Physical complaints with no apparent organic pathological explanation are very frequent in both primary and secondary care, but medicine has found them difficult to conceptualise and classify. Reference Mayou, Kirmayer, Simon, Kroenke and Sharpe1,Reference Sharpe, Mayou and Walker2 All those who treat such problems need labels for the type of somatic symptom and to have an additional way of classifying the minority who also have a mental disorder. Some of the psychiatric subgroup satisfy criteria for established mental disorders (neurocognitive, mood, anxiety and personality disorders) but many do not. DSM-III 3 invented a whole new terminology for this situation, included in a chapter entitled ‘Somatoform Disorders’. This was an advance in that it recognised a substantial clinical problem. However, the complicated categories (somatisation disorder, undifferentiated somatoform disorder, pain disorder and somatoform disorder not otherwise specified (NOS)) were derived from an idiosyncratic preoccupation of the original DSM-III authors with the least prevalent clinical subgroup - those with chronic multiple ‘unexplained’ symptoms (somatisation disorder). This resulted in categories being largely defined negatively in terms of lack of organic pathology (‘medically unexplained’). A more useful psychiatric classification would have neutrally named categories based on positive psychological criteria. It would also take account of very similar psychological and behavioural reactions to proven organic illness.
DSM-5 has at last appeared. 4 How does it seem to someone who before retiring in 2005 argued that somatoform disorder should be abandoned? Reference Mayou, Kirmayer, Simon, Kroenke and Sharpe1,Reference Sharpe, Mayou and Walker2 Judgement must consider not just the classification but also the chapter text, because the DSM manuals have often been used as textbooks of psychiatry. Indeed, the chair of the DSM-5 taskforce has been quoted as believing that DSM-5 ‘is the best manual for helping clinicians care for patients’. Reference Gornall5
The renamed chapter ‘Somatic Symptom and Related Disorders’ does not depend on any new evidence, but the plain English descriptive title is undoubtedly better. There are three major changes.
(a) Most importantly, the alleged distinction between medically explained and unexplained is abandoned. Reference Sharpe6 It is unequivocally stated that ‘It is not appropriate to give an individual a mental disorder diagnosis solely because a medical cause cannot be demonstrated’. 4 The most controversial DSM-IV 7 category of somatisation disorder (together with undifferentiated somatoform disorder, pain disorder and somatoform disorder NOS) is replaced by a new diagnosis, somatic symptom disorder. This is defined by the psychological criterion: ‘excessive thoughts, feelings, or behavors related to the somatic symptoms or associated health concerns… ’.
(b) It rationalises the categories and subcategories. Pain disorder has gone, body dysmorphic disorder is moved out, and hypochondriasis is renamed and redefined. Conversion disorder is tidied up, and psychological factors affecting other medical conditions and factitious disorder are moved in.
(b) It is no longer dualist stating, as an example: ‘The diagnosis of somatic symptom disorder and a concurrent medical illness are not mutually exclusive […] an individual may become seriously disabled […] after an uncomplicated myocardial infarction even if the myocardial infarction itself did not result in any disability.’ 4
Illness anxiety disorder
Hypochondriasis is renamed illness anxiety disorder and defined as ‘a preoccupation with having or acquiring a serious, undiagnosed medical illness (Criterion A). Somatic symptoms are not present or, if present, are only mild in intensity (Criterion B)’ and ‘the individual’s distress emanates not primarily from the physical complaint itself but rather from his or her anxiety about the meaning, significance, or cause of the complaint’. Spruced up criteria are welcome but the restriction to symptoms of mild intensity goes against all clinical experience and the evidence. Reference Starevic8 The failure to move the category to the ‘Anxiety Disorders’ chapter is surely illogical and a mistake.
Somatic symptom disorder
The new single major and broader category somatic symptom disorder is a significant improvement on somatisation disorder and related DSM-IV categories. Although there have been concerns that the new diagnosis is overinclusive, this is an argument for clarifying a fairly high threshold.
