Public Health Implications of Changes in Psychiatric Classification (September 25-28, 2007)
Prepared by Michael B. First, M.D., DSM Consultant to the American Psychiatric Institute for Research and Education (APIRE), a subsidiary of the American Psychiatric Association
The APA and the World Health Organization (WHO), in collaboration with the National Institutes of Health, convened a research planning conference focusing on the public health aspects of the diagnosis and classification of mental disorders at WHO headquarters in Geneva, Switzerland, on September 26-28, 2007. The conference was the twelfth in a series of NIH-funded conferences on “The Future of Psychiatric Diagnosis: Refining the Research Agenda” that is administered by APA’s American Psychiatric Institute for Research and Education (APIRE). The conference co-chairs were Benedetto Saraceno, M.D., Director, Division of Mental Health and Prevention of Substance Abuse, WHO (Geneva, Switzerland) and Norman Sartorius, M.D., Ph.D., (Geneva, Switzerland). The goals of this conference were to review the available evidence and experience concerning the public health implications of the diagnosis and classification of mental disorders and to make recommendations on how DSM-V and ICD-11 can best serve public health needs globally. Twenty-four invited scientists from around the world participated. Ten Conference Expert Groups (CEGs), involving more than 80 contributors, were created prior to the conference in order to develop background papers in advance of the conference focusing on various public-health-related topics, including the public health implications of definition of mental disorders, stakeholder perspectives, implications for prevention, the use of mental health classification in primary care, forensic implications, economic implications, implications for statistics and information systems, the interface of the classification with ICD Z codes and the International Classification of Functioning (ICF), implications for education and training, and translating psychiatric diagnosis and classification into public health usage. The conference was divided into three parts. The first part of the conference consisted of a series of presentations summarizing each of the background papers, presented by the lead member of each CEG. The second part included presentations of summaries of recommendations from each of the ten prior diagnosis-related conferences in the Refining the Research Agenda series which covered personality disorders, substance use disorders, stress-induced and fear circuitry disorders, dementia, deconstructing psychosis, obsessive-compulsive behavior spectra, dimensional approaches, somatic presentations, externalizing disorders of childhood, and depression and generalized anxiety disorder. The final part divided the conference participants into three working groups that were given the charge to formulate recommendations for the conference, with one group focusing on structural and process issues, disease-specific issues, and dissemination and use issues.
After presentations by Shekhar Saxena, M.D., (Geneva, Switzerland), Coordinator of the Team Mental Health: Evidence and Research at WHO, outlining the working plan of the meeting and by Darrel Regier, M.D., (Arlington, VA), Director of Research and Executive Director of APIRE, summarizing the methods and goals of the Refining the Research Agenda conference series, conference Co-Chair Benedetto Saraceno, M.D., (Geneva, Switzerland), described WHO’s perspective on the involvement of stakeholders in the revision of the classification system. He noted that the logical structure of the classification system is vital for an organization like WHO which provides global statistics of morbidity and mortality. Due to the diversity of its function, classification revisions should be done in consultation and collaboration with its stakeholders. Stakeholders can be divided into indirect stakeholders and direct stakeholders. Indirect stakeholders include those sectors that might have an interest in classification, like the pharmaceutical industry, for whom the introduction of a new disease or deletion of an existing disease might be of interest. Direct stakeholders include health care providers, governments, and service users and family members. Providers include not only psychiatrists but also other mental health professionals, like psychologists, nurses, social workers, and others. Historically only psychiatrists have been influential in the development of the classification. Issue of representativeness also needs to be considered; groups from rich countries may have excessive influence over the process. The influence of the pharmaceutical industry on professionals must be taken into account to counteract the tendency of medicalizing ordinary life experiences to sell drugs. It is therefore necessary for WHO and APA to examine and scrutinize potential conflicts of interest. Governments are also responsible for providing treatment, social services, and pensions which depends on having clear classifications of diseases. The needs of psychiatric service users and family members (consumers, clients, patients) also need to be considered. Classification involves not only scientific facts but also a consideration of values. It is not immediately clear how best to include their input, however. Involvement of users brings with it the problem of representation; only a few users recognize themselves as part of an organization, some of whom only represent a few people. Service users and patient organizations reflect diverse and sometimes opposing views that are not easily amenable to resolution. Some consumer organizations are supported by pharmaceutical companies, creating a potential conflict of interest. Nonetheless, some level of stakeholder involvement in the revision process is necessary; classification should not solely be the province of health care professionals.
