#dsm-iii

Millon's subtypes:

(Millon, ed.).

About Ma/SDPD

Ma/SDPD is similar to avoidant, dependent, obsessive-compulsive, depressive / melancholic, negativistic / passive-aggressive and borderline PDs. It's part of what Millon & Bloom term the "Compliant Personality Patterns", along with OCPD, DPD & HPD.

Differential diagnoses include ongoing abuse, anxiety disorders, somatic disorders, and mood disorders.

The most common PD comorbidities with Ma/SDPD are AsPD (22.17%), Depressive / Melancholic PD (16.74%), & SzPD (12.22%). The least common was HPD (4.07%). Less than 8 percent (7.24%) had only ("pure") Ma/SDPD [much higher than those who had pure Negativistic / Passive-Aggressive, Depressive / Melancholic and Sadistic PDs] (Millon & Bloom).

Millon defines it on a spectrum from aggrieved -> masochistic (self-defeating) (Millon Personality Group); or alternatively from abused [personality type] -> aggrieved [style] -> self-defeating [disorder] (Millon). It also exists on a spectrum from self-sacrificing -> yielding -> masochistic (Millon, ed.).

Originally called masochistic PD, the name was changed to self-defeating PD in the DSM-III-R "to avoid the historic association of the term masochistic with older psychoanalytic views of female sexuality and the implication that a person with the disorder derives unconscious pleasure from suffering" (DSM-III-R). However, Millon & Bloom write that the specific name chosen is pointless, because "all personality disorders are “self-defeating.”"

Childhood trauma is a predisposing factor (DSM-III-R).

Herman argues that Ma/SDPD is a misdiagnosis of Complex PTSD, due to victim blaming and sexism (she also argues the same for Dependent and Histrionic PDs).

In the DSM-III-R it was described as being characterised by “ubiquitous self-defeating behavior such as repeatedly entering into unsatisfying and hurtful relationships, avoiding opportunities for pleasure, rejecting relationships with seemingly caring people, and repeatedly rendering ineffective reasonable efforts by others to help the person" (Coolidge & Segal).

Ma/SDPD "is a mixing or confusion of the usual pleasure-seeking drive with the pain avoidance. As a result, these individuals appear to create personal and social discomfort in their lives. Although it is often reported that they seem to feel comfortable only with guilt and shame, they are also believed to use their self-deprecation as a social strategy to gain support from others" (Millon & Bloom).

In Ma/SDPD, "the individual experiences what is emotionally painful as a means of fulfilling his or her survival aims" (Millon, ed.).

Ma/SDPD only ever appeared in the appendix of the DSM-III-R, and it was dropped because it was associated with 'feminine sexual submissiveness' (Millon, ed.).

References

Coolidge, Frederick L., & Segal, Daniel L., ‘Evolution of Personality Disorder Diagnoses in the Diagnostic and Statistical Manual of Mental Disorders’, Clinical Psychology Review, 1998, vol. 18, no. 5, pp. 585-599.

Herman, Judith Lewis, Trauma and Recovery, 2015.

Millon, Theodore, & Bloom, Caryl, The Millon Inventories, 2008.

Millon, Theodore, Disorders of Personality, 2011.

Millon, Theodore, ed., Personality Disorders in Modern Life, 2004.

Millon's subtypes:

(Millon, ed.).

About SaPD

SaPD is similar to antisocial, negativistic / passive-aggressive, paranoid, and narcissistic PDs. It's part of what Millon & Bloom term the "Aggressive Personality Patterns", along with AsPD, NPD, & Negativistic / Passive-Aggressive PD.

Differential diagnoses include anxiety disorders, mood disorders, substance abuse, AsPD, psychotic disorders (especially with paranoid & persecutory features), and coercive sexual sadism disorder.

The most common PD comorbidities with SaPD are AsPD (36.45%), Negativistic / Passive-Aggressive PD (24.04%), & NPD (14.09%). The least common was OCPD (0.77%). Less than 1 percent (0.31%) had only ("pure") SaPD [less than those who had comorbid OCPD] (Millon & Bloom).

Millon defines it on a spectrum from assertive -> negativistic (Millon Personality Group); or alternatively from assertive [personality type] -> denigrating [style] -> sadistic [disorder] (Millon).

In the DSM-III-R it was described as being “characterized by pervasive cruel, demeaning, humiliating, and aggressive behavior directed toward others, along with a basic lack of empathy and respect for others" (Coolidge & Segal).

Millon also calls it "Aggressive Personality Disorder" (Millon & Bloom).

People with this disorder "might engage in risky behavior undaunted by danger, and [they] might be forward and inhibiting to others. This kind of person is often described as strongly opinionated, closed-minded, unbending, energetic, hardheaded, competitive, and malicious. [They] might be cold-blooded and detached from awareness of the impact of [their] own actions. Sexual energy might lead to imprudent and unseemly behavior" (Millon & Bloom).

People with SaPD are "socially aggressive and seek to dominate those around them", and "are likely to identify vulnerabilities in [people] and exploit them, producing strong fearful or painful emotional responses" (Millon & Bloom).

Not everyone with SaPD is violent, and some inflict pain by other non-physical means; "[t]he person may inflict pain or suffering by lying; for example, a woman may call her former husband and lie to him about their son's having been seriously hurt" (DSM-II-R).

"People with this disorder rarely experience depression, and their reaction to feeling abandoned is usually anger" (DSM-III-R).

SaPD is often caused by childhood trauma, especially domestic abuse (DSM-III-R). The underlying reason for symptoms is often pain avoidance; "[a]lthough sadistic individuals do seem to acquire a perverse pleasure in inflicting pain on others, the underlying motivation seems to be in their own pain avoidance by inflicting overwhelming pain on others before it can be done to them" (Millon & Bloom).

"... sadists may think of themselves as energetic, assertive, and realistic. What is dominating and callous to others is competitive and not overly sentimental to the sadist" (Millon, ed.).

SaPD only ever appeared in the appendix of the DSM-III-R, and if a person met criteria for this PD they were diagnosed with PDNOS (Coolidge & Segal).

"... it was dropped because of scientific concerns, such as the relatively low prevalence rate of the disorder in many settings. However, there were also political reasons. Physically abusive, sadistic personalities are most often male, and it was felt that any such diagnosis might have the paradoxical effect of legally excusing cruel behavior” (Millon, ed.).

References

Coolidge, Frederick L., & Segal, Daniel L., ‘Evolution of Personality Disorder Diagnoses in the Diagnostic and Statistical Manual of Mental Disorders’, Clinical Psychology Review, 1998, vol. 18, no. 5, pp. 585-599.

Millon, Theodore, & Bloom, Caryl, The Millon Inventories, 2008.

Millon, Theodore, Disorders of Personality, 2011.

Millon, Theodore, ed., Personality Disorders in Modern Life, 2004.

The much awaited fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) was published by the American Psychiatric Association (APA) in 2013 (APA, 2013). When one looks at the list of disorders included, and at the sets of diagnostic criteria, the DSM-5 is strikingly similar to the DSM-IV (APA, 1994). At the same time, at the structural, or conceptual, level there have been radical changes; for example, the definition of ‘mental disorder’, which delimits the scope of the classification, is importantly different, and the multiaxial approach to diagnosis, characteristic of earlier editions of the DSM, has been abandoned.

This paper aims to understand the DSM-5 through situating it within the context of the historical development of the DSM series. The first part, ‘Stasis’, considers why it is that the DSM-IV and DSM-5 are so similar in the disorders they include, and in the sets of diagnostic criteria. Of course, a few changes have been made, and those changes are important; the DSM is now so influential that even the smallest revisions can potentially affect the lives of millions worldwide. However, considered overall, when it comes to the diagnostic criteria there is largely stasis. In a piece introducing the new edition, David Kupfer and Darrel Regier, who chaired the revision process, tell readers ‘by and large, there were not sweeping changes in the diagnostic criteria for most disorders’ (Regier, Kuhl and Kupfer, 2013: 95).

