Introduction

About every 12–18 years, a revision of the DSM is done in order to include advancements in our knowledge that are relevant to improving our diagnostic system. The DSM-IV was first published in 1994, with a text revision being published in 2000. The process for the DSM-5 was begun in 2006—after a series of meetings and publications preparing for this new classification during the previous 4 years—with the recruitment of diagnostic group chairs and 13 committees to address diagnostic groups represented in the DSM-IV. This article is an effort to describe what the current thinking is about the DSM-5 at approximately the halfway point in our work, with special focus on mood disorders. These positions could change substantially with additional outside input, field trial results, and decisions made by the DSM-5 Task Force. The projected publication date for the DSM-5 is now 2013.

The DSM-5 Process

It might be helpful to broadly describe the process thus far and describe what has yet to be done. It is also very important for the reader to realize that the iterative nature and uncompleted aspects of this process render any positions taken at this time subject to considerable change in the “final” product.

A brief overview of the project’s dynamics will help in understanding just why the current positions are subject to change. A total of 13 diagnostic workgroups were formed by the chairpersons appointed by Drs. David Kupfer and Darrel Regier, the director and vice chair, respectively, of the DSM-5 project. Each workgroup chair recruited a workgroup of 10–15 individuals who were willing to serve and to forgo any honoraria from pharmaceutical companies and submit a statement of their earnings from such activities covering the previous 4 years. The mood disorders group initially broke itself into subworkgroups dealing with major depressive disorder and bipolar disorder, headed by Bill Coryell and Trisha Suppes, respectively. Each workgroup freely consulted with outside “advisors” brought into specific aspects of their meetings to provide input on specific aspects of their work. Three additional subworkgroups were formed: the anxiety subworkgroup, the suicide subworkgroup, and a premenstrual dysphoric disorder (PMDD) subworkgroup headed by Kimberly Yonkers and made up of advisors with expertise in this disorder. Workgroup members attended face-to-face meetings in Arlington, VA, about every 6 months. I attended conference call meetings of all the subworkgroups and conducted full Mood Disorders Workgroup meetings, which scheduled conference calls every 2 weeks, in addition to attending the Arlington workgroup meetings and separate task force meetings.

All recommendations of the subworkgroups must be approved by the full Mood Disorders Workgroup and were reconsidered in light of input from an Internet posting of positions considered by each workgroup. Recommendations for the DSM-5 approved by the Mood Disorders Workgroup will be forwarded for testing in field trials, which were being put together as this was being written. The field trials were scheduled to begin in late-summer and require about 1 year to complete. Once the results of these trials become available, the Mood Disorders Workgroup and other workgroups will present their recommendations for the DSM-5 to the task force, which is directed by Drs. Kupfer and Regier and made up of all the workgroup chairs and other members who are chairs of cross–DSM-5 committees and study groups. This group will determine what is written in the DSM-5, which then must be approved by the trustees of the American Psychiatric Association.

We are about halfway through this process, so I have tried to convey the positions the Mood Disorders Workgroup has taken entering the field trials. These positions were modified in response to the comments resulting from the Internet posting of the DSM-5 positions (http://www.dsm5.org) and will be modified to varying degrees after field test results are available, and perhaps further modified by the task force on the DSM-5.

Changes to the DSM-IV Under Consideration

The major themes suggesting the need for changes are the addition of symptom severity scales to the categorical diagnoses for symptom dimensions that affect outcome, the change to specifiers for mixed states that can be added across the bipolar spectrum as well as to major depression, and the addition of a suicide assessment that will result in the clinician declaring the level of concern driving the treatment plan with respect to preventing suicide in each patient. In addition, two new diagnoses are being considered, mixed anxiety depression and PMDD, as well as two new diagnoses for childhood and adolescent patients, temper dysregulation and dysphoric disorder (TDD), which may be renamed, and self-injurious behavior. The TDD category is an effort to provide an alternative to the diagnosis of pediatric bipolar disorder, which has been reported to be increasing at a disturbing rate. A study from Spain has shown a 40-fold increase in office visits for youth (1–19 years of age) during the period from 1995 to 2003, compared with only a doubling of adult bipolar visits [1]. The category of self-injurious behavior is an effort to provide an alternative for the assumption of suicidal intent in patients with repeated self-injurious behavior without suicidal intent.