Norman Sartorius, MD, PhD (Geneva, Switzerland) gave a presentation on the public health implications of the definition of mental disorders. He noted that the challenge is to harmonize the definitions that have been made by different groups since the consequences of the definition are not the same from one group to another. For example, it was announced the all mentally ill persons in France are now classified as being disabled and can now get a pension. While this may be a help to the families of these patients, it ma not be as helpful to the individuals actually suffering from mental illness or to society as it may make it more difficult for such individuals to work. The threshold of disease can be set in absolute terms (e.g., the presence of a particular symptom, like psychosis, equals disease) or certain treatments might define the disorder. In general, disorder threshold involves three sources of information, impairment and consequent disability, distress, and symptoms. ICD has made an effort to keep disability out of the definition because disability depends on the social environment. DSM, in contrast, includes disability as one of the defining characteristics of disorder. It has also been proposed that the social desirability of the condition may play a role in the definition; for example, if an individual has damage of the corpus callosum, it must not be considered a disease in a dis-literate society where reading skills have no social value. Dr. Sartorius then discussed the issue of the stigma and the inevitable discrimination associated with being labeled as having a mental disorder. Stigma is sometimes related to the disease name, raising the question of whether it is possible to define a disorder without giving it a name. Changes in Japan in the name of schizophrenia (from a Japanese word meaning essentially “broken brain” to a term with less severe connotations) will allow us to study the impact of a disorder’s name on stigma. Dr. Sartorius concluded by noting that the main points he is making will be covered in the other presentations at the conference, i.e., that the public health implications relevant to the area in question (e.g., forensic, economic) depend on the definition of mental disorder that is used.
Diana Rose PhD and Graham Thornicroft MD (London, UK) presented next on service user and care stakeholder perspectives on diagnosis and classification. There are a wide range of service user’s reactions to receiving a diagnosis, including feelings of relief (e.g., naming the unspeakable, providing directions towards help from professionals or other service users), feelings that the diagnosis is unhelpful because of anticipated stigma and disadvantages, rejection (e.g., “the label will define me and be something that I can never get rid of”), and denial (e.g., “this cannot be me”). Moreover, many service users receive more than one diagnosis throughout their psychiatric career, sometimes more than one at the same time. User research (i.e., research conducted by service users) has blossomed in the last ten years in the UK. For example, SURE [Service User Research Enterprise] at the Institute of Psychiatry in London sponsored a study on consumers’ perspectives on ECT conducted by two researchers who had themselves received ECT. It concluded that professional research overestimates consumer satisfaction with ECT. Therefore it would be informative to carry out service user research into consumers’ views of diagnostic and classificatory systems. The CEG made several recommendations, including: 1) that the new diagnostic manuals could contain a chapter explaining to clinicians how psychiatric diagnoses, especially those which may be seen to be more severely disabling, can have stigmatizing effects which could also include references to resources which can assist individuals and families in coping with stigmatization; 2) that clinicians be given guidance on assisting service users to recognize their assets and the positive aspects of their lives in any diagnostic consultation and that they should include statements such as “X diagnosis does not preclude positive behavior such as being able to work or marry”; and 3) a short user-friendly version of any new diagnostic manual should be prepared so that service users know what their doctor is talking about. There is very little research on how service users and families react to psychiatric diagnoses and what research there is suggests great diversity in such reactions. Drs Rose and Thornicroft concluded with a number of recommendations for future research, including: 1) researching whether changing the names and diagnostic terms changes their degree of stigmatization; 2) exploring the effects of different methods of giving and offering diagnoses, the comprehensibility of different rationales for the diagnosis, and how far these are accepted by service users; 3) determining the extent to which diagnostic information needs to be tailored to the needs of each person in terms of rationale, vernacular and implications for care; 4) evaluating whether decision-aid methods assist the communication of diagnoses; 5) determining the impact of providing opportunities to engage carers and family members for information-sharing at the time of diagnosis within proper limits of confidentiality; and 6) determining the degree to which receiving a diagnosis is experienced as an empowering or a disempowering process by services users.
