“The Madness of Normality” – On why the DMS-5 is fundamentally wrong

The DSM (Diagnostic and Statistical Manual of Mental Disorders) is the world widely recognized classificatory system of psychiatric disorders, published by the American Psychiatric Association (APA). It is currently under major revision; the release version DSM-5 is expected in May 2013. The “psychiatrist´s bible“ has overwhelming impact: Inclusion in the DSM carries weight far beyond the psychiatrist’s office. It has major influence on whether insurers will cover therapy for a condition, whether research will be pursued for a specific disease or whether the health technology assessment agencies will approve medications that can be marketed for it.

Many interesting issues in DSM-5 could be discussed: the prevailing categories “substance abuse” and “dependence” will be substituted by “addiction and related disorders”, gender and ethnicity specific distinctions will we introduced and instead of distinguishing different entities like “Autistic disorder” or “Asperger´s disorder”, the manual introduces the term “Autism spectrum disorder”.

In this blog post, I want to focus on one particular innovation: The introduction of so-called risk syndromes. This is a collective term for all those conditions, which do not “yet” meet the “full” clinical diagnostic criteria, e.g. for schizophrenia: In this case, you would suffer from the “attenuated psychotic syndrome”. The aim is obvious: “Young people at risk for later manifestation of a psychotic disorder can be identified“. (http://www.dsm5.org/ProposedRevision/Pages/proposedrevision.aspx?rid=412#). This is a clear paradigm shift towards early diagnosis and prediction in psychiatry. Is this shift ethically justified?

Although these changes seem to be well intentioned, I think the DSM-5 suggestions are fundamentally mistaken and must be perceived as one step more towards medicalizing normality. If an adolescent for example suffers from hallucinations or disorganized speech for at least a month and at least once a week, he can be labelled diagnostically as “pre-schizophrenic”. Interestingly, this diagnosis can be made purely clinically regardless of any biomarkers, genetic testing or neuroradiological examinations. At first sight it seems quite convincing to compare this procedure to diagnosing cardiovascular diseases: “The diagnosis and therapy of schizophrenia at the moment of the first psychosis is comparable to the treatment of cardiovascular diseases at the stage of the first heart attack – it is too late“, says Andreas Meyer-Lindenberg, head of the Central Institute of Mental Health, Mannheim, Germany. I personally think this comparison is inappropriate, because the false positive cases are not comparable at all.

The key to understanding my objection is to look at the natural relationship between specificity and selectivity of a test: “The problem is that every increase in the sensitivity of a psychiatric diagnosis is accompanied by a concomitant drop in its specificity“ (quoted from Allen Frances; interestingly, he – the former head of the DSM-4 committee – is one of the main critics of DSM-5). You can only reduce false negatives by the cost of producing more false positives. And the consequences for these supposedly “pre-schizophrenics“ are grave:

1. They receive unnecessary treatment. The only “sure” thing for these patients personally is the severe side effects of the medication. Furthermore, in a health system with scarce resources, I perceive this to be a big waste!
2. They have to deal with the stigma and the label of psychiatric conditions. I imagine that even the diagnosis itself can worsen the course of disease; comparable to a self fulfilling prophecy.
3. They will have massive difficulties to get life insurance etc.

Another interesting issue is raised by Allen Frances: He predicts “yet another round of costly and dangerous iatrogenic epidemics“, if the DSM-5 would be released in the current version.

Bearing in mind that studies predict, that about 30% of patients with “attenuated psychosis” will actually develop schizophrenia in the future, – in my opinion – the possible benefits for 30% do not outweigh the above named burdens of the false positives.

I would find the DSM-5 suggestions more adequate, if there was a more specific diagnostic test and a more beneficial treatment (better therapeutic-effect / side-effect ratio) available. But the dilemma of false positives remains. This can only be solved, if the scientific progress made in genetics and neurosciences finally had great impact in the clinic. Otherwise the space between disease and health is likely to become a medicalized purgatory!