At the same time, at the structural, or conceptual, level much has changed. The second part of this paper, ‘Change’, considers four key revisions: (1) the multiaxial approach to diagnosis has been abandoned; (2) the definition of mental disorder, included in the introduction, has been revised; (3) the DSM-5 embraces theory, while earlier editions sought to be atheoretical; (4) ethical questions have been systematically considered in revising the classification for the first time. Thus at the conceptual, meta-, level the DSM-5 and the DSM-IV are importantly different.

This then raises a question: How is it that the conceptual framework of the DSM can be radically revised while the diagnostic criteria that it contains remain the same? In the third part of this paper, ‘Looseness’, I argue that the way the DSM-5 was constructed means that the overall conceptual framework of the classification only barely constrains the sets of diagnostic criteria it contains.

Stasis

I will argue that the diagnostic criteria in the DSM-5 are similar to those in the DSM-IV because it has become difficult to change them. With recent editions of the DSM, revision takes many years, costs millions of dollars, and involves hundreds of experts working in many workgroups and subcommittees. The chairs of the DSM-5 Task Force, David Kupfer and Darrel Regier, were appointed by the APA in 2006 and oversaw the process of constructing the DSM-5 (APA, 2013: 6). Workgroups of specialists worked on the various sections of the manual (mood disorders, substance-related disorders, and so on). Each workgroup reviewed relevant literature published since the publication of the DSM-IV, drafted proposals for revisions to diagnostic criteria in their area, and then solicited and responded to feedback on their proposals. Throughout the revision period, mental health conferences and journals were full of discussion about proposals for revision. Draft versions of the DSM-5 were periodically posted online, and anyone who wished could contact the workgroups with their ideas and suggestions. Field trials, where clinicians used draft criteria to check they could be reliably applied, were used to test some draft diagnostic criteria. Behind the scenes, various interest groups (representing different types of therapist, and patient, but also business groups) lobbied the APA in attempts to get the classification to fit their interests. Finally, the DSM had to be voted through by the various committees of the APA. The revision process was lengthy, and involved many distinct actors. As a consequence, the initial plan for the new edition, and the finished product, could markedly diverge.

In the case of the DSM-5, a radical reshaping of the classification was initially proposed. A Research Agenda for DSM-V (the Latin numeral changed later) set out the ambitions (Kupfer, First and Regier, 2002). It began by detailing problems with the DSM series to date. The DSM-III, published in 1980, sought to be a purely descriptive classification that made no use of unproved theoretical assumptions (APA, 1980: 6–8). In the late 1970s (when the DSM-III was under construction) many researchers believed that the unreliability of diagnosis was holding back research. The descriptive diagnostic criteria of the DSM-III were designed in part to improve the reliability of diagnosis. The idea was that once syndromes could be reliably identified, research could be expected to make progress in identifying the causes of psychopathology. Since DSM-III, however, research into the causes of psychopathology has failed to live up to expectations. A Research Agenda outlined how the descriptive syndromes included in the DSM may now be slowing progress. Many researchers suspect that theoretically interesting populations will likely not map on to DSM categories. A Research Agenda agreed that researchers thus need a different style of classification, but it was unclear exactly what sort of classification might be better, and the publication concluded that some ‘as yet unknown paradigm shift may need to occur’ (Kupfer et al., 2002: xix).

In the event, seeking to construct a radically different sort of classification proved too ambitious. Partly this was because the exciting research breakthroughs that A Research Agenda saw on the horizon – in genetics, neuroscience, developmental science, and so on – failed to produce results quickly enough to inform a new classification. But even in cases where researchers agreed that there was sufficient evidence for change, it turned out to be unexpectedly hard to revise the sets of diagnostic criteria included in the DSM. At the end of the process, co-chair David Kupfer described it as ‘an aggressive, conservative document’; in his view the committees were aggressive in their pursuit of revision, but conservative in their decisions in the end (Levine, 2013). Commentators agree that the sets of diagnostic criteria in the DSM-5 and DSM-IV are highly similar (Aragona, 2015; Paris, 2015: 52; Poland, 2015: 34). In this part, I argue that diagnostic criteria included in the DSM-5 have become hard to revise because classifications such as the DSM are ‘path dependent’, and the DSM has become ‘locked-in’.

The QWERTY keyboard layout is the classic example of technological path dependence leading to lock-in (David, 1985). Currently, the QWERTY layout is nearly universally used, but the reasons for this are historical. When keys were mechanical, QWERTY was the design that reduced the chances of physically adjacent keys being pressed in succession and so jamming together. Typists trained on a QWERTY keyboard found it not worth their while to shift to any other. Nowadays keyboards no longer jam, and if keyboards were being introduced de novo it is likely that a different design (such as Dvorak) would enable faster typing. Still, given that the use of QWERTY is established, the costs of moving to a different layout are too great. QWERTY has become ‘locked-in’.

I suggest that when classification systems come to be employed widely, by diverse actors, and where their continued existence relies on those actors continuing to be satisfied, they become prone to lock-in. To see how the diagnostic criteria included in the DSM have come to be locked-in, we need to consider the multiple users of the DSM, and how their needs constrain the classification.

Most straightforwardly, the DSM is a classification produced by and for the members of the American Psychiatric Association. Once developed by the various workgroups and Task Force, the draft DSM has to be agreed by various APA committees before it is accepted. The DSM thus has to be broadly acceptable to its members. Some members are researchers, most are clinicians. This gives rise to diverse requirements, for example clinicians tend to prefer simpler classifications than researchers (Mullins-Sweatt, Lengel and DeShong, 2016).

Members of other mental health professions in the USA also commonly use the DSM. They buy copies in large part because they use codes from the book for filling in insurance forms (Frazer, Westhuis, Daley and Phillips, 2009). These buyers are numerous, and sales to them are important for the APA’s finances. Between 2005 and 2011, the APA earned $5–6 million each year from sales of the DSM-IV (APA Treasurer, 2012).

Research in mental health typically employs DSM criteria. The use of the DSM in research cements the scientific respectability of the DSM. As a consequence, it is advantageous to the APA that the DSM be broadly acceptable to mental health researchers around the world.

The World Health Organization (WHO) publishes its own classification: the International Classification of Diseases (ICD) (WHO, 1992). The ICD covers all health conditions, but includes a chapter on mental disorders. Over the last few decades, the development of the DSM and the ICD has been coordinated, and the two classifications are now nearly identical (APA, 1994: xx–xxi). The compatibility of the ICD with the DSM helps to ensure the respectability of the DSM as ‘the’ classification of mental disorders.

Non-mental health professionals and the lay public rely on the DSM to define mental disorders, and conversely normality. DSM categories feature in legislation and the guidelines of many bureaucracies (particularly in the USA, but also around the world). The APA is sensitive to public criticism of psychiatry,1 and is thus motivated to attempt to ensure that the DSM continues to appear respectable in the eyes of lay people. Of course, many of those who use the DSM are not completely satisfied with the system and grumble about it. But they continue to use the classification because it is ‘good enough’ for their purposes. I will show that revising the diagnostic criteria while ensuring that the DSM continues to be ‘good enough’ for all its users has become extremely difficult.

Dense networks of uses lead to constraints

Let us consider the constraints that can be created by the interconnections among the DSM, US medical insurance, the ICD, the needs of service users, and legislation. The DSM is used by many mental health professionals in the US for completing forms that are required for insurance coverage. Keeping the insurance industry on board places some constraints on the DSM – most obviously, a very radical expansion of the domain of mental illness would threaten its continued use of the system. To ensure that insurance companies continue to be satisfied with codes taken from the DSM, it is not sufficient simply to check with insurers whether they are happy with any proposed revisions. The codes included in the DSM used on insurance forms are actually ICD codes, which are mandatory because they are required by the US Health Insurance Portability and Accountability Act (1996). This means that for sales of the DSM to be ensured, the codes in it have to be compatible with ICD codes. The DSM and ICD cannot diverge very far from each other before this ceases to be the case.