Pratap Sharan MD, PhD (New Delhi, India) presented next on the implications for revisions of diagnosis and classification systems on the prevention of mental disorders. Given the limitations in access and effectiveness of available treatments, prevention can be an important method for reducing the burden imposed by neuropsychiatric conditions (estimated to be 13% of the global burden of disease). The structure of current classificatory systems is inadequate for providing comprehensive coverage for preventive approaches, for example, their exclusive focus on the individual vs. risk groups and populations. Moreover, many distal risks to health, such as socioeconomic disparity, cannot be defined appropriately at the individual level. The CEG proposed that an axis for protective and risk factors be added either to the main body of the classificatory system, in an appendix, or else be included in the form of a descriptive section on protective and risk factors in the text. Subthreshold conditions have considerable public health importance and may better be seen as indicators of risk and a focus for indicated prevention. These conditions, however, are usually excluded from classificatory system because current diagnostic thresholds have a treatment rather than a prevention perspective. The group recommended that prevention-relevant subthreshold conditions be included in various phenomenological groups as categories or text descriptions and that as etiologic/pathogenetic data accumulate, prevention-relevant categories be considered for future inclusion of early manifestations and risk states predictive of disorders, (e.g. family members of patients with Alzheimer's disease who are homozygous for Apo E -4 alleles). High risk studies have shown that prodromal features and biobehavioural anomalies can be detected in the offspring of persons with mental disorders and that trials of preventive interventions have been found to be effective in conditions like ADHD, affective disorders, and substance use disorders. The CEG proposed that, where there is adequate evidence, the classification system should provide a listing of early or prodromal symptoms and biobehavioral markers and features of parental psychopathology and relevant risk factors associated with disorders. Given that most risk factor profiles have continuous distribution (due to multiple underlying causes of vulnerability), the CEG also recommended that an axis for classification of agreed upon dimensions with good literature support (e.g., dependence severity) should first be included in the appendix. Current classificatory systems have a cross-sectional approach. This neglect of a longitudinal perspective, however, is at odds with the life-course approach to the study of health and illness. The CEG recommended the use of diagnostic criteria that are developmentally informed, and to include a listing of relevant developmental information on risk and protective factors on an axis or in descriptive text. The preoccupation with over-refined categories has led to complex diagnostic systems that are difficult to use in the community setting. As several mental disorders have major proximal risk factors in common, subdivisions are less relevant to prevention. Most evidence-based preventive trials focus on generic protective and risk factors and typically address broad categories such as depression. The CEG recommended that in making refinements to the classification, broader categories should be preferred and, where feasible, that categories should be combined (e.g. the category of adjustment disorders could be merged with subthreshold categories in relevant phenomenological groups). Finally, the CEG recommended revision of the multiaxial system as follows: Axis I: Categorical diagnoses; Axis II: Dimensions related to mental disorders; Axis III: Protective and risk and factors; Axis IV: Function and dysfunction (ICF); and Axis V: Quality of life.
Linda Gask, PhD (Manchester, UK) presented on the adaptation of psychiatric diagnosis and classification systems for primary care use and the implementation of these systems in primary care settings. Primary care is where most mental disorders are seen and treated. There are a number of challenges in diagnosing psychiatric disorders in primary care settings. Although primary care providers may not be perceived as seeing complex mental health problems, psychiatric presentations in primary care are quite complex; some patients are distressed without meeting criteria for a psychiatric disorder and other patients will meet criteria for a disorder without experiencing distress. There is also a tremendous amount of comorbidity in primary care settings, both with physical illness and other common psychiatric disorders; Subthreshold conditions are prevalent and often associated with disability (e.g., mixed anxiety depression). DSM-IV and ICD-10 were developed in Western specialist care settings and were not applicable to primary care in much of the world. Adaptations of these systems have been made for use in primary care; the DSM-IV-PC does not work very well, however, and the ICD-10-PC works well but does not address many issues that primary care providers would like to see addressed. The International Classification for Primary Care (ICPC-2, developed by WONCA) has many advantages that should be incorporated; it includes codes for indicating the patient’s reasons for the encounter and the general practitioner’s diagnosis. A classification system for primary care should also have provisions for indicating chronicity (i.e., how long the patient has experienced symptoms or problems), disability: using reliable self-rated measure (e.g. SFQ) and social problems (e.g. using ICPC-2 Z codes). An effective system has to be simple, address diagnosis, severity, and chronicity, should include both dimensional and categorical approaches, be linked to disability assessment, easily utilized for routine data gathering, facilitate training of primary care workers, aid communication with specialist care and be locally adaptable for specific populations and countries. The collective experiences of WHO (in developing and testing ICD10-PHC) and WONCA (in developing and testing ICPC) provide a strong platform on which a new classification can be built.
