Book Review: Diagnostic and statistical manual of mental disorders

Basic Organization

The organization and structure of the manual has changed considerably. In DSM Fourth Edition Text Revision (DSM-IV-TR; American Psychiatric Association [APA], 2000), childhood disorders came first (within the portion of the table of contents presenting categories of disorder and their respective symptoms). After that, however, other categories of disorder did not seem to be presented in any specific order. In DSM-5, changes in the order of presentation were made, designed to reflect “a lifespan approach, with disorders more frequently diagnosed in childhood…at the beginning and disorders more applicable to older adulthood…at the end of the manual” (p. xliii). In effect, however, the manner in which this plan has been carried out remains less than fully satisfying. Many categories of disorders that were formerly included within the DSM-IV-TR chapter on “Disorders usually first diagnosed in infancy, childhood, and adolescence” have now been lifted out and given their own chapters, such as “Feeding and eating disorders,” “Elimination disorders,” and “Disruptive and impulse control disorders.” If a life span order of presentation were the goal, one might expect these new chapters to follow more closely on the heels of the newly retitled chapter on childhood disorders, now called “Neurodevelopmental disorders,” rather than 8 or more chapters later.

The concept of a life span order of presentation of categories seems to break down further, the deeper one gets into the details of which disorders are now included in which chapters. Many conditions formerly construed as relating primarily to childhood and adolescence have now been removed from the “neurodevelopmental” category, and embedded within other sections more typically related to disorders of adulthood. For instance, reactive attachment disorder, formerly defined as specific to infancy and early childhood, is now included within the chapter on “Trauma and stressor-related disorders,” where most of its categorical “neighbors” have traditionally been thought of as adult conditions. Similarly, selective mutism, also formerly a childhood disorder, may now be found within the chapter on “Anxiety disorders,” also traditionally an “adult” category for the most part. Other examples may be found throughout the manual. The reasoning behind these changes seems based on a hope of encouraging clinicians to be increasingly cognizant of “how disorder presentations may change across the lifespan” (p. xlii), yet, to this reader at least, these changes do not seem to be fully justified. The bottom line for users will probably be a need to consult the index (pp. 917–947) just as frequently as the table of contents when trying to locate specific diagnostic categories they are interested in reviewing, at least until they have become more fully familiar with DSM-5 and its new organization. Use of the table of contents may steer readers wrong, since particular categories of disorder are not where they used to be, nor where one might expect.

Following the chapters presenting disorders and symptomatology, toward the back of the manual, is a new section on “Emerging Measures and Models” that is potentially very helpful. A chapter on “Assessment measures” is located here, including a “Self-Rated DSM-5 Level 1 Cross-Cutting Symptom Measure, Adult” (p. 738); a “Parent-Guardian-Rated Cross-Cutting Symptom Measure” for children (p. 740); and a “World Health Organization Disability Assessment Schedule” (p. 743). The applied usefulness of these resources could be considerable, for instance, as screening tools in integrated primary care settings, offering merged health/behavioral health services in one location. Also in this section is a chapter on cultural formulation, including interview schedules designed for use with a culturally diverse clientele. These are new and potentially helpful and relevant in contemporary practice.

Removal of Multiaxial System

The five-axis system of assessment has been removed from DSM-5, and the fifth axis, global assessment of functioning (GAF), is no longer in use. According to the text, “The approach of separately noting diagnosis from psychosocial and contextual factors is…consistent with established WHO [World Health Organization and ICD [International Classification of Diseases] guidance to consider the individual’s functional status separately from his or her diagnoses and symptom states” (p. 16), which offers a rationale for this difference. In other words, no distinction is now made between diagnoses on the former Axes I and II, and data formerly noted on Axes III–V can now be captured elsewhere in the clinical record. If other clinicians and academics are of a similar mind, the former multiaxial system will not be missed, having seemed too generic and/or imprecise to be worth using.

Categorical Versus Dimensional Conceptualization of Disorders

DSM-IV-TR utilized “a categorical classification that divides mental disorders into types based on criteria sets with defining features” (p. xxxi). With DSM-5, there is movement away from this categorical conceptualization of disorders toward a more dimensional viewpoint “meant to stimulate new clinical perspectives and to encourage researchers to identify the psychological and physiological cross-cutting factors that are not bound by strict categorical designations” (p. 10). Additional articulations of this shifting of perspective are seen in a number of places throughout the new manual, as reflected, for instance, in the following passage:

The influence of this new, dimensional perspective contributes to several of the more dramatic and sweeping changes throughout DSM-5, particularly within the revised criteria for autism spectrum disorders, substance use disorders, and neurocognitive (formerly cognitive) disorders. Additional discussion about each of these newly revised categories follows; meanwhile, however, the question arises, does this conceptual evolution (from the categorical to the dimensional) actually improve DSM, from Edition IV to 5? For the most part, it does. In many instances, these changes do allow more latitude for relativity and clinical judgment and for improved congruence between diagnostic criteria and perceived clinical realities. What this conceptual evolution does not do is to make diagnostic criteria generally more “concise” (per the above passage). Instead, many disorders now have additional criteria to be met, while others now include several sets of criteria that must be reviewed, according to a sequential progression of considerations, before the complete, final diagnosis is made. New criteria for the “neurocognitive disorders” provide an example of this, about which more will be said subsequently. However, some of these new, stepwise diagnoses are rather complicated, and it seems predictable that some of these are now going to be slower and more cumbersome to formulate. If performed as intended, these new, layered diagnostic formulations show promise of the hoped-for increases in diagnostic specificity. However, that could be a big if, were clinicians tempted to take shortcuts, out of a belief that greater pinpoint diagnostic specificity may not always be worth the time and mental energy spent. If the pinpoint accuracy of a diagnosis does not lead to equally significant, case-specific implications for treatment planning, then clinicians may lose interest in the details, given that they sometimes feel that gold standard diagnostic specificity is more important to researchers and basic scientists than to themselves in their therapeutic work.

Removal of Not Otherwise Specified Diagnoses

A deliberate decision was made during revisions leading up to DSM-5 to remove all “not otherwise specified” (NOS) diagnostic categories from the new edition. The APA has recognized that the NOS categories have long been over used, and the hope seems to be that updated criteria will yield greater diagnostic accuracy in so many instances that the former NOS categories will no longer be needed. However, these have been replaced by two new potential catch-all categories: “other specified” (i.e., other specified schizophrenia spectrum disorder) and “unspecified” (i.e., unspecified schizophrenia spectrum disorder). One is doubtful that the average manual user will understand how these categories are particularly different from the previous NOS categories, other than the name change.

Bipolar and Related Disorders

Bipolar disorders are no longer subsumed under “mood disorders”; they now have their own chapter. As before, for bipolar I, criteria must be met for a major depressive episode at some point in the patient’s history. Now, however, these criteria are embedded within the rest of the criteria for the bipolar disorder, freeing users from the need to skip to another section to find them, a user-friendly upgrade.

There is a new cautionary note included among the criteria for bipolar I and II to the effect that “caution is indicated so that one or two symptoms (particularly increased irritability, edginess, or agitation…are not taken as sufficient for a diagnosis of a hypomanic episode, nor necessarily indicative of a bipolar diathesis” (pp. 125, 133). The insertion of this text appears to be a deliberate effort to curtail a possible tendency toward relaxation of diagnostic standards for bipolar disorder, leading to a potential increase in reported cases of uncertain validity. It resonates as a responsible and forward-thinking reminder for users to exercise defensible clinical judgment and to avoid trendiness.

Depressive Disorders

A new diagnosis has been added to this chapter, “disruptive mood dysregulation disorder,” which is described as follows: “The core feature of disruptive mood dysregulation disorder is chronic, severe persistent irritability” (p. 156), with age of onset being less than 10 years. This diagnosis is intended to be applied to children with frequent temper outbursts and episodes of behavioral dysregulation. This code could be of interest to those who contend that there is a worrisome contemporary trend toward overdiagnosis of pediatric bipolar disorder; perhaps this diagnosis would be a better fit in some situations.

Other revisions in criteria for mood disorders have been designed to address their expression or presentation among children and youth, a significant step forward.

Obsessive-Compulsive and Related Disorders

“Hoarding disorder” is new in DSM-5. Criteria are explicit and seem to capture clinical (and community) experience of this phenomenon very well. One wonders how this condition escaped becoming “diagnosable” as long as it did, although its inclusion now is no doubt due to greatly increased media attention to the condition in recent years.

Trauma- and Stressor-Related Disorders

Important changes have been made to criteria for posttraumatic stress disorder (PTSD). One such change is the separation of criteria for children and adults, which adds clarity. However, another significant change that is more difficult to embrace is seen in the softening of Criterion A, where the former requirement that “the person’s response (to a traumatic stressor) involved intense fear, helplessness, or horror” (APA, 2000, p. 467) has now been dropped. This criterion has been broadened further to the point that the experience of “learning that the traumatic events(s) occurred to a close family member or close friend” (APA, 2013, p. 271) is now sufficient. According to this criterion, a military wife who receives a home visit or letter letting her know that her spouse was killed in a combat theater could presumably be as susceptible to the development of PTSD as a soldier who was wounded at her husband’s side and was witness to his death. This relaxation of the diagnostic standard may meet stiff resistance in certain settings, such as the Veteran’s Administration, which has a history of requiring veterans seeking disability benefits for PTSD to be able to document the intensity of the traumatic stressor alleged to be causally related to the symptoms. This softening of the causal/experiential criterion for PTSD also seems counterintuitive and inconsistent with the notion of a searing, lightening-like spike of limbic arousal thought to be produced by a traumatic stressor. Contemporary brain science suggests that experiences such as these often predispose victims to many of their subsequent invasive symptoms, such flashbacks and episodes of reexperiencing. Is the diagnosis of PTSD heading in a direction that could blur the boundaries between experiences that are sad, or even tragic, yet within expectable range of human experience, as compared with those that are truly “traumatic”?