To make matters more complicated, the USA does not use the standard version of the ICD, but a ‘clinical modification’ developed especially for use in the USA. Development of the US modification lags years behind the revision schedule of the ICD. The WHO published the ICD-10 in 1990, but the US version of this system was not ready until 2014. This delay meant that the DSM-5 had to be broadly compatible with different versions of the ICD. An APA website explains:

DSM-5 contains both ICD-9-CM codes for immediate use and ICD-10-CM codes in parentheses. The inclusion of ICD-10-CM codes facilitates a cross-walk to the new coding system that will be implemented on October 1, 2014 for all US health care providers and systems, as recommended by the Centers for Disease Control and Prevention’s National Center for Health Statistics (CDC-NCHS) and the Centers for Medicare and Medicaid Services (CMS). (Anon., 2013)

But the complications are not yet done. The USA has only just moved to ICD-10, but the rest of the world has moved on, and ICD-11 is due to be published by the WHO in 2018. Compatibility with this version is also a prerequisite to ensure long-term use of the DSM. The APA staff work ‘closely with staff from the WHO, CMS, and CDC-NCHS to ensure that the two systems are maximally compatible’ (Anon., 2013).

Concerns about maintaining viability for insurance constrain the general structure of the DSM – which has to maintain compatibility with the ICD. Insurance concerns can also limit revision at the level of individual diagnoses. In the DSM-5 the constraints are clearly visible if one considers the revisions that have been made to the autism-related disorders. In DSM-5 a number of previously distinct autism-related conditions have been merged to create a new category of autistic spectrum disorder (ASD). In DSM-5, ASD includes most of those diagnosed under DSM-IV with autism and most of those diagnosed with Asperger’s disorder. The change was justified on the basis that the supposed distinction between autism and Asperger’s was of questionable significance, and in any case could not be reliably drawn (Happé, 2011).

Having a diagnosis of Asperger’s or autism often makes a big difference to someone’s life. In the USA, children with such diagnoses can often be entitled to very expensive forms of one-to-one therapy. Importantly, such provision is often conditional on the diagnosis; if the diagnosis goes, so does the support. When the classification alters, certain changes make little difference, but others can have a huge effect. For someone who had a diagnosis of Asperger’s under DSM-IV, to be relabelled as having ASD would probably be of little consequence; this would merely be seen as a renaming of the same condition. It would be far more problematic if someone with a DSM-IV Asperger’s diagnosis was to end up with no mental disorder label, or a label with very different connotations – say, a speech disorder. In such a case, a child would be at risk of losing their access to therapy.

During the period of DSM-5 development, the potential effects of the proposed revisions to autism-related disorders were hotly contested. The difficulty in estimating the effects of proposed changes arose because multiple small changes to the criteria were to be implemented simultaneously. In the DSM-IV, the main difference between children with autism and children with Asperger’s was that those with Asperger’s showed no significant delays in early language skills. In merging the disorders, in DSM-5 the criteria relating to problems with language development, previously included in the DSM-IV as symptoms of autism, have been removed. Other changes were also made, for example in the age by which symptoms must be manifest.

A number of studies using draft DSM-5 criteria predicted that a significant number of those diagnosed with Asperger’s under DSM-IV would not receive a diagnosis of ASD under DSM-5. Instead, some would receive a diagnosis of ‘pragmatic speech disorder’, a diagnosis with very different connotations (Matson et al., 2012; Mattila et al., 2011).

Autism groups are well informed and well organized, and were alert to the potential ramifications of changes to DSM criteria on service provision. They came together to protest against the proposed revisions, and to argue that broad definitions of ASD should be maintained (Greenberg, 2013: 296–9). Politicians were drawn into the disputes. In a number of states, legislation was proposed to ensure that all those with DSM-IV diagnoses would continue to be eligible for services (e.g. Connecticut General Assembly, 2013; Illinois General Assembly, 2013).

Eventually the DSM committees back-tracked. The published DSM-5 includes a note after the new diagnostic criteria for ASD, stating that ‘Individuals with a well-established DSM-IV diagnosis of autistic disorder, Asperger’s disorder, or pervasive developmental disorder not otherwise specified should be given the diagnosis of autism spectrum disorder’ (APA, 2013: 51). Given that there are clear differences between the old diagnostic criteria and the new, this note is extraordinary. In effect, it says that some people who do not meet the diagnostic criteria for ASD should still be given the diagnosis.

In the case of ASD, we can see that it has not been possible to revise the DSM and also to produce a revision that will be ‘good enough’ to satisfy the needs of its various users. The note has been inserted to satisfy patients and their supporters, and maybe it will enable people diagnosed with Asperger’s under DSM-IV to retain access to services. But it seems inconceivable that these instructions could be employed by researchers seeking to investigate ASD. It would make no sense to select a subject population for research consisting of people who meet DSM-5 criteria for ASD plus a few who do not but who satisfy DSM-IV criteria for Asperger’s. Here attempts to keep diverse groups satisfied have resulted in a botched definition.

It is true that, in the case of ASD, revisions have been made to the DSM, thus this is not a pure case of ‘lock-in’. However, the fact that the revision could only be done clumsily is a good illustration of how the DSM has become highly resistant to revision. Revision has become difficult and here the strains show.

The personality disorders section of the DSM-5 provides another clear example, showing how hard it has become to revise the sets of diagnostic criteria included in the DSM. Early on in the revision process, the dominant view among researchers was that the DSM should shift towards a dimensional system for personality disorders, and a radically revised classification for personality disorders was proposed (Skodol et al., 2011). There were fears, however, that the proposed new system would be hard for clinicians to use in practice (Skodol et al., 2013). It proved impossible to find a satisfactory compromise that balanced the needs of both researchers and clinicians. Eventually, attempts at revision were abandoned, and the DSM-5 personality disorder section simply reprints the diagnostic criteria from DSM-IV.

I conclude that at the level of diagnostic criteria the DSM-5 is very similar to the DSM-IV, because at this level the DSM has become locked-in. It has become very difficult to make revisions and keep all users of the system on board, and so revisions have been avoided.

Change

At the structural, or conceptual, level the DSM-5 differs from earlier editions in important respects. Here I discuss a number of key changes that have been selected to demonstrate that significant change has occurred.

Multiaxial diagnosis abandoned

At the structural level, the most obvious change between the DSM-5 and earlier editions is that the multiaxial diagnostic system has been abandoned. In DSM-IV, a complete diagnosis involved attention to five axes. Diagnostic criteria for axis I (clinical mental disorders) and axis II (pervasive psychological issues, such as mental retardation and personality disorders) were provided within the DSM-IV itself. Axis III recorded potentially relevant general medical conditions (using ICD codes). Axis IV noted psychosocial and environmental problems. These were to be selected from a possible list, for example ‘Problem related to the social environment: living alone’, or ‘Occupational problems: discord with boss’. Axis V recorded a Global Assessment of Functioning (GAF). The GAF score varied from 0 (worst) to 100 (best) and gave an indication of the patient’s overall level of functioning. In practice, many clinicians failed to record information on all five axes. While those who employed the DSM always used axis I, some used only this, and many used only axes I to III, although a complete DSM diagnosis required information on all five axes.