Julio Arboleda-Flórez, MD, FRCPC, PhD (Kingston, Ontario, Canada) presented on the public health implications of diagnosis in forensic psychiatry. Forensic psychiatry deals with issues arising in the interface between psychiatry and the law, and with the flow of mentally disordered offenders along a continuum of social systems. Dr. Arboleda-Flórez noted that there is a poverty of diagnosis in forensics. Symptoms, rather than diagnoses, are supreme because of their severity and incapacitating qualities. An important component of forensic evaluations is the assessment of the individual’s disabilities and functional incapacities, which are particularly relevant in declarations of incompetence. In the legal process diagnosis is not essential, but mental states and functional disabilities are of paramount importance. There is also the danger of providing inappropriate excuses to unlawful behaviors via medicalization through diagnostic labeling. A diagnosis suggesting a link to reduced functional capacity, which is formulated in such as way as to equate the diagnosis with impaired functional capacity will carry with it significant implications in law. Therefore any classification system should avoid suggestions that a diagnosis leads ipso facto to volitional paralysis or inability to differentiate right from wrong. Furthermore, forensic psychiatry should not prescribe which concept of mental disorder or mental illness is adequate from a legal point of view for this is a normative question. Better validity and reliability of psychiatric diagnosis can improve discussions with lawyers. Empirical matters of science are for psychiatrists to decide, but socio-political and legal questions are for lawmakers to resolve.
Darrel Regier MD, MPH (Arlington, VA), filling in for Howard Goldman, MD, PhD (Baltimore, MD) presented the recommendations of the CEG on the economic implications of diagnosis and classification. These economic implications are apparent in several important areas; when doing any economic analysis of cost, when talking about burden of illness, when examining insurance or disability assessment and when predicting some of the costs of hospitalization. When studying the relationship of diagnosis to service need, the financial implications depend on whether it is related to any service needs or whether it is related to a specific need like hospitalization. In general, the characteristics of diagnostic criteria at the syndrome level are of much less value than considering clinical significance and level of disability. Furthermore, how the diagnosis is used in terms of a target for service policy can have very different impacts; for example, a community mental health center program that is setting out to cover a wide range of disorders and aims to prevent mild disorders has a different impact that one that aims to reduce institutionalization. Often many reforms that were designed initially to deal with severe mental illness are transformed into systems of care for milder patients that are much easier to treat. The relationship between diagnosis and changes to economic policy can make a difference in how a division of labor is established between specialists and primary care generalists. If it is determined that the primary care physician should take the bulk of treatment responsibility, it will have some impact on the relative costs of care as well. Different perspectives can guide rationing of care; rationing can be based on need (i.e., the sickest individuals should have the bulk of resources) vs. who might benefit most in terms of productivity. When conducting burden of disease studies with epidemiological data, how you do diagnosis is very important. One needs a good set of criteria with good data on incidence, prevalence, duration, and disability level. A significant weakness of current global burden of disease measurements is the lack of some of these data. If the next revision could have better information on duration and disability, it would result in better estimates of the relative burden associated with specific disorders.
Diagnosis also plays a role in insurance eligibility. Some plans will accept all DSM and ICD diagnoses. When states started getting parity for coverage, the term “biologically-based” disorders, which usually included Major Depressive Disorder, Panic Disorder, Schizophrenia, OCD, Bipolar, or Schizoaffective, was created to limit coverage to those disorders. Many other disorders were excluded, especially substance use disorders and eating disorders. After managed care become more widely adopted, diagnosis became less important to assess for “medical necessity,” instead assessments of impairment, potential for self harm, and treatment need took precedence over diagnosis.