In addition, we are making an effort to require more specific clinical information associated with the not-otherwise-specified (NOS) diagnosis or what may be considered conditions not elsewhere classified—diagnoses that are used for a growing proportion of mood disorder cases in large Medicare and private insurance databases. Specific diagnostic criteria changes will be suggested, including dropping the bereavement exclusion for major depression and separating psychotic depression from the severity of depression. Major changes across the DSM-5 will be suggested, including dropping the Global Assessment of Functioning scale in exchange for the addition of the World Health Organization Disability Assessment Schedule, with specific cutoffs recommended for caseness (need for treatment) and an overall severity scale that will be diagnostic category specific. In the case of mood disorders, the Patient Health Questionnaire (PHQ)-9 scale will be used for mood disorders, and the Altman Self-Rating Scale for the bipolar disorder spectrum. In the DSM-5 will be the requirement that a patient not only meet diagnostic criteria but also adequate disability and severity criteria to establish caseness. On the less severe end of the mood disorders spectrum, we are proposing that a diagnosis of mixed anxiety depression be added to accommodate patients often seen in primary care practice with anxiety and depressive symptoms that do not meet the criteria for generalized anxiety disorder or major depressive disorder but have enough anxiety and depressive symptoms to result in significant disability. Perhaps reviewing the thinking behind these suggested changes will provoke some questions concerning these recommendations.

In addition to a disability dimension, another dimension—category-specific symptom severity scales to convey the overall severity of each diagnosed disorder—is being field tested. Each workgroup has been asked to submit a severity scale for testing in the field trials. For the Mood Disorders Workgroup, some difficulty was encountered in deciding on these measures. We believed that it was not appropriate to reinvent the wheel without the ability to test reliability and validity, so we looked for scales that had been authenticated in the literature. The first thought was to simply to use a version of the Clinical Global Impression that is frequently used as an overall measure of outcome in medication trials [2]. Although it solved the problem of applicability in major depression and bipolar disorder, it was discarded as not being sufficiently reliable or valid. The PHQ-9 was favored by some for major depression, as it would be useful for measurement-based care and has been fully tested, but there has been some concern about the time that this would require in light of the other dimensional scales being proposed [3]. The workgroup has also considered the Longitudinal Interval Follow-up Evaluation (LIFE) scale, which was used extensively in the National Institute of Mental Health Collaborative Depression Study [4]. The bipolar subworkgroup is favoring the Altman scale, a brief, easy-to-rate scale that has been used in prior studies [5]. These are being tested for utility and feasibility in the field trials.

Addition of Symptom Dimensions to the Categorical Mood Disorder Diagnoses

The Diagnostic Assessment Instruments Study Group headed by Jack Burke is proposing a set of self-rated cross-diagnoses measures (World Health Organization Disability Assessment Schedule II [impairment-disability]) and other level one screening scales [6]. If the patient rates positive on entry level one screening scales, he or she will be offered more specific level two scales to further explore areas ranked positive on level one. With this information filled out before the clinical visit, the clinician will then use this evidence and the clinical interview to make final determinations regarding the relevant dimensions in the individual case. This system was scheduled to be tested for feasibility and clinical utility in field trials beginning this summer.

Medicine and psychiatry have traditionally used categorical diagnoses. In psychiatry, for example, a patient must meet a criterion consisting of a certain minimum number of symptoms. For example, to meet the criteria for a major depressive episode, a patient must be found to have five of the nine criteria symptoms concurrently almost every day for a minimum of 2 weeks. These symptoms cannot be caused by other conditions, such as a medical diagnosis (e.g., head injury, hypothyroidism) or a substance (e.g., steroids). There is an underlying assumption that the severity of the disorder is related to the number of symptoms and the degree of impairment combined with suicidality or psychosis. This is measured by the Global Assessment of Functioning scale and is expressed by the fifth digit of the diagnosis as follows in the DSM-IV: mild (1), moderate (2), severe without psychotic features (3), severe with psychotic features (4), in partial remission (5), or in remission (6). The current proposal for the DSM-5 is to add a pure measure of impairment (World Health Organization Disability Assessment Schedule II) as well as a diagnostic category–related measure (PHQ-9). This assesses the nine criteria symptoms of a major depressive episode on the basis of number of days present in the past week as an estimate of severity across three severity levels. These scores can be easily added to a total severity scale. It is anticipated that this scale of symptom severity will allow measurement-based estimates of the degree of improvement or lack of improvement over the course of clinical treatment. The World Health Organization Disability Assessment Schedule II and PHQ-9 thus will allow a measurement of severity for diagnosis and measurement of treatment outcome by the treating clinician.