In DSM-5, these axes have gone. Clinicians are now instructed to document the DSM diagnostic criteria that a patient meets and any relevant non-psychiatric medical conditions (APA, 2013: 16). There is no longer a distinction drawn between Axis I (primary mental disorders) and axis II (mental retardation and personality disorders). Clinicians can also record social and environmental problems that may be relevant (and ICD codes are listed in the DSM to facilitate this). It is left unclear whether clinicians should also be recording level of disability; the introduction tells clinicians that there is now a ‘separate’ (i.e. non-DSM) notation for documentation of disability, and then a few lines later says that this system (the WHO Disability Schedule) is included in an appendix of ‘Emerging Measures and Models’ (thus implying that it is not yet to be used in standard clinical practice) (APA, 2013: 15, 23–4).

One might wonder whether the loss of the multiaxial system is simply a change in formatting – what does it matter whether one records information on a variety of ‘axes’ or in a simple list? A clinician who recorded all psychiatric and non-psychiatric conditions, and also noted social and environmental problems, and made use of the WHO Disability Schedule, would end up recording much the same information as would have been recorded using DSM-IV, but in a different format. With the abandonment of the multiaxial system, however, there has been a clear shift in emphasis. The introduction to the DSM-IV gave the impression that a diligent clinician was expected to employ the multiaxial system. Nine pages explained the system, and a suggested form was included that might be used to record information on all axes (APA, 1994: 25–34). Now, although it is true that clinicians can still record psychosocial and environmental problems, and that a disability assessment scale is included in an appendix, one gets the impression that clinicians are not really expected to record this information. The instructions telling clinicians that they may do so are buried in a single sentence in dense text, and are so poorly copy-edited that it is unclear whether or not the WHO Disability Schedule is intended for everyday clinical use (APA, 2013: 16).

When revisions have been made to the actual sets of diagnostic criteria included in the DSM-5, it is easy to track the rationale. At this level, for each revision the responsible workgroups published papers explaining their proposals (either in print, or online on the DSM-5 revisions pages maintained by the APA).2 In contrast, it is much harder to track the rationale for changes to the DSM at the conceptual or structural level. At this level, explanations were seldom published.

The multiaxial system was originally introduced with DSM-III (APA, 1980). The basic system was maintained from DSM-III through to DSM-IV-TR (APA, 2000). When the DSM-III was published, the multiaxial system was widely praised (Frances and Cooper, 1981). It was seen as implementing a biopsychosocial model of disorder; the clinicians’ attention was systematically directed towards biological, psychological, social and environmental factors. This thinking was maintained into DSM-IV, where readers are told that ‘the multiaxial system promotes the application of the biopsychosocial model in clinical, educational and research settings’ (APA, 1994: 25)

During the development of the DSM-III and IV, the multiaxial approach played a key strategic role in addressing the concerns of those who feared that the DSM was an overly ‘medical’ or ‘biological’ diagnostic system. Those who wished to emphasize the importance of environmental or social factors could be assured that a full DSM diagnosis would take such factors into account (Spitzer, 2001: 357). Axis IV required that psychosocial and environmental problems be considered. Axis II (although mainly used for personality disorders and mental retardation codes) could be used for recording ‘personality traits’, and from DSM-III-R onwards could also be used to note ‘defence mechanisms’ (APA, 1987: 16; 1994: 27).

The multiaxial system also kept open the possibility that with time the DSM might come to include more psychodynamic, social or environmental information. Some thought that a new axis should be developed for defence mechanisms (Frances and Cooper, 1981; Skodol and Perry, 1997), others suggested an axis be added for relational functioning (Williams, 1997). These efforts met with some success, and the DSM-IV included these axes in an appendix entitled ‘Axes provided for further study’ (APA, 1994: 751–9).

The multiaxial system managed to partially accommodate psychodynamic, social and environmental concerns, but it did so in a way that simultaneously avoided threatening those who were happy to embrace a ‘medical’ or ‘biological’ approach to mental disorders. While recording information on axes IV and V was encouraged, it was generally considered optional (Skodol, 1997). Thus, the multiaxial system functioned to create liminal spaces in the DSM system. Axes IV and V were not outside the system, but not quite in it either.

There are probably a number of reasons for the abandonment of the multiaxial system. The split between axes I and II caused a number of difficulties: insurance companies often refused to reimburse for axis II disorders, and some hoped that doing away with axis II would make it easier for patients to secure coverage for therapy (Wakefield, 2013). There was also ongoing uncertainty about the placement of certain conditions, for example some argued that Borderline Personality Disorder should be moved to axis I (Gunderson, 2009).

Yet it would have been possible to merge axes I and II while retaining axes III, IV and V, and so the difficulties with the axis I/axis II split cannot fully account for the abandonment of multiaxial system. Probst (2014) argues that the removal of axis IV may both reflect and produce a reduced interest in social and environmental factors. With the loss of axis II, which allowed defence mechanisms to be recorded, psychodynamic factors have also been downplayed. A brief comment in a paper co-authored by Kupfer, chairman of the DSM-5 committee, indirectly supports the suggestion that the abandonment of the multiaxial system signals a shift away from a biopsychosocial towards a more narrowly medical approach to mental disorder. The paper notes that the multiaxial system was abandoned, ‘… largely due to its incompatibility with diagnostic systems in the rest of medicine’ (Kupfer, Kuhl and Regier, 2013). Given the history of the multiaxial system, it is reasonable to conclude that its abandonment goes along with a reduced emphasis on psychodynamic, social, relational and environmental approaches to mental disorder.

The definition of disorder

Since DSM-III, the DSM has included a definition of mental disorder in the introduction. In the DSM-IV definition, there is a conceptual link between disorder and harm (or at least the risk of harm):

each of the mental disorders is conceptualized as a clinically significant behavioral or psychological syndrome or pattern that occurs in a person and that is associated with present distress (a painful symptom) or disability (impairment in one or more important areas of functioning) or with significantly increased risk of suffering death, pain, disability, or an important loss of freedom … (APA, 1994: xxi)

In the DSM-5, however, there is a revised definition and the role of harm has been downgraded. The new definition states only: ‘Mental disorders are usually associated with significant distress or disability in social, occupational, or other important activities’ (APA 2013: 20, emphasis added). Following this definition, someone who meets diagnostic criteria but who is not harmed (or at least at risk of harm) by their condition can now be diagnosed with a mental disorder. Potentially, this change is very significant. Many people have ‘symptoms’ suggestive of, for example, schizophrenia or Asperger’s or Tourette’s, but are not harmed by their condition (or at risk of harm). Such people do not have a mental disorder under the DSM-IV definition, but could under the DSM-5 definition.

The idea that there is a conceptual link between mental disorders and harm (as per the DSM-IV) became the dominant view among mental health professionals following the 1970s debates over homosexuality (Bayer, 1981). Homosexuality might turn out to be caused by some sort of evolutionary dysfunction, but a consensus developed that, in so far as being gay is not harmful, homosexuality should not be considered a disorder. The DSM-IV definition of disorder acted to ensure that those who are different, but whose difference causes no harm, could not be given a diagnosis.

In the DSM-5 the definition of disorder was changed, but figuring out exactly how it came to be changed is not straightforward. There was no workgroup appointed to revise the definition of disorder, and no papers discussing or justifying the changes were published. The change to the introductory definition of disorder was made late on in the revision process and was little discussed. If we look at the drafts of the DSM-5 that were published online at various points as the edition was being put together,3 we can see that early drafts of the DSM-5 (from at least February to July 2010) included a definition of mental disorder that differed only slightly from the one included in DSM-IV (this proposed definition is detailed in Stein et al., 2010). The possibility of a radical revision was first discussed in the DSM-5 draft available in December 2011, which included two competing definitions of mental disorder, each developed by a different group. A note explained that a decision would be made later about which definition to adopt. While the first proposed definition continued to be that presented in the earlier draft (a slight revision of the DSM-IV definition), the second proposed definition was very different. This was developed by the Impairment and Disability Assessment Study Group, chaired by Jane Paulsen; in this definition all reference to distress and impairment had been removed.