Walter Gulbinat, Dipl.-Math. (Lichtenstein, Germany) presented on the implications of diagnosis and classification on mental health information systems. Mental health information systems are primarily used for patient management, program and policy management, administrative purposes, and in support of research programs. A prerequisite for a meaningful use of information generated from different sources by different people representing different professions is an agreement on the use of a common language. To date, there is no common language to cover all the different types of information entered into and processed by mental health information systems. Nevertheless, great efforts have been devoted to developing classifications for international use in a great number of domains relevant to health and mental health. The WHO constitution mandates the production of international classifications on health. In particular, WHO maintains three core or reference classifications: the International Classification of Diseases, the International Classification of Functioning, Disability and Health, and the International Classification of Health Interventions. In addition, the WHO Family includes a number of related and derived classifications, such as the International Classification of Primary Care (ICPC) and International Classification of External Causes of Injury (ICECI). The United Nations Statistics Division also maintains a registry of classifications of direct relevance to health and mental health. This family is comprised of reference classifications on such matters as economics, demographics, labour, education, social welfare, geography, environment and tourism (e.g., International Classification of Status in Employment [ICSE]). Three other classifications and nomenclatures of wide international distribution with a direct bearing on mental health and mental health information systems include the DSM, the International Classification of Nursing Practices (ICNP) and the Systematized Nomenclature of Medicine (SNOMED). The CEG made a number of recommendations for improving use of classifications in mental health information systems including: 1) increasing awareness of these international classifications among mental health professionals and promoting their use; 2) intensifying efforts to improve the compatibility of successive revisions of mental health classifications and the compatibility of classifications pertaining to mental health issues given that incompatibility of various versions can jeopardize communications; 3) minimizing the time lag between release and the actual use of a new version of a classification (e.g. by making the new classifications available in the languages required); 4) ensuring the adaptability of classifications of mental health importance for use at different levels of health care, and in low resource countries or situations; 5) encouraging the use of uniform definitions for representing and recording clinical data and ensuring that systems incorporating multiple methods of record keeping or reporting (e.g. both paper-based and electronic) are compatible; 6) Training on proper use of classifications and definitions should be provided to users of mental health information system; 7) developing, validating, and standardizing screening, diagnostic and psychometric instruments for use in various social-cultural contexts (including developing countries) to create a data base comparable across facilities and across countries (including developed countries); 8) intensifying efforts to reach agreement on definitions, nomenclature, and classifications of the data elements commonly used in electronic health records; and 9) ensuring that information in electronic health records and data bases is protected and access restricted to avoid inappropriate access.
Bedirhan Ustun MD (Geneva, Switzerland) presented on functioning and disability
and contextual factors in the diagnosis of mental disorders in DSM and ICD. The CEG focused their recommendations on two tasks: the relation of the classifications to the ICD Z codes which are available for classifying contextual factors, and the potential linkage of the diagnostic classifications to the International Classification of Functioning. DSM-IV currently classifies contextual factors on Axis IV (psychosocial and environmental problems); a similar axis is included as part of the triaxial system of ICD. ICD Z codes try to capture health status and contact with health services (e.g., Z40−Z54 for Persons encountering health services for specific procedures and health care; Z55−Z65 Persons with potential health hazards related to socioeconomic and psychosocial circumstances). There are similarities between those two systems so that they could be made compatible. The CEG laid out three options for classifying contextual and environmental factors: use the ICF environmental factors codes, use the ICD Z codes within DSM axis IV; or merge the multiaxial systems of ICD and DSM. To evaluate the coverage and possible utility of these three options, a crosswalk between the DSM-IV-TR, the ICD-10 Multiaxial Presentation, and the ICF Environmental factors codes should be completed to determine joint coverage areas and possible missing dimensions. As a potential research area, the scope of contextual factors could be explored by computerized search techniques of patient health records to create an "item pool" that covers the universe of commonly used contextual factors. DSM-IV and ICD-10 have adopted different approaches to the incorporation of disability in the classification of mental disorders. DSM-IV requires impairment or distress in addition to the symptomatic presentation for all disorders, although no guidance is provided regarding how distress or impairment in functioning should be operationalized or independently assessed. In contrast, caseness in ICD is defined as a combination of a symptomatic picture from the ICD-10 and a level of disability from the ICF. The ICF is divided into 10 areas of functioning (e.g., communication, interpersonal interactions, community, social, and civil life) which can be assessed using an instrument called the WHO-DAS. The CEG proposes that diagnosis of a mental disorder in DSM-V should be uncoupled from disability as in the case with physical disorders. Specifically, they recommend that no functioning or disability phenomena should appear in a diagnostic criterion, that there be a separate rating of the disorder severity (i.e., mild, moderate, or severe) which would rely on the individual’s symptom count or nature of symptoms (e.g., suicidal ideation or any significant symptom) rather than disability (functional impairment), and that "functional impairment" criteria should be operationalized using common domains. Furthermore, they recommend that in DSM-V "disability" or “functioning” could be formulated as a redefined axis replacing current Axis V that would serve to document only disabilities (excluding any sign or symptoms) and have clinically meaningful anchor points depending on the extent of functioning and sensitivity to change by clinical interventions.