Another example of a dimensional approach is the addition of symptom severity for symptoms that are not presently included as criteria for the diagnosis but have been demonstrated by recent studies to affect the outcome of the disorder and thus should be considered targets of treatment. Anxiety severity is one such dimension proposed for addition to the categorical mood disorder diagnoses, as studies have shown that the presence of this comorbid symptom portends poor treatment response and suicidal behavior. We are also proposing an assessment guide for the clinician to use in making a clinical determination of the proportion of clinical attention that should be given to the prevention of suicide in any given patient. Although suicide cannot be predicted in an individual, a clinician is expected to make a determination of the degree of treatment planning that should be devoted to the prevention of suicide in a patient who is being clinically evaluated. Considering the suicide assessment guide (made up of traits, situations, and clinical states that have been found in the literature to confer high-risk status to a patient), the clinician will be asked to indicate a score on a four-point scale (ranging from little or no clinical attention necessary to the total treatment plan being focused on suicide prevention) for any patient being evaluated. It is thought that this recording of a suicide assessment across disorders with increased rates of suicide will draw clinical attention to this problem and provide evidence that a thoughtful assessment was made, no matter what the clinical outcome. It has been suggested that suicidal behavior would qualify as a separate diagnosis that could be added to various categorical diagnoses. Although this could be defended on the grounds of focusing attention on individuals who have histories of prior attempts and would focus attention on the problem, it would not fully address the question of clinical management of immediate risk.

Most specifiers will be retained in the mood disorders section of the DSM-5. Specifiers are used to denote aspects of the diagnosis, such as clinical course (e.g., first episode, recurrent [currently one prior episode], chronic [most of the time in a major depressive episode over 2 years]). These course specifiers are currently indicated by numbers in the DSM-IV. The dimensions mentioned above will be recorded in a different, yet-to-be-decided system.

A clinical example might help illustrate dimensions and specifiers. A 63-year-old industrialist was seen in evaluation for symptoms of depression, sleeplessness, hopelessness, and agitation occurring after his corporation was forced to undergo bankruptcy reorganization in the midst of a major economic downturn. The patient endorsed symptoms of poor appetite, sleeplessness (including trouble falling asleep and middle and early awakening), anhedonia, poor concentration, decreased energy, and suicidal ideation. He also endorsed severe anxiety and worry (about financial ruin) day and night without being able to distract himself and thus to make decisions about his business. Although he denied that he would commit suicide, on examination, he admitted that he felt his family would be better off if he died, and when asked directly, he admitted to thoughts of driving his sport utility vehicle off a mountain road cliff, although he reassured me he would not act on his idea or plan. He also admitted to agitation and pacing. He had been hospitalized recently but talked the admitting psychiatrist into releasing him. He had no prior history of depression, suicidal thoughts, or substance (including alcohol) abuse. He had always been a “bigger-than-life character,” wearing a cowboy hat and boots with many powerful politicians as friends. No hypomanic or manic episodes could be defined in talking with the patient or his wife of 25 years, even though the patient seemed to have functioned at a hyperthymic level.

He was diagnosed by DSM-IV criteria as having had a major depressive episode (first episode, severe, without psychotic symptoms). Because of his high level of anxiety/agitation and admitted suicide plan/rehearsal, he and his wife were told that his life was at imminent risk and that he must be hospitalized immediately. His wife took over and promised to drive him directly to the hospital. Over a 2- to 3-week period, the patient improved while on antidepressant medication and was discharged with follow-up. He readily acknowledged that when he was admitted, he had a plan to drive his sport utility vehicle off the cliff side road on his property so that his wife could inherit his life insurance and thus survive financially. He was subsequently able to reorganize his company and pull it out of bankruptcy reorganization. If diagnosed under the proposed DSM-5, the patient would be assessed as having had a major depressive episode (first episode, severe with severe anxiety/agitation [dimensional measure]) and a level four suicide risk category level indication (dimensional level). Most of the treatment plan would be aimed at preventing his suicide. The patient was observed for 1 year without recurrence of depression, his business stabilized, and at his request, he was tapered off antidepressant medication over a 6-month period with a warning to the patient and his wife to contact me in the event of any signs of depressive recurrence of anxiety. The addition of the anxiety dimension and the suicide assessment dimension makes explicit factors that are key to the management of this case. Had the patient refused hospitalization while denying suicidal intent, he probably could not have been involuntarily hospitalized. In this case, the clinician would have a basis for aggressively treating the patient’s anxiety/agitation and depression while instructing his wife not to leave him alone or let him drive.