The motivation of the Impairment and Disability Assessment Study Group was to bring the DSM definition of disorder in line with that implicit in the WHO’s ICD. Alongside the ICD, the WHO also publishes a distinct classification, The International Classification of Functioning, Disability and Health (2001), which supplies codes for all forms of disability. The ICD conceives of disorder and disability as being quite distinct. According to the social model of disability, impairment and disability should be conceived of separately; impairment refers to the biological difference (e.g. having no legs), and disability refers to problems in everyday living that are thought of as arising from the social response to the impairment (Oliver, 1996). Following this logic, in the case of someone who, for instance, hears voices but is not bothered by them and has a good life, the Impairment and Disability Assessment Study Group would say that the person has the disorder of schizophrenia (supposing that criteria for duration, etc., are met),4 but is not disabled.

The proposed changes to the definition of disorder attracted very little attention and were little discussed. Eventually, from June 2012 draft versions of the DSM-5 included something very close to the definition actually published in the DSM-5. In full, the new DSM-5 definition is:

A mental disorder is a syndrome characterized by clinically significant disturbance in an individual’s cognition, emotion regulation, or behavior that reflects a dysfunction in the psychological, biological, or developmental processes underlying mental functioning. Mental disorders are usually associated with significant distress or disability in social, occupational, or other important activities. An expectable or culturally approved response to a common stressor or loss, such as the death of a loved one, is not a mental disorder. Socially deviant behavior (e.g. political, religious, or sexual) and conflicts that are primarily between the individual and society are not mental disorders unless the deviance or conflict results from a dysfunction in the individual, as described above. (APA, 2013: 20)

The published definition appears to be an uncomfortable compromise between a definition like the DSM-IV definition (which considers harm conceptually tied to disorder) and the definition of the Impairment and Disability Assessment Study Group (which does not mention harm).

It remains the case in the DSM-5 that many of the individual sets of diagnostic criteria include a requirement that the particular disorder can only be diagnosed if symptoms cause problems. There is often a criterion requiring that the symptoms cause ‘clinically significant distress or impairment in social, occupational, or other important areas of functioning’, or some such. These criteria have mainly been inherited from DSM-IV. But the introduction to the DSM-5 makes it clear that these might be removed at some point in the future. The introduction explains:

There have been substantial efforts by the DSM-5 Task Force … to separate the concepts of mental disorder and disability (impairment in social, occupational, or other important areas of functioning) … However, in the absence of clear biological markers or clinically useful measurements for severity for many mental disorders, it has not been possible to completely separate normal and pathological symptom expressions contained in diagnostic criteria. … Therefore, a generic diagnostic criterion requiring distress or disability has been used to establish disorder thresholds, usually worded ‘the disturbance causes clinically significant distress or impairment in social, occupational, or important areas of functioning’. (APA, 2013: 21)

In DSM-5, the idea that certain disorders cause distress or impairment is included in diagnostic criteria only pending the development of some better means of drawing the line between the normal and the pathological. The DSM-5 thinks of references to harm as only being necessary for the time being, because science is as yet underdeveloped. Harm is no longer considered conceptually necessary for diagnosis.

Descriptive approach abandoned

The DSM-III set out to be a purely descriptive, atheoretical classification system (except in the case of conditions with established causes, such as the organic mental disorders). In the DSM-IV, the claim to be atheoretical was quietly dropped, but the structure and contents of the classification remained much as they had been in DSM-III. Now, with the DSM-5, the classification aims to be based on theoretical knowledge about aetiology.

The reasons that the DSM-III set out to adopt a descriptive, atheoretical stance have already been partly discussed. When the classification was being developed, in the mid- to late 1970s, there was much concern with improving the reliability of diagnosis. The adoption of ‘operational definitions’, which sought to set out the symptoms required for diagnosis in language that was as descriptive as possible, was seen as a way of dealing with the problem of reliability (e.g. see Kendell, 1975). In addition, at the time, mental health professionals in the USA worked under a range of different paradigms. In particular, psychoanalysis remained an important approach. Making the DSM-III atheoretical was presented as a means by which the classification could be rendered acceptable to those working in a range of different paradigms (APA, 1980: 7).

Following the publication of the DSM-III, the claim that the classification was atheoretical came in for much criticism. Commentators pointed out that the DSM-III was not, in fact, purely descriptive. Many of the diagnostic criteria require theory-laden inferences to be made, such as ‘identity disturbance’, which is a symptom of Borderline Personality Disorder (Cooper and Michels, 1981). Many DSM diagnoses contain exclusion clauses. Agoraphobia, for example, can only be diagnosed if the symptoms cannot be better explained by a major depressive episode or schizophrenia. These disorders are thought to be more ‘deep-rooted’ than that underlying agoraphobia, and so trump an agoraphobia diagnosis (Cooper, 2005: 78). A different line of criticism has taken a more conceptual path. Philosophers of science commonly hold that observation is theory-laden, and that thus no classification can be theory-free (Kuhn, 1970; Quine, 1960).

In the DSM-IV the introduction quietly dropped the claim to be atheoretical. Possibly this was under the influence of its chairman, Allen Frances, who had criticized the DSM-III’s claim to be atheoretical on the basis that observation is always theory-laden (Frances and Cooper, 1981: 1199; Frances et al., 1990). Still, the DSM-IV is very clearly a descendant of DSM-III. Many sets of diagnostic criteria included in the DSM-IV remain the same as in the DSM-III, and new criteria sets follow the same style as those that have been inherited.

The DSM-5 sought to take a different approach. As discussed in the first part of this article, the classification set out to be moulded by theories about aetiology. In particular, a body of work focused on the ‘metastructure’ of the DSM. The idea was that disorders would be grouped together in a way that provided information about their nature.

Early on in the revision process, there were indications that the structure of the classification might be overhauled. A subgroup of the DSM-5 Task Force worked on the ‘metastructure’, and a special issue of Psychological Medicine published their proposals. It was suggested that the DSM-5 might be reorganized into a number of clusters reflecting ‘aetiological risk factors’ (Andrews et al., 2009: 1999). Five clusters were proposed: (1) neurocognitive (grouping delirium, dementias, amnesic and other cognitive disorders); (2) neurodevelopmental (grouping mental retardation, learning, motor skills and communication disorders, and pervasive developmental disorders); (3) psychosis (grouping schizophrenia, bipolar disorder, schizotypal personality disorder); (4) emotional (a large grouping, including unipolar depression, anxiety, post-traumatic stress disorders, and avoidant personality disorder); and (5) externalizing (including substance-related disorders, antisocial personality disorder, and impulse–control disorders). Most disorders fell outside these clusters and formed a final cluster: ‘disorders not yet assigned’.

In the aforementioned special issue, following the papers setting out the Task Force proposal, a number of commentaries were published. These were uniformly negative in their assessment (First, 2009; Jablensky, 2009; Wittchen, Beesdo and Gloster, 2009). The proposal was criticized on the basis that it lacked adequate empirical support, too many conditions were unassigned, and some of the proposed groupings departed radically from clinical tradition (the personality disorders, for example, were to be split between distinct groups).