Horst Dilling, MD (Lübeck, Germany) presented on the implications of diagnosis and classification on education and training. He noted that there were no specific instructions provided by WHO for how to do ICD-10 training and no scientific account of the efficacy of training. By and large the various WHO collaborating centers accepted the task of distributing ICD-10. Detailed examples were given about the introduction of ICD-10 and DSM-III in different countries. The CEG made a number of recommendations regarding education and training for ICD-11. There should be preparation and publication of materials for education, training, use in the clinic, and research; such material could include classification textbooks, a research version of the classification, a primary care version, short versions of the classification textbooks, internet and PC-versions, a lexicon or glossary, case books, and a structured clinical interview and other instruments. Initial seminars should be organized for trainers at the international, national, language, regional, and institutional levels (i.e., teaching the teachers). Centers and national representatives should be appointed who are responsible for the implementation process. Common training approaches should be developed consisting of lectures and seminars, as well as internet conferences and interactive learning programs. Different training modules, however, will need to be developed to serve the needs of the various target groups. The CEG defined four target groups: Group A, consisting of psychiatrists, psychologists, psychotherapists, other physicians (specialists), will need in-depth knowledge on the use of ICD-11 in clinical practice and research; Group B, consisting of social workers, nursing staff, and other auxiliary staff in the mental health care system, will need less detailed training in how to apply the ICD-11 in practice; Group C, consisting of administrative and health insurance personnel, would need only basic knowledge of ICD-11; Group D, consisting of those affected and their relatives, partners, friends and the general public, would get general information on the ICD-11. There will also need to be quality assurance of the introduction process at both a scientific level (e.g., studies of inter-rater reliability, implementation in different health care settings) and the individual level (e.g., evaluation of training courses, re-evaluation of the training outcome after a defined period). Finally, there needs to be a public relations effort to include an ICD-11 web page in different languages, press conferences, and seminars for special groups (e.g., teachers, law enforcement agents). The amount of educational work needed for the introduction of a new classification will be related to the extent of changes actually made to the classification (e.g., if the changes from ICD-10 to ICD-11 are relatively minor, the educational efforts will be by necessity less complex).
Alberto Minoletti, MD (Santiago, Chile), presented on translating psychiatric diagnosis and classification into public health usage. ICD and DSM have proved to be useful for public health, facilitating the determination of prevalence of mental disorders in countries with different income levels, the determination of burden of disease and disability, the determination of service needs, detecting treatment gaps and inequalities, and for advocacy for mental health. ICD and DSM have also been useful to the implementation of national mental health policies and plans. Because of the different estimates of prevalence and incidence that result from the use of different classifications, a universal mental health diagnostic system is needed, raising the question of whether it is possible to merge ICD and DSM classifications into one system. A diagnostic system should be feasible for use in low and middle income countries. For this to be so, it must be simple, with names accessible to the general population; be trans-cultural with validation studies in different countries; be applicable to promotion and prevention; be feasible for use in primary care settings; have both high reliability and diagnostic sensitivity, include diagnostic categories that respond to specific interventions, be widely accepted by MH professionals, be able to minimize stigma, and be free of conflicts of interest. There needs to be an interaction between the designers of the classification systems and public health researchers so that public health views can be integrated with clinical views. Diagnostic systems should take into account a community perspective and include sub-threshold conditions associated with high utilization of services (especially those with impairment and comorbidity). Dr. Minoletti concluded with a proposal for a four axis multiaxial diagnostic model that would be useful for both public health and clinical work. Axis I would be for the diagnosis of syndromes, with categories divided into small groupings only when there is hard evidence that they require different treatments. A simplified Axis I would have several advantages including increasing the likelihood that the classification would be learned and applied by general practitioners, it would have higher diagnostic reliability; be easier to incorporate into health information systems, and be easier to validate in different cultures. Axis II would be for indicating the level of distress, as the level of subjective suffering is directly associated with the frequency of help-seeking behaviour. The challenge is to develop distress scales that can be brief, simple, time efficient and trans-cultural. Axis III would be for indicating the level of functional impairment or disability. Such an axis would have strong financial implications since disability is associated with high health and social services utilization, diminished work performance, sick leaves, and disability pensions. Axis IV would be for indicating dangerousness to self or others. There is evidence that this variable also correlates with the utilization of health resources and it might also facilitate early identification of persons that require urgent and intensive treatment.