Evidence Supporting the Anxiety Dimension

Recent evidence has shown that the severity of comorbid anxiety in mood disorders is a strong predictor of outcome. This includes treatment response [7•, 8–11], time spent in depressive episodes [7•, 12•], and increased risk of suicide [13–15, 16•]. The presence of comorbid anxiety in primary depression (without a prior anxiety disorder diagnosis) has been shown to have a high prevalence with a wide range of severity [17]. Although the argument could be made that anxiety may be as common in major depression as the other nine current criteria symptoms, no studies have focused on this, and the severity of the anxiety may in itself carry valuable information to focus clinical management. For this reason, the Mood Disorders Workgroup is proposing that a separate anxiety severity dimension using a single psychic anxiety scale be used to create an anxiety severity dimension.

In a recent study by Coryell et al. (personal communication), it was shown that anxiety severity at baseline assessment predicted the proportion of time spent in a depressive episode over the following 5–25 years, and that of all the anxiety scales used, only the Schedule of Affective Disorders and Schizophrenia-Change (SADS-C) predicted this across both unipolar major depression and bipolar disorder patients. In this study, obsessive-compulsive disorder symptoms were found to be most strongly predictive of poor outcome over time. This and studies of the outcome of bipolar disorder as well as studies of suicide show that the SADS-C psychic anxiety scale is a useful proxy for the severity of the range of anxiety symptoms seen in mood disorder patients [12•, 13, 14].

Thase [11] recently showed that in the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study, patients rated as having high levels (above the median) on a somatization-anxiety subscale extracted from the Hamilton Depression Rating Scale scores from level 2 ( having not achieved remission on citalopram) responded at one third to one half the level of the patients in the lower median score range for anxiety on this scale. The above studies have influenced the Mood Disorders Workgroup to decide that anxiety severity measurement would make a valuable addition to encourage more comprehensive and successful management of mood disorders.

The workgroup is also suggesting a severity of substance abuse scale, as this dimension also predicts poor outcome and therefore requires therapeutic attention in treating mood disorder patients. The actual scale for substance abuse has not yet been determined, although self-rating scales given to patients as screening tools may suffice.

Another addition will be a suicide assessment guide that provides a list of historical factors, traits, situational factors, and clinical factors that a review of the literature has shown are associated with high risk of suicide. Although as stated previously, it is understood that suicide is not predictable in an individual based on the literature, this assessment will provide a guide for a clinician to make a full assessment. The assessment guide helps provide information to the clinician to record on a four-point scale, which reflects the level of the clinician’s concern for suicidal behavior that will guide the management of a patient. A clinician using this guide and estimating the level of clinical concern reflected in the treatment plan will be establishing evidence of consideration of the risk factors for suicide in each individual case. The weighting of the various factors will be left to the clinician’s judgment in reaching a level of concern, as no evidence indicates that one factor is more important than any other in a given individual. The assessment would tend to focus more attention on reviewing relevant clinical factors for suicide risk and would try to maximize the clinical completeness of the assessment without claiming to be able to make an accurate prediction of suicide. Because it is a standard of care to assess the level of suicide risk yet impossible to predict suicide in an individual case, it is hoped that this assessment guide will improve the assessments in general. If completed by the treating clinician, it will provide evidence that a thorough clinical assessment was done, even in the case of a suicide.

It is not uncommon to find clinicians listing as their only evidence that a suicide assessment was performed the notation “no SI” or “no plan” in many charts. In teaching clinics, it is my experience that a history of a prior suicide attempt generally was not asked about or recorded. Oquendo et al. [18] found in a series of mood disorder patients that those with a history of prior suicide attempt who are thus at a higher risk of suicide received no more intensive pharmacologic treatment than patients with no previous attempt. It is the hope of the Mood Disorders Workgroup that a suicide assessment guide applied across relevant diagnoses will increase the attention paid by clinicians to managing suicide risk.