The idea of a radical restructuring was abandoned, but some of these ideas made it into the DSM-5, albeit in diluted form. As the introduction makes clear, the ordering of disorders is now intended to convey information about supposed aetiology:

The proposed organization of chapters of DSM-5, after the neurodevelopmental disorders, is based on groups of internalizing (emotional and somatic) disorders, externalizing disorders, neurocognitive disorders, and other disorders. It is hoped that this organization will encourage further study of underlying pathophysiological processes that give rise to diagnostic comorbidity and symptom heterogeneity. (APA, 2013: 13)

In actual implementation, the changes that have been made to the DSM on the basis of aetiological theories are extremely subtle, and I suspect they will be missed by most readers. There are no headings that make it clear that ‘internalizing’ and ‘externalizing’ conditions are being grouped together. Those aspects of the proposal that would have resulted in a classification that diverged sharply with tradition have been avoided; antisocial personality disorder, for example, is an ‘externalizing disorder’ but remains with the personality disorders, where it has always been. Those changes that have been made are modest; for example, the chapter on ‘Disruptive, impulse-control and conduct disorders’ has been moved to be next to ‘Substance-related and addictive disorders’, so that these ‘externalizing’ disorders are together. Antisocial personality disorder now appears in both the personality disorder section and under ‘Disruptive, impulse-control and conduct disorders’. Changes of this type are insignificant in themselves. But they do show that there has been a profound reconceptualization of the DSM. Once, the classification set out to be descriptive and atheoretical; now, it seeks to reflect theoretical knowledge.

The role of values

In the DSM-5 revision, for the first time the potential of the classification to produce harm (via facilitating stigma, over-treatment, and so on) was considered in an explicit and systematic way.

In contrast, when the DSM-IV was developed, although the harms that the system could cause were sometimes considered, this was only in a limited and piecemeal way.5 In the construction of the DSM-IV, the key rhetorical themes were empirical evidence and transparency. Guidelines for DSM-IV workgroups emphasized that proposals for change had to be supported by comprehensive and systematic reviews of empirical evidence (Frances, Widiger and Pincus, 1989; Widiger, Frances, Pincus, Davis and First, 1991). An accompanying four-volume Sourcebook (each volume of which runs to about 1000 pages) published the literature reviews that had been undertaken by the various workgroups to document the support for changes to the classification (APA, 1994: xx; Widiger et al., 1994, 1996, 1997, 1998). The guidelines for workgroups made no mention of the harms that revisions might cause.6

Those developing the DSM-IV considered their responsibilities to be limited. Allen Frances, chairman of the DSM-IV, wrote:

It is unclear to what extent the DSM-IV task force should be influenced by the effects its decision might have beyond the immediate clinical domains for which the system is designed. To ignore obvious detrimental effects would be irresponsible, but for DSM-IV to have credibility as a system of diagnosis its clinical uses and foundations in research must have priority. (Frances et al., 1990: 1442)

Conceptually, in this period the harms that diagnosis can cause were generally characterized as stemming from ‘misuse’. Talk of ‘misuse’ implies that harm might result when the DSM was used inappropriately. The idea seemed to be that various bad or irresponsible types – maybe the marketing departments of pharmaceutical companies, or overly zealous lawyers – might potentially exploit the letter of the diagnostic criteria of the DSM in ways that departed from its spirit. Those writing the DSM-IV sought to be careful in their choice of wording to avoid such possibilities.

Prior to publication of the DSM-IV, an Options Book was published, which contained details of the changes that were being proposed. The introduction explains: ‘The provision of an Options Book at this time will help ensure that we are receiving the widest possible input of data and opinion and that we are not missing inconsistencies, errors, or potential for misuse’ (Task Force on DSM-IV, 1991: A.1). The Options Book included a form for readers to note any problems they found in the text and to send it to the APA. Some readers did report errors and possible problems. I have visited the archives of the American Psychiatric Association and seen some of the materials related to development of the DSM-IV. The files do contain some returned forms, with notes of problems and suggested solutions. Note, however, that the Options Book accords identifying ‘potential for misuse’ a rather low priority. Readers are asked to report concerns about potential harms alongside copy-editing errors.

In contrast, when the DSM-5 was being developed, the potential for harm was systematically considered. Guidelines for revisions to DSM-5 asked those proposing new diagnoses to consider whether ‘… the harm that arises from the adoption of the proposed diagnosis exceed[s] the benefit that would accrue to affected individuals’ (Kendler et al., 2009: 6), and potential for harm was cited as a possible reason for keeping a diagnosis in the appendix rather than promoting it to the main classification (p. 9). Significantly, in these guidelines harm is seen as potentially resulting from proper use of the DSM and not just from ‘misuse’. The aim now was to try to ensure that the DSM was designed such that diagnoses based on the DSM on balance do more good than harm.

Workgroups developing proposals for the DSM-5 did seek to demonstrate that their proposed revisions would bring about more good than harm. The ‘do no harm’ criterion was referred to in many papers arguing for or against proposed revisions to the DSM-5 (e.g. Boelen and Prigerson, 2012; Huprich, 2012; Mataix-Cols et al., 2010; Selby, Bender, Gordon, Nock and Joiner, 2012; Woods, Walsh, Saksa and McGlashan, 2010). Benefits that were commonly expected from proposed revisions included the facilitation of appropriate treatment and other needed services, providing a category useful for future research, and improving clinical communication. Types of harm that were commonly considered included stigmatization and self-stigmatization, the inappropriate treatment of false-positives, legal and bureaucratic consequences, and the potential medicalization of normality.

In the DSM-5, the requirement that harm be considered was limited in scope. Considerations of harm were only explicitly required from those proposing to introduce new diagnoses, or to move diagnoses from the appendix to the main classification. Presumably, however, if considerations of harm should play a role in assessing these types of proposed revision, there is no principled reason why they should not also play a role in evaluating other types of proposed revision (such as proposals for name changes, for lumping or splitting categories, or for altering diagnostic criteria).

In the DSM-5, for the first time it was accepted that proper use of the DSM can produce harm, and this is something that those writing the DSM should systematically consider. The classification shifted from being one that presented itself as based foremost on empirical evidence to one that was developed on the basis of both empirical evidence and consideration of values.

Summary

We have seen how the conceptual underpinnings of the DSM-5 differ from those of the DSM-IV in important ways. Table 1 summarizes the main differences (various nuances, discussed earlier, have been ignored in the interests of clarity).

Table 1. Conceptual differences between DSM-IV and DSM-5.

Looseness

In the first section, I argued that the sets of diagnostic criteria included in the DSM-5 remain much as they were in DSM-IV – at this level, the system has become locked-in. In the second section, I argued that at the conceptual level there are important changes between the DSM-IV and DSM-5. This raises two questions: How can the DSM be locked-in at the level of diagnostic criteria, but flexible at the conceptual level? What is the relationship between the conceptual framework of the DSM and the sets of diagnostic criteria it contains, such that radical changes can occur at the conceptual level but have little impact on the actual diagnostic criteria?

Understanding how the DSM is used and written is key to addressing these two questions. Fundamentally, the reason it has become hard to revise the diagnostic criteria included in the DSM is that diverse users employ these criteria. If diagnoses are added or removed from the classification, or diagnostic criteria are revised, people notice and care. The various users have different requirements, and it is hard to revise the classification in such a way that these requirements continue to be met.

Plausibly, most users read only those portions of the DSM that they use directly.7 The introduction in particular will probably pass unread, and this, of course, is the part of the book that outlines the conceptual framework. As a consequence, most readers will not know or care how the DSM defines disorder, or whether concerns about harm are considered, or whether the classification seeks to be based on theoretical knowledge.

The conceptual framework of the DSM is of little interest to most users, and it has also been of little interest to those who develop the classification. The various sections of the classification had workgroups assigned to them. These workgroups published and presented their proposals, which were intensely scrutinized and discussed. Multiple conferences and journal issues were devoted to discussion of proposed revisions for autism-related conditions, or ADHD, for example. In contrast, few committees were appointed to consider issues connected with the conceptual framework. In her article on the demise of Axis IV, Barbara Probst (2014: 128) notes that ‘The paucity of APA resources assigned to the issue both reflected and determined its perceived unimportance; something that did not even merit its own Work Group could not inspire dialogue’. Where committees did work on conceptual issues, such issues were accorded low priority. As mentioned previously, proposals to revise the definition of disorder were made available late and little discussed. By the time the definition of mental disorder was decided, almost all the sets of diagnostic criteria were already close to their final form. I suggest that those who have revised the conceptual framework of the DSM have been free to make revisions partly because very few people have known, or cared, about the revisions.

