Diagnosis, diagnosis, diagnosis: towards DSM-5

The allocation of diagnoses in psychiatry has always been controversial. Diagnoses usually indicate possible treatments, as well as determining who will receive support from health and social services. Some regard the formulation and sharing of a psychiatric diagnosis as itself therapeutic, in that psychological symptoms can be given meaning and effectively discussed with the patient (e.g., Brody & Waters, 1980). However, detractors also comment on how diagnosis can medicalize patterns of behaviour (Conrad, 2007) as well as the human condition itself (Chodoff, 2002), compound stigma (Sartorius, 2002), pre-determine which interventions are deemed appropriate, and also narrowly define the frameworks through which mental health problems might be addressed. The production of an updated diagnostic manual exposes these controversies, and this has certainly been the case in recent discussions of the forthcoming fifth edition of the Diagnostic and Statistical Manual of Mental Disorders, published by the American Psychiatric Association. Initially, we were led to believe that there would be sweeping changes that would include an increased “dimensional” rather than categorical approach to disorders that would better describe phenomena that are continuous and lacking clear boundaries, and would allow clinicians a rating of severity. However, this emphasis has decreased over time. Along the way, there have been critics of the confidentiality agreements that have to be signed by the working group members (see Collier, 2010), the rush to field trials before the end of the process, and also the likelihood that sub-syndromal diagnoses will be introduced. The initial drafts of the diagnostic system are now available for public review and comment (www.dsm5.org), with the expectation of version five of the manual (known as DSM–5) being published in May 2013. It is important to stress that changes in the manual are backed up by informed opinion, complex negotiations between committee members, as well as clear research evidence, and that there is usually a force of conservatism that prevents major changes. This is clear in looking at the current release. However, even small changes can have unintended consequences. Changes to the previous version seem to have contributed to three false positive “epidemics” (Frances, 2010) – high rates of attention deficit hyperactivity disorder, autistic disorder, and childhood bipolar disorders. Clearly, there were other factors that also contributed, in particular drug companies marketing drugs for these diagnoses directed not only at doctors but also at the general public (Moncrieff & Timimi, 2010; Moynihan, Heath, Henry, & Gotzsche, 2002).

There are of course different diagnostic systems. The one proposed by the World Health Organization called the International Classification of Diseases (ICD–10) is similar but not identical to the system devised in DSM–5, and its current revision to ICD–11 is likely to produce similar difficulties as for DSM–5. However, we have concentrated here on DSM–5 as it is very influential not only in the USA but also across the world in both Western and non-Western cultures as medical education takes a more global form. It has become prevalent in defining groups who are the participants for research studies. In fact it is hard to use any other system when trying to publish in prestigious journals, many of which are US-based. This pervasive influence of the DSM criteria has even been termed “The Americanization of Mental Illness” (Watters, 2010).

The current release for public consideration includes proposals for new diagnoses – including mixed anxiety depression, binge eating, psychosis risk syndrome and temper dysregulation disorder with dysphoria – where the symptoms are shared with the general population. It is also proposed that the threshold for inclusion for some existing disorders be lowered, and a few (but not many) diagnoses are scheduled for removal. Most of these changes imply a more inclusive system of diagnoses where the pool of “normality” shrinks to a mere puddle.

There are specific problems associated with potential “diagnoses” which are made in advance of knowing that a disorder will in fact present itself. This is the case for the Psychosis Risk Syndrome. In this “disorder”, attenuated symptoms of schizophrenia are present which, in some cases, may lead to a later florid onset of psychosis. The reason for identifying this is because many clinicians believe that early treatment will provide benefits and avert some of the toxic effects of psychotic experiences. This belief has some validity, which is currently being further tested, but it is not, as yet, based on sound evidence and there are clear negative consequences of such a diagnosis for those concerned, particularly issues of stigma covered, for example, by Ben-Zeev, et al. (2010). There is also the obvious problem of providing diagnoses to some who will never experience the full-blown disorder, usually known as the problem of false positives. It is a bit like telling ten people with the common cold that they are “at risk for pneumonia syndrome” when only one is likely to get the disorder. In addition, one of those people might also have developed pneumonia because they had a predisposition which was nothing to do with the presence of attenuated symptoms, for instance they might be HIV positive. Their proneness to the disorder was therefore entirely unrelated to the preceding cold. If this were to be replicated in the Psychosis Risk Syndrome, then identifying the diagnosis would have negative consequences for many who would never develop the disorder in terms of increased stigma and possible discrimination, as well as in terms of changing their very sense of personhood (in being described as a person “at risk” of developing a disorder commonly regarded with great apprehension). As well as the distress likely to be experienced by themselves and their families, and the suffering entailed by treatment irrespective of whether it is drug or psychological therapy, the logic entailed by Psychosis Risk Syndrome might also divert attention away from understanding the underlying causes of schizophrenia.