Neurodevelopmental Disorders

The chapter on “Neurodevelopmental disorders” begins with “intellectual developmental disorder” (also known as mental retardation). This diagnosis is now freed from its formerly inexplicable consignment to Axis II, in DSM-IV-TR. Particular points of IQ are no longer seen among diagnostic criteria. This is the first of the disorders in DSM-5 intended to be diagnosed in a more dimensional, less categorical manner than was formerly seen. Here, a continuum of severity is outlined to be determined by indicators seen across three separate domains: conceptual ability; social skills; and the practical domain (life/adaptive skills). This continuum is presented in a separate table that is to be consulted in order to specify severity, after a determination that other basic criteria have been met. Throughout, the term intellectual disability has replaced mental retardation, and all these changes allow diagnosis to be made in a notably less arbitrary and formulaic (and more humanistic) way than before.

Major changes are also seen in the autism spectrum disorders, now also conceptualized in a more dimensional way, again along a severity continuum. To arrive at a diagnosis, a determination is first made of each of the primary diagnostic criteria: (a) deficits in social communication and (b) restrictive, repetitive patterns of behavior and interest. Next, severity is assessed (using a separate table), on a scale from 1 to 3, depending on the level of support that the patient is expected to need in a given case situation.

A new category of disorder is introduced in this chapter, “social (pragmatic) communication disorder,” which includes criteria for social, communication, and interpersonal deficits but without the need to meet criteria for restrictive, repetitive patterns of behavior and interest. Also, the former Asperger’s disorder has been eliminated, and a note following the listing of criteria for autism spectrum disorder directs clinicians to reassign those with “a well-established DSM-IV pattern…of Asperger’s Disorder” (p. 51) to the new diagnosis of autism spectrum disorder, with the appropriate level-of-severity specifier. This note goes on to caution clinicians that “individuals who have marked deficits in social communication, but whose symptoms do not otherwise meet criteria for autism spectrum disorder, should be evaluated for social (pragmatic) communication disorder” (p. 51). This is one of a number of notations throughout DSM-5, suggesting the possibility of a conscious effort on the part of the Task Force to delimit a trend toward overutilization of a particular diagnosis; in this case, that of Asperger’s disorder. These changes appear to be carefully and thoughtfully made and seem timely and welcome.

Substance-Related and Addictive Disorders

Movement toward a more dimensional approach is also evident in this chapter. In DSM-IV-TR, a distinction was made between patterns of “abuse” and “dependence” in patients with problematic substance involvement. In DSM-5, the distinction between abuse and dependence has been eliminated to be replaced by a single category of “substance use disorder,” further qualified by a continuum of severity. This change in conceptualization is helpful, because it frees clinicians from having to make (and possibly defend) a judgment call as to whether a given patient merely “abuses” his or her substance(s) of choice; or whether she or he is actually “hooked” (“dependent”). Since this distinction has often been lost in an unknowable gray area, it may be a relief to clinicians not to have to make this “call.”

New to the chapter on Substance-Related and Addictive Disorders is “gambling disorder.” In DSM-IV, there was a disorder called “pathological gambling,” but it was placed among “impulse control disorders.” The symptoms were similar, but the conceptualization of problematic gambling as an addictive condition seems more authentic and more likely to suggest appropriate clinical treatment considerations. Also new is “Internet gaming disorder,” and although this appears toward the back of the manual in the section on “Conditions for Further Study,” its appearance is an interesting sign of the times.

Dissociative Disorders

The goal of increased diagnostic specificity has not been achieved in the case of “dissociative identity disorder (DID).” Rather, criteria have been broadened and made more diffuse. The former criteria that there should be “two or more distinct identities or personality states (each with its own relatively enduring pattern of perceiving, relating to, and thinking about the environment and the self)” and that “at least two of the identities or personality states recurrently take control of the person’s behavior” (APA, 2000, p. 529) have been dropped. The new Criterion A reads, “Disruption of identity characterized by two or more distinct personality states, which may be described in some cultures as an experience of possession” (APA, 2013, p. 292). It seems possible that many clinicians will find this ambiguous description difficult to interpret, particularly in light of Criterion D that goes on to say “The disturbance is not a normal part of a broadly accepted cultural or religious practice.” These differences in conceptualization do seem likely to steer the controversies surrounding DID in new directions, yet at the same time appear unlikely to resolve them, any time soon.