In discussing the question of how the conceptual framework of the DSM can shift without this affecting its contents, we must consider each conceptual change in turn. Let us first consider the loss of the multiaxial system. I have argued that the multiaxial system of the DSM-IV can be understood as implementing a biopsychosocial model of disorder. The DSM-5, in contrast, has abandoned the multiaxial system and has a more narrowly medical model of disorder. This is an important conceptual change, but it is not one that could be expected to affect the sets of diagnostic criteria. In DSM-IV the diagnostic criteria were used for Axes I and Axes II. In effect they coded the ‘medical’ bit of a multiaxial (biopsychosocial) DSM-IV diagnosis. This information continues to be routinely collected, and so the sets of diagnostic criteria remain unaffected.

The DSM-III sought to be a descriptive, atheoretical classification. The DSM-IV dropped the philosophical claims, but continued in the same vein. In contrast, DSM-5 set out to reflect theories about aetiology. This is an important conceptual shift. It has had little impact on the sets of diagnostic criteria, however, partly because there is not yet sufficient theoretical knowledge available for many changes to have been motivated.

In the DSM-5, for the first time the potential for the DSM to produce harm was systematically considered. A committee was put in place to make sure that all proposals for adding or deleting conditions included a consideration of the potential harms that might ensue (Yager and McIntyre, 2014). However, this has only had a modest impact on the sets of diagnostic criteria included because the guidelines only required that harm be considered when disorders were suggested for inclusion or deletion. Relatively few such changes were suggested, and so the altered stance has had relatively little effect.

Finally, the DSM-5 uses a definition of mental disorder that differs from that in earlier editions. Disorder is no longer conceptually tied to harm. Potentially, this is an important revision. If taken seriously, it would mean that people who tic, or hear voices, or avoid social interaction, but who lead flourishing lives, could be diagnosed with Tourette’s, or schizophrenia, or ASD. In actual fact, there has been no such impact. The definition of disorder was produced too late to feed into other committee processes.

For the work of various DSM committees to fail to connect is not that unusual.8 The DSM is constructed by multiple committees, which work semi-independently. A Task Force provides some oversight, but the degree of independence is sufficient for distinct groups to end up with divergent ideas such that, when their work is patched together to form the overall text, the joins show. Either the overall text becomes inconsistent, or last-minute compromises are forged. Committee processes mean that the DSM is not always a conceptually coherent classification.

Notes

Papers in the DSM-III and DSM-IV archives reveal much concern with press coverage.

These pages have now been removed but can still be located using the internet archive, The Wayback Machine, by searching for http://dsm5.org

These pages can still be located using the internet archive, The Wayback Machine, by searching for http://dsm5.org

In contrast to the DSM-IV position, the ICD has always allowed for schizophrenia to be diagnosed in the absence of disability or distress.

Sometimes the DSM-IV came to be shaped by concerns about possible harms. For example, the Child Disorders workgroup worried about the potential for ADHD to be over-diagnosed, and were concerned that ‘subsuming oppositional defiant disorder under conduct disorder may be unnecessarily stigmatizing’; Shaffer, Widiger and Pincus, 1998; for quote see p. 968.

A later chapter by Pincus (2012) cites the guidelines for revision (Frances et al., 1989) and says they included a step ‘Consider and delineate risks and benefits of each option’, which makes it sound as though harms might have been systematically considered. But in Frances et al. (1989) there is no mention of any such step.

Few studies have been conducted that show how the DSM is employed in practice. Those that have been done are mainly fairly old, but suggest that users of the DSM tend to consult it on a need-to-know basis and tend not to read it very carefully. For a review of such studies see Mullins-Sweatt et al., 2016.

Consider, for example, the DSM-5 stance on Bipolar Disorder in children. A proposal for a distinct code for Paediatric bipolar disorder was rejected. Instead, Disruptive mood dysregulation disorder was introduced in the Depressive Disorders chapter. But elsewhere in the DSM, in the bipolar chapter, the text advises that ‘Other specified bipolar and related disorder’ might be used to diagnose children with bipolar-like phenomena; APA, 2013: 123.

References

American Psychiatric Association [APA] (1980) Diagnostic and Statistical Manual of Mental Disorders, 3rd edn. Washington, DC: American Psychiatric Association.

American Psychiatric Association [APA] (1987) Diagnostic and Statistical Manual of Mental Disorders, 3rd edn rev. Washington, DC: American Psychiatric Association.

American Psychiatric Association [APA] (1994) Diagnostic and Statistical Manual of Mental Disorders, 4th edn. Washington, DC: American Psychiatric Association.

American Psychiatric Association [APA] (2000) Diagnostic and Statistical Manual of Mental, 4th edn text rev. Washington, DC: American Psychiatric Association.

American Psychiatric Association [APA] (2013) Diagnostic and Statistical Manual of Mental Disorders, 5th edn. Washington, DC: American Psychiatric Association.

American Psychiatric Association [APA] Treasurer (2012) Treasurer’s Report 2012. Accessed (2 Sep. 2016) at: https://docs.google.com/file/d/0BzWdENl1wkVSYk5aXzRZelFYUjA/edit?pli=1 (Posted on dx revision watch blog, May 29 2012, by MEAgenda: https://dxrevisionwatch.com)

Andrews, G, Goldberg, DP, Krueger, RF, Carpenter, WT, Hyman, SE, Sachdev, P, Pine, DS (2009) Exploring the feasibility of a meta-structure for DSM-V and ICD-11: could it improve utility and validity? Psychological Medicine 39: 1993–2000.

Anonymous (2013) Insurance implications of DSM-5. Psychiatric News 48(9): 1.

Aragona, M (2015) Rethinking received views on the history of psychiatric nosology: minor shifts, major continuities. In: Zachar, P, St.Stoyanov, D, Aragona, M, Jablensky, A (eds) Alternative Perspectives on Psychiatric Validation. Oxford: Oxford University Press, 27–46.

Bayer, R (1981) Homosexuality and American Psychiatry. New York: Basic Books.

Boelen, P, Prigerson, H (2012) Prolonged grief disorder as a new diagnostic category in DSM-5. In: Stroebe, M, Schut, H, van den Bout, J (eds) Complicated Grief: Scientific Foundations for Health Care Professionals. London: Routledge, 85–98.

Connecticut General Assembly (2013) Amendment to an Act concerning health insurance coverage for autism spectrum disorders. SB0102906154SDO; accessed (2 Sep. 2016) at: https://www.cga.ct.gov/2013/amd/S/2013SB-01029-R00SA-AMD.htm

Cooper, A, Michels, R (1981) Diagnostic and statistical manual of mental disorders. American Journal of Psychiatry 138: 128–129.

Cooper, R (2005) Classifying Madness. Dordrecht: Springer.

David, P (1985) Clio and the economics of QWERTY. American Economic Review 75: 332–337.

First, M (2009) Reorganizing the diagnostic groupings in DSM-V and ICD-11: a cost/benefit analysis. Psychological Medicine 39: 2091–2097.

Frances, A, Cooper, A (1981) Descriptive and dynamic psychiatry: a perspective on DSM-III. American Journal of Psychiatry 138: 1198–1202.

Frances, A, Pincus, H, Widiger, T, Dabis, W, First, M (1990) DSM-IV: work in progress. American Journal of Psychiatry 147: 1439–1448.

Frances, A, Widiger, T, Pincus, H (1989) The development of DSM-IV. Archives of General Psychiatry 46: 373–375.

