disorders. They result from lengthy discussions that have taken place on the mailing-list of the Society for the Scientific Study of Reading (SSSR), following a post linking to the International Dyslexia Association's petition (http://www.ipetitions.com/petition/include-dyslexia-in-dsm-5/). Nevertheless I am not representing any particular group and the observations below only reflect my personal views.

The first contentious point raised by the DSM5 proposal is whether there should be a single category for all specific learning disorders, rather than one for each disorder. Reasonable grounds for such a change would be, for instance, epidemiological evidence that triply comorbid cases of dyslexia, dysorthography and dyscalculia vastly outnumber cases of a single disorder. Evidence that these disorders are to a large extent independent would count against the proposed change. Either way, I note that no evidence at all is given in the Rationale to justify such a change. Because any change in the classification has important consequences (I will come back to this point below), I would suggest that either robust scientific evidence in favour of the single category be provided, or that the previous categories for each learning disorder be preserved.
   
Secondly, whether the decision is to preserve categories for each disorder or to group them all together, it is important to preserve as much continuity as possible with earlier category names. In the proposed version, the names "dyslexia", "dysorthography" and "dyscalculia" have all but disappeared. Such a change would likely have dramatic consequences, as names have important symbolic functions, and many policies have been organised around such names. Therefore, unless there are very good reasons to change names, they should be preserved by all means. I am not aware of any good, scientifically-grounded reason for changing "developmental dyslexia" to "specific learning disorder/reading". Tremendous efforts have been made in many countries to communicate about dyslexia, to convince many people that it exists, to design social policies to cater for it, etc. Dropping the name altogether would cause immense unnecessary disruptions at many levels from diagnosis to intervention. This was the whole point of the IDA's petition. Therefore I would recommend the most conservative approach to category naming, that is, to preserve the terms "developmental dyslexia", "dysorthography", "developmental dyscalculia", if only as specifiers of "specific learning disorders". Furthermore, if the single category of "specific learning disorders" remains in DSM5, the systematic use of the relevant specifiers should be recommended, as they are absolutely necessary to characterise the disorder of each child and the type of intervention required.

The third problem that I see is that the current proposal is self-contradictory with respect to whether discrepancy criteria should be used for the diagnosis of specific learning disorders. On the one hand, the very term "specific learning disorders" implies a discrepancy (as specific means that the disorder does not affect all cognitive functions equally). On the other hand, the Rationale explicitly excludes discrepancy criteria, arguing for consistency with the IDEA regulations. I think that this contradiction reflects a confusion between two distinct issues: definitions of and criteria for diagnostic categories on the one hand; evaluation of and provision for disability on the other hand. Provisions should be based on the assessment of disability; diagnostic criteria are only one element that enter into the assessment of disability. I will illustrate this in the case of reading and dyslexia.
With respect to disability, most experts of reading difficulties (as represented in the SSSR) seem to agree that all poor readers below a certain level of performance, no matter what the cause may be, deserve appropriate attention. And in practice, in the current state of knowledge, all poor readers typically benefit from the same kind of (mostly educational) intervention. As long as this is the case (this might change with the invention of new forms of intervention for certain poor readers), it may make sense in practice to ignore the various causes of poor reading, and to treat all those children as "poor readers". This is the point of the IDEA regulations. This is akin to a decision tree in medical practice.
However a classification of diseases such as DSM should not attempt to build-in such decision trees. Indeed, the very definition of a mental disorder by the DSM5 task force states that "the diagnosis of a mental disorder is not equivalent to a need for treatment". Diagnostic criteria should be based on scientific evidence as to what seems to constitute the most "natural", coherent categories at some level of analysis (etiology, symptoms, etc.). In the case of dyslexia, the question is: do all poor readers have the same problem, to such an extent that they should be given the same diagnostic category? I don't think so. Poor reading because of low IQ (even within the normal range), or because of poor education, is not the same natural entity as a specific reading disability, "specific" implicitly embedding a discrepancy criterion. This is why in dyslexia research many of my colleagues and I still use such a discrepancy criterion, even though we might not recommend it for intervention purposes.
I would therefore recommend that diagnostic categories be defined on the most scientifically sound criteria (which, in the present case, include a discrepancy criterion), and that DSM leave guidelines for intervention and provision for disability to a separate document or to different institutions.
Sincerely yours,

Franck Ramus

Publier un commentaire