Several experienced investigators have made a case for inclusion of a diagnostic category for suicidal behavior disorder. Given that prior suicidal behavior is the most potent predictor of suicidal behavior [19–22], even though follow-up studies show that about 10% of patients with prior attempts commit suicide over a 10-year follow-up period, it is a most potent actuarial predictor of future suicidal behavior. Evidence from a familial genetic component for endophenotypes such as impulsiveness adds further evidence supporting suicidal behavior as a diagnosis [23, 24]. Also adding support are findings of the presence of negative affect traits (neuroticism, impulsiveness) in prisoners with family histories of suicide and findings of a specific gene interaction with abuse resulting in alterations in the hypophyseal-pituitary-adrenal system response relating to suicidal behavior [25, 26].

The clinical problem with creating a category of suicidal behavior disorder is that it would depend on a history of prior suicidal behavior. It would not address the problem of clinical recognition of acute suicide risk factors related to recent major loss or threat, recent worsening, recent hospital discharge, and severe anxiety or agitation, which are the most difficult and important for clinicians to assess. We know from the studies of Isometsa and Lonnqyist [27] that in a large sample, 62% of men and 38% of women committed suicide in their first attempt, and of 100 patients who committed suicide on the same day that they saw their clinicians, only 22% admitted to suicidal ideation in this last clinical visit [28]. Recognition and intervention with regard to acute suicide risk is the most important and most difficult task for a clinician. A suicidal behavior diagnosis will not add much to solve this problem and could detract from it in patients with no prior suicidal history. On the other hand, a suicidal behavior disorder diagnosis would exist separately from any other categorical diagnosis and would promote awareness of increased risk. It might also promote further research studies yielding more evidence relating to causes and more data concerning acute suicide risk factors. Objections to suicidal behavior disorder as a diagnosis may be made on the basis of stigma by patients or patient self-help groups.

The Mixed Specifier

Another area of anticipated change in the DSM-5 is an attempt to better document the spectrum from major depression to bipolar depression. Recent studies from the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD), a dearth of studies showing efficacy of selective serotonin reuptake inhibitors in bipolar depression, as well as data suggesting worsening of the course of bipolar disorder with new-generation antidepressants and tricyclic antidepressants attest to the importance of differentiating patients along this spectrum [29–33]. Studies pointing to bipolar features existing in 25% to 40% of patients diagnosed with major depression and studies showing switching over 5- to 10-year follow-up press the case for trying to distinguish these disorders [32, 34–36].

On the bipolar side, the DSM-IV permits the diagnosis of a mixed state only in the presence of full criteria for mania and full criteria for a major depressive episode. Practicing clinicians see patients meeting criteria for bipolar II disorder and even bipolar disorder NOS who exhibit mixed features of the coexistence of depressive symptoms with hypomanic symptoms of increased activity, crowded thoughts, agitation, anxiety, and irritability across the entire severity system of bipolar disorder. For this reason, the Mood Disorders Workgroup is recommending a mixed specifier that can be applied across the spectrum of bipolar disorder and across the spectrum of major depression.

In one move to deal with the soaring rates of diagnosis of childhood bipolar disorder [1], a new diagnosis, currently named temper dysregulation and dysphoric disorder, which had been found in initial follow-up studies over time to commonly lead to major depression or anxiety disorders and not bipolar disorder, is being considered. It would deal with the symptom of severe irritability that often leads to a bipolar diagnosis. Self-injurious behavior is being considered as a new diagnosis principally for adolescents who show repeated self-injurious behavior without suicidal intent. This would help make this distinction diagnostically. In this case, if a diagnosis of suicidal behavior were adopted, it would be paired with this diagnosis.

Other Changes

The DSM-IV category of major depressive episode will not be altered in terms of criteria, not even after the Mood Disorders Workgroup considered a reanalysis by Kendler (personal communication) of his large twin sample, which showed that adding hopelessness to the criteria for major depressive episode added nothing to the prevalence or reliability of the diagnosis.

The Mood Disorders Workgroup has decided to remove the bereavement exclusion from the major depressive episode diagnosis based on data indicating that when a patient meets the criteria for a major depressive episode, the response to treatment is identical to that for any major stressor preceding a major depression [37].