Neurocognitive Disorders

The dimensional approach has been thoroughly incorporated into the new conceptualization of neurocognitive disorders (formerly “cognitive” or “brain” disorders) in DSM-5. Also in this chapter, layered or multistep diagnostic procedures are the new norm. The term dementia is gone to be replaced by neurocognitive disorder or impairment across an array of capacities (complex attention; executive function, learning and memory, language, perceptual-motor, etc.). To document a neurocognitive disorder, manual users consult tables offering a series of observational and quantitative assessment strategies for use in determining a patient’s level of decline across these various capacities. The next diagnostic consideration is to judge whether the patient’s neurocognitive disorder is “mild” or “major,” which is determined by the level of functional impairment, measured and/or observed, in the relevant cognitive domains. The next step is to judge the most likely etiological subtype (Alzheimer’s disease, Lewy body disease, HIV infection, Prion disease, etc.). The next step is to decide whether the chosen etiological subtype is “probable” or “possible.” Using this assessment procedure, the former Dementia of the Alzheimer’s Type becomes “Neurocognitive Disorder (either major or mild) due to Alzheimer’s Disease (either probable or possible).”

Although this dimensional, stepwise approach to making these diagnoses is doubtless more precise, it also involves a lengthier and more cumbersome procedure than before. This may also be one of those areas in which pinpoint specificity of diagnosis is not guaranteed to yield proportionately significant implications for treatment planning.

Personality Disorders

The treatment of personality disorders in DSM-5 has been a highly controversial topic. This chapter was due to undergo major revisions, also along the lines of movement from a categorical to a more dimensional perspective. The Task Force believed that diagnostic categories for personality disorders were especially vulnerable to criticism on the basis of being too rigidly defined and that the NOS catch-all code was especially prone to overutilization with personality disorders. In other words, someone with a broad antisocial streak in his or her character might also be seen to have one or more narcissistic traits, while someone with heavy dependency needs might also possess a number of avoidant traits. All too often, this “mixed bag” presentation could only be captured using the NOS descriptor and hence, only in a nonspecific and incomplete manner. The new dimensional approach to these diagnoses was intended to resolve this problem with the validity and the utility of the personality disorder diagnoses.

When public commentary was invited, however, the proposed revisions drew especially heavy criticism, and in the end, the Task Force decided to put the issue aside for future consideration. As a result, criteria for personality disorders are among the least impacted by the process of revision and are presently included in the manual in terms and parameters that are little changed from the traditional.

This public outcry may have been due in part to difficulties fully articulating the intent and nature of the proposed revisions to interested parties during the public commentary interval. Among others, I looked into the issue at the time but found the description of proposed changes dense, cryptic, and hard to follow. It talked about evaluating personality function along continua of four dimensions: identity; self-direction; empathy; and intimacy. Additionally, aberrant or pathological trait domains of personality were to be evaluated along another set of continua: negative affectivity (vs. emotional stability); detachment (vs. extroversion); antagonism (vs. agreeableness); disinhibition (vs. conscientiousness); and psychoticism (vs. lucidity). However, insufficient detail was included in the summary description of this new conceptualization to allow readers to appreciate it in full. My reaction was one of confusion melded with indignation, as it seemed that these were too few descriptors and/or dimensions there to fully capture the richness and complexity of human personalities. However, an expanded explanation of the entire plan is included in a separate section toward the back of DSM-5, in Section III, “Emerging Measures and Models.” Here, enough additional detail may be found to make the intent and operational diagnostic structure of these revisions more apparent, and there is more to like about these changes than I was prepared to find, before getting into them in more detail. Space limits prevent a full defense of this assertion, but readers are encouraged to get into these details and try the new standards out on hypothetical cases you may know. The system works better than I thought it might; and it does have potential for solving several important inadequacies of the present set of standards.

Summary and Conclusions

On the whole, DSM-5 is a significant improvement over the previous edition. It is more clear, more specific, and more complete than before. Huge efforts have been put forth to harmonize DSM with the World Health Organization’s International Classification of Diseases (ICD) standards, in recognition of increasing globalization in scientific and humanitarian endeavors generally. Much attention has been put into the matter of differential diagnoses, and encouragement is provided throughout to stimulate users to think, “is this code the best, most congruent diagnostic fit with the clinical features of the case I’ve encountered? If I’m not fully sure, what other information would I need to achieve greater clarity?”

Despite facetious predictions among mental health professionals as well as the public at large that with DSM-5, “everyone will be in there, except for you and me, and I’m not so sure about you,” care appears to have been taken in several instances to delimit, rather than expand, overutilization of diagnoses. DSM has always been controversial and will never be perfect, but this edition moves the work in a very respectable future direction.