What is diagnosis for?

Making a diagnosis is not – as many might mistakenly imagine – an essential part of treatment decisions. It may be one of the reasons, but clinicians also use a range of other information to make judgements over treatment, and making a diagnosis in psychiatry rarely leads directly to a recipe for treatment success. Treatment success is, after all, highly dependent on the relationship between the clinician and the patient, as well as the patient's own views about treatment acceptability as well as their personal circumstances. However, diagnoses might be useful in providing both patients and families with the recognition that the array of symptoms is known, “real”, and that it may lead to treatment. Diagnoses are also accepted by society as reasons for a claim for health and social care services, as well as providing the patient with “a mantle for his [sic] distress that society will accept” (Cassell, 1976).

The framework of psychiatric diagnosis employed in the DSM operates by assuming that disorders are stable entities that transcend their embodiment in, and meaning for, any individual patient (Lakoff, 2005). Such a model, of course, eases the task of developing and operationalizing treatment as well as research protocols. But a number of philosophers, sociologists, medical anthropologists and service user researchers have pointed to the complex way in which diagnoses can help to bring into being the very phenomena and self-attributions that they purport to describe, as well as produce varied responses from individuals living “under the description” of a psychiatric diagnosis (Horn, Johnstone, & Brooke, 2007; Jutel, 2009; Martin, 2007; Young, 1995). Both the meanings that people attach to the feelings and behaviours they are experiencing, as well as societally sanctioned explanations for these feelings and behaviours play an important role in shaping people's ways of embodying a psychiatric diagnosis. Consider, for example, the recent report of a new phenomenon: patients in Britain approaching psychiatrists with self-identified bipolar disorder (Chan & Sireling, 2010). This suggests that, despite the ongoing and pervasive stigma attached to a “severe mental illness” diagnosis, individuals are beginning to understand as well as self-describe some of their own patterns of behaviour and emotional variability as a list of symptoms that – when gathered together – make them people “with bipolar”. It is likely, then, that the introduction of new diagnostic categories to DSM–5 will play an important – though as yet not fully understood or anticipated – part in reshaping the ways in which both individuals as well as society conceptualize both mental distress and “patienthood”.

Given the public interest and the debates within the newspapers and other media, the editors of the Journal of Mental Health thought that it was timely to produce a special section that deals with the question of diagnosis in more detail. The papers are mostly critical of current conceptualizations of psychiatric diagnosis, but do make some suggestions about how these may be changed. The paper by van Os (2010) sets out the problem for one diagnosis, schizophrenia. Van Os argues for a dimensional approach not a categorical one, because the symptoms reported in the clinic are so often reported by members of the general public. He also suggests more clarity in our labelling, particularly for disorders where we now understand more about the underlying causes and which have attained a lot of stigmatizing baggage. His name for schizophrenia would be “Salience Syndrome”, which he explains in detail in the paper. More criticism comes from Jerome Wakefield whose thesis is that DSM–5 is missing a trick. Firstly, by using a symptom-based set of criteria, there is little room for discussions about what is ordinary trait variation in the human condition and what is part of a disorder. Secondly, by attempting to use reliable and scientific criteria (symptoms) it has taken the context out of diagnosis. For him, context is vital to distinguish, for instance, between sadness and depression. Finally, he argues that disorders are not well predicted by context factors alone, e.g., stressors do not well predict depression, rather it is the personal meaning that these stressors have for the person that affects their emotional response. This last issue is echoed in the paper on the personal meaning and conceptualization of mental disorder by Bolton (2010). The effects of diagnosis on public and self-stigma are highlighted and discussed in the paper by Ben-Zeev et al., 2010. The commentary suggests that the proposals for DSM–5 make some possible beneficial additions but also warns, alongside the other authors, that some of the proposed changes are likely to increase rather than decrease stigma.

Our journal is unique in emphasizing the value of contributions from and with service users (or consumers), and so we have also not only asked for papers from esteemed clinical experts but also from service users themselves. These service user papers are not directly related to the potential changes in DSM–5, but rather reflect on how diagnosis has had an impact on the lives of those in the public eye. We asked a number of people to contribute and our list here includes doctors, psychiatrists, politicians and authors. They constitute only a few of the people we asked to contribute and surprisingly our final sample contains only men. We understand that the stigma of mental ill health is pervasive and that this affects whether people wish to dwell on their experiences of mental ill health when they are devoting time to a career in the arts, industry or politics. We are therefore grateful to all who felt they could contribute to an edition that we hope will dispel some of the stigma of mental ill health as well as interrogate the diverse and sometimes surprising ways in which individuals respond to receiving a psychiatric diagnosis.

Acknowledgements

Both authors acknowledge the financial support of the NIHR Biomedical Research Centre for Mental Health, South London and Maudsley NHS Foundation Trust/Institute of Psychiatry (King's College London).