Frazer, P, Westhuis, D, Daley, J, Phillips, I (2009) How clinical social workers are using the DSM-IV: a national study. Social Work in Mental Health 7: 325–339.

Greenberg, G (2013) The Book of Woe: The DSM and the Unmaking of Psychiatry. New York: Blue Rider.

Gunderson, J (2009) Borderline personality disorder: ontogeny of a diagnosis. American Journal of Psychiatry 166: 530–539.

Happé, F (2011) Criteria, categories, and continua: autism and related disorders in DSM-5. Journal of the American Academy of Child and Adolescent Psychiatry 50: 540–542.

Huprich, S (2012) Considering the evidence and making the most empirically informed decision about depressive personality disorder in DSM-5. Personality Disorders: Theory, Research, and Treatment 3: 470–482.

Illinois General Assembly (2013) Amendment to the the Illinois Insurance Code SB0679; accessed (2 Sep. 2016) at: http://www.ilga.gov/legislation/fulltext.asp?DocName=&SessionId=84&GA=97&DocTypeId=SB&DocNum=0679&GAID=11&LegID=55508&SpecSess=&Session

Jablensky, A (2009) A meta-commentary on the proposal for a meta-structure for DSM-V and ICD-11. Psychological Medicine 39: 2099–2103.

Kendell, R (1975) The Role of Diagnosis in Psychiatry. Oxford: Blackwell Scientific Publications.

Kendler, K, Kupfer, D, Narrow, W, Phillips, K, Fawcett, J (2009) Guidelines for making changes to DSM-V; accessed (2 Sep. 2016) at: http://www.dsm5.org/ProgressReports/Documents/Guidelines-for-Making-Changes-to-DSM_1.pdf

Kuhn, T (1970) The Structure of Scientific Revolutions, 2nd edn. Chicago, IL: University of Chicago Press.

Kupfer, D, First, M, Regier, D (eds) (2002) A Research Agenda for DSM-V. Washington, DC: American Psychiatric Association.

Kupfer, D, Kuhl, E, Regier, D (2013) DSM-5 – the future arrived. JAMA 309: 1691–1692.

Levine, M (2013) Pitt prof oversees psychiatric guide revision. University of Pittsburgh University Times 45(9) 10 Jan.; accessed (2 Sep. 2016) at: http://www.utimes.pitt.edu/?p=23893

Mataix-Cols, D. (2010) Hoarding disorder: a new diagnosis for DSM-V? Depression and Anxiety 27: 556–572.

Matson, J, Belva, B, Horovitz, M, Kozlowski, A, Bamburg, J (2012) Comparing symptoms of autism spectrum disorders in a developmentally disabled adult population using the current

DSM - IV - TR diagnostic criteria and the proposed DSM - 5 diagnostic criteria. Journal of Developmental and Physical Disabilities 24: 403–414.

Mattila, M (2011) Autism spectrum disorders according to DSM-IV-TR and comparison with DSM-5 draft criteria: an epidemiological study. Journal of the American Academy of Child and Adolescent Psychiatry 50: 583-592.e11.

Mullins-Sweatt, SN, Lengel, GJ, DeShong, HL (2016) The importance of considering clinical utility in the construction of a diagnostic manual. Annual Review of Clinical Psychology 12: 133–155.

Oliver, M (1996) Defining impairment and disability: issues at stake. In: Barnes, C, Mercer, G (eds) Exploring the Divide. Leeds: The Disability Press, 29–54.

Paris, J (2015) The Intelligent Clinician’s Guide to the DSM-5, 2nd edn. Oxford: Oxford University Press.

Pincus, H (2012) DSM-IV: context, concepts and controversies. In: Kendler, K, Parnas, J (eds) Philosophical Issues in Psychiatry. II: Nosology. Oxford: Oxford University Press, 145–160.

Poland, J (2015) DSM-5 and research concerning mental illness. In: Demazeux, S, Singy, P (eds) The DSM-5 in Perspective. Oxford: Oxford University Press, 25–42.

Probst, B (2014) The life and death of axis IV: caught in the quest for a theory of mental disorder. Research on Social Work Practice 24: 123–131.

Quine, W (1960) Word and Object. Cambridge, MA: The MIT Press

Regier, D, Kuhl, E, Kupfer, D (2013) The DSM-5: classification and criteria changes. World Psychiatry 12: 92–98; doi: 10.1002/wps.20050

Selby, E, Bender, T, Gordon, K, Nock, M, Joiner, T (2012) Non-suicidal self-injury (NSSI) disorder: a preliminary study. Personality Disorders: Theory, Research, and Treatment 3: 167–175.

Shaffer, D, Widiger, T, Pincus, HA (1998) DSM-IV child disorders, Part II: Final overview. In: Widiger, T. (eds) DSM-IV Sourcebook, Vol. 4. Washington, DC: APA, 963–977.

Skodol, A (1997) Axis IV. In: Widiger, TA. (eds) DSM-IV Sourcebook, Vol. 3. Washington , DC: APA, 409–422.

Skodol, A. (2011) Proposed changes in personality and personality disorder assessment and diagnosis for DSM-5. Part I: Description and rationale. Personality Disorders: Theory, Research, and Treatment 2: 4–22.

Skodol, A, Morey, L, Bender, D, Oldham, J (2013) The ironic fate of the personality disorders in DSM-5. Personality Disorders: Theory, Research and Treatment 4: 342–349.

Skodol, A, Perry, J (1997) Defense mechanisms. In: Widiger, TA. (eds) DSM-IV Sourcebook, Vol. 3. Washington, DC: APA, 503–518.

Spitzer, R (2001) Values and assumptions in the development of DSM-III and DSM-III-R: an insider’s perspective and a belated response to Sadler, Hulgus, and Agich’s ‘On values in recent American psychiatric classification’. Journal of Nervous and Mental Disease 189: 351–359.

Stein, D, Phillips, K, Bolton, D, Fulford, K, Sadler, J, Kendler, K (2010) What is a mental/psychiatric disorder? From DSM-IV to DSM-V. Psychological Medicine 40: 1759–1765.

Task Force on DSM-IV (1991) DSM-IV Options Book. 9.1.91. Washington, DC: American Psychiatric Association.

Wakefield, J (2013) DSM-5: an overview of changes and controversies. Clinical Social Work Journal 41: 139–154.

Widiger, T, Frances, A, Pincus, H, Davis, W, First, M (1991) Toward an empirical classification for the DSM-IV. Journal of Abnormal Psychology 100: 280–288.

Widiger, TA. (eds) (1994) DSM-IV Sourcebook, Vol. 1. Washington, DC: APA.

Widiger, TA (eds) (1996) DSM-IV Sourcebook, Vol. 2. Washington, DC: APA.

Widiger, TA. (eds) (1997) DSM-IV Sourcebook, Vol. 3. Washington, DC: APA.

Widiger, TA. (eds) (1998) DSM-IV Sourcebook, Vol. 4. Washington, DC: APA.

Williams, J (1997) The DSM–IV multiaxial system. In: Widiger, TA. (eds) DSM-IV Sourcebook, Vol. 3. Washington, DC: APA, 393–400.

Google Scholar

Wittchen, H, Beesdo, K, Gloster, A (2009) A new meta-structure of mental disorders: a helpful step into the future or a harmful step back to the past? Psychological Medicine 39: 2083–2089.

Woods, S, Walsh, B, Saksa, J, McGlashan, T (2010) The case for including attenuated psychotic symptoms syndrome in DSM-5 as a psychosis risk syndrome. Schizophrenia Research 123: 199–207.

World Health Organization [WHO] (1992) The ICD-10 Classification of Mental and Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines. Geneva: World Health Organization.

World Health Organization [WHO] (2001) International Classification of Functioning, Disability and Health (ICF). Geneva: World Health Organization.