Problems of differential diagnosis arise if there is also a medical condition (i.e. a proven or assumed pathological basis). Although oddly they do not appear in the differential diagnosis, it is apparent elsewhere in the chapter that somatic symptom disorder overlaps almost indistinguishably with two other categories, which seem to have received little working group attention:
(a) Adjustment disorder: ‘Abnormal psychological or behavioral symptoms that develop in response to a medical condition are more properly coded as an adjustment disorder’. It is regrettable that DSM-5, like its predecessors, regards adjustment disorder as a residual category, dismissing highly prevalent clinical problems in three pages. Reference Casey and Doherty9
(b) Psychological factors affecting a medical condition: ‘The difference is one of emphasis, rather than a clear-cut distinction. In psychological factors affecting other medical conditions, the emphasis is on the exacerbation of the medical condition […] In somatic symptom disorder, the emphasis is on maladaptive thoughts, feelings and behavior’. It may be that this category is redundant and it certainly deserves greater critical thought and research.
The chapter authors admit: ‘nonpsychiatric physicians found the DSM-IV somatoform diagnoses difficult to understand and use’ but claim ‘These reconceptualized diagnoses […] are more useful for primary care and other medical (nonpsychiatric) clinicians’. Sadly, the text is incoherent and obscure. Unlike the classification, it remains firmly dualist with one reference to ‘somatizing’ individuals and many to medically unexplained symptoms. It is unable to grasp that symptoms not totally attributable to known or assumed pathology can have a medical aetiology in terms of physiological, psychological and social factors.
The opportunity to set out a clear approach to aetiology of all somatic symptoms is lost. A dense paragraph buried in the text for somatic symptom disorder has the meaningless conclusion: ‘Thus, somatic presentations can be viewed as expressions of personal suffering inserted in a cultural and social context’. There is no sense of the widely accepted common principles of aetiological models which emphasise individuals’ interpretation or experience of minor pathological and physiological bodily perceptions. Reference Mayou, Kirmayer, Simon, Kroenke and Sharpe1 There is no understanding of the importance of the reactions of others and especially of reinforcement by overcautious or inadequate medical explanation and care. Discussion in such terms should have been central to the chapter and would have provided the clinical understanding with which to design interventions.
My verdict must be that although the DSM-5 classification is an improvement on DSM-IV it is a predictable disappointment. The most important change is that it does not require symptoms to be medically unexplained (although the text still does) and instead emphasises psychological criteria. It may be more successful than DSM-IV in satisfying its main motivation of enabling psychiatrists to charge for their services and may be more useful to planners, insurers and lawyers. However, it will not be much used in primary and general medical care or be popular with our patients. A text which reads as poorly collated and contradictory preliminary drafts by numerous working groups will baffle those who want help in treating patients. We need to find ways forward.
Developing the classification
ICD-11 is an opportunity for the logical revision that DSM-5 should have been. This should refine and simplify the categories and modify illness anxiety disorder and move it into ‘Anxiety Disorders’. It should clearly state that any mental disorder coding is additional to a somatic condition or symptom diagnosis. The DSM-5 chapter needs to be rewritten urgently in clear, jargon-free prose directed at primary care and general physicians.
Contributing to general medicine
Even more important than classificatory quibbles is need and opportunity for psychiatrists who work in the general medicine field to move on from the narrow preoccupations of mental disorder to encouraging general understanding of psychological reactions to physical illness and especially of symptoms not entirely attributable to organic pathology. I suggest the key points are:
(a) showing that the range of psychological and behavioural responses to symptoms is very similar, whether caused by cancer or by minor pathology or by awareness of physiological processes;
(b) outlining an aetiological approach that can inform decisions about treatment;
(c) describing how good routine medical care can prevent many somatic complaints becoming chronic and how specialist interventions can treat persistent disabling complications.
It will require a determined effort to work with physicians, including those revising their own symptom classifications such as those for pain and headache. An equally demanding task will be persuading sceptical patients that psychological understanding is not a denial of the reality of their suffering but rather an essential component of all medical care.
I am grateful to Michael Sharpe for discussion of the issues covered in this editorial and to an anonymous reviewer.