The workgroup has considered whether to recommend that the postpartum specifier for major depressive episode be dropped (not bipolar depression) based on extensive reviews showing no increased incidence of major depressive episode after pregnancy compared with other times among females. Data concerning the clinical importance of recognizing and treating postpartum depression may lead to a recommendation to change the periods of time following parturition to 2 weeks for bipolar disorder and to 6–8 weeks for a major depressive episode receiving the postpartum specifier.

The decision has been made to recommend that the antidepressant induction for mania be dropped as an exclusion criterion for the diagnosis of bipolar disorder if the patient meets criteria for bipolar mania for 1 week after the antidepressant has been discontinued.

Changes Recommended for NOS Diagnoses

As previously noted, efforts are being made to find a way to enable the NOS category that goes across many diagnoses in the DSM-IV to convey more specific information. The Mood Disorders Workgroup’s proposal is to force the use of subcategories that provide more specific information, such as xxx.1 inadequate number of symptoms, xxx.2 inadequate duration of symptoms, xxx.3 inadequate information, and xxx.4 hypomania without history of major depression. As large insurance carriers (Medicaid, private insurance) frequently use NOS diagnoses, we would like to find a way to enable these diagnoses to convey more useful clinical information. Mood disorder NOS is being recommended to be dropped, whereas depressive disorder NOS is being retained from the DSM-IV.

Premenstrual Dysphoric Disorder

A separate subworkgroup of the Mood Disorders Workgroup headed by Kimberly Yonkers and made up of advisors who are experts in the field of PMDD undertook an extensive review of the criteria and emerging evidence regarding prevalence, impairment and course, and treatment response differences from other mood disorders and likely will recommend that PMDD be included as a new diagnosis in the mood disorders section of the DSM-5. This category was included in the appendix of the DSM-IV.

Role of Feedback and Field Trials

All the diagnostic workgroups, including the Mood Disorders Workgroup, were influenced by helpful and astute comments obtained in response to the online posting of our preliminary recommendations from March through May 2010. Changes in our recommendations resulted from this exercise. We are presently preparing for the field trials, which will take in specialty academic settings, practice settings (e.g., offices of psychiatrists, psychologists, and social workers), as well as in primary care settings, and will test about 35 diagnoses. This process will take place in two phases in 2010–2011, and during this time, workgroups will look at various issues that can be addressed before the field trial results are available.

At present, we are working on ensuring that the field trials adequately test the recommendations that we are making at this point. We are debating and considering the meta-structure of chapter titles that will include the diagnostic categories that we are trying to make with as much congruence with the ICD-10 and ICD-11 as possible and to reflect what is known about biological etiologic factors. One piece of evidence of this is a reduction of chapter headings to 10 overall groupings.

After we take into account all the input we have received, the results of the field trials, and reconsidered and perhaps modified our decisions, we will forward our recommendations to the task force. This group will decide what recommendations to accept and how to format all the proposed changes going forward in the DSM-5.

DSM-5 as a “Living” Document

There is a hope that a “living document” form of the DSM-5 can be created on a website that can be updated by a committee as it monitors changes resulting from new evidence in the field as it becomes available. This may allow the inclusion of new scientific findings as they emerge in the DSM without having to wait 12–18 years for another revision. This change, made possible in the Internet age, would allow for incorporation of new scientific findings that are carefully vetted over time, which would accommodate the pace of progress as it occurs. Although it would raise many issues, it could potentially incorporate new knowledge at a much faster rate.

The reader should understand that all the recommendations discussed in this presentation are subject to reconsideration and change as new evidence becomes available during the next 2 years.

Conclusions

The DSM-5 process thus far can be described as a bottom-up process in which workgroups are encouraged to suggest changes to our current diagnostic system, the DSM-IV, on the basis of evidence that some aspect of the current system is not working properly or on the basis of new information in the literature that indicates that a change could improve the utility of the system for the clinician or attain improved reliability or validity. The current work product of the Mood Disorders Workgroup represents the current state of suggestions for changes to be incorporated into the DSM-5. It has been emphasized that the positions cited in this presentation are subject to change in response to outside input, field trial results, and interaction with the task force before they are officially incorporated into the DSM-5.

The positions presented are still subject to considerable change as more information and input become available; thus, changes are definitely likely to be made before the finalized version of the DSM-5 is published, The hope is to produce a “living document” that can be updated on an Internet site as new findings allow us to approach the problem of greater validity of our diagnostic system in psychiatry.