The conference continued with presentations that summarized the recommendations from each of the ten prior diagnosis-focussed research planning conferences. Thomas Widiger, PhD (Lexington, KY) presented a summary of the Personality Disorders research conference which was concerned with the development of a research foundation for dimensional models of personality disorders. The impetus for developing dimensional models of personality grew out of dissatisfaction with the arbitrary boundaries of the current categories which lead to wide differences in prevalence rates, excessive diagnostic co-occurrence, lack of coverage, and an inadequate scientific base. Although there are at least 18 different dimensional models of personality that have been proposed by various researchers, the models converge on five broad domains: emotional instability, introversion vs. extraversion, antagonism vs. compliance, constraint vs. irresponsibility, and unconventionality vs. closedness to experience. Considering the public health implications of dimensional models of personality, Dr. Widiger noted the potential for reduced stigma because of the opportunity provided for individual descriptions instead of lumping traits into potentially stigmatized categories. Dimensions based on normal personality traits also allow for the identification of personality strengths as well as weaknesses. Dimensions also facilitate the adoption of specialized cut-off points for hospitalization and other types of decisions.
John Saunders, MD (Sydney, Australia) presented a summary of the Substance Use Disorders research conference. Key issues identified by the conference included: 1) establishing an addictive disorders domain, which encompasses both substance use disorders and non-substance addictions (e.g. internet addiction); 2) determining the biological basis of substance use disorders and its relevance to diagnosis, including developments in knowledge of the neurobiology of substance use and dependence, the impact of genetics, and the value of imaging techniques for diagnosis; 3) considering how to integrate categorical versus dimensional approaches in substance use disorders; 4) trying to reconcile the DSM and ICD concepts of dependence, including the issues of operationalizing craving, weighting items, and the efficiency of diagnosis; 5) comparing the performance and usefulness of ICD 10 harmful use and DSM-IV abuse categories, determining the relationship of abuse to dependence, and the residual non-dependence category (i.e., NOS); 6) considering modification of diagnostic criteria for special groups, including adolescents, the elderly, different cultures, and chronic pain patients; 7) determining the applicability of current diagnoses and criteria to specific substances; 8) delineating the clinical syndromes of cannabis withdrawal, caffeine withdrawal, anabolic steroid use, MDMA use and dependence, and nicotine abuse; and 9) clarifying the comorbidity between substance use disorders and mental disorders, including the definition and delineation of substance-induced mental disorders, the overlap among addictions (e.g. poly-substance dependence), and criteria for diagnosing mental disorders within the context of substance use and its disorders. Conclusions of the conference were that: 1) the generic DSM-IV and ICD-10 dependence syndromes perform well; 2) the differences between DSM-IV and ICD-10 dependence are slight and could probably be reconciled; 3) DSM-IV substance abuse is a less satisfactory diagnostic entity than dependence; 4) ICD-10 harmful use is very unsatisfactory, both psychometrically and in terms of utility; 5) there are many people in the general community and in clinical samples who are not captured by any of the above diagnoses; and 6) DSM-IV and ICD-10 cater better for the clinical rather than the epidemiological and public health communities. Conference recommendations were grouped on the basis of whether they could be addressed by analysis of existing data sets (e.g., does the general approach for defining substance use diagnoses apply for specific age or ethnic groups such as adolescents, the elderly and people from different cultures?); those issues requiring further work to generate specific research questions (e.g., is there support for a protracted abstinence syndrome?); questions that should be addressed by literature reviews but where the research opportunities were considered to be limited by a lack of data (e.g., how can the validity and reliability of substance use disorder criteria be established across different cultural and ethnic groups?) and recommendations not specially requiring data analysis that could be taken up by the DSM-V Committee.
Gavin Andrews, MD (Darlinghurst, Australia) presented a summary of the Stress-Induced and Fear Circuitry Disorders conference which covered four disorders: PTSD, panic, social phobia, and specific phobia. In considering possible ways to restructure the disorder groupings, comorbidity among disorders may be a useful guiding factor; suggesting a major grouping of internalizing disorders divided into misery disorders (major depressive disorder, dysthymic disorder, generalized anxiety disorder, PTSD, somatoform disorders) and fear disorders (social phobia, panic and agoraphobia, specific phobia, OCD). Although a genetic factor for neuroticism significantly affected all of the internalizing disorders, a common genetic factor independent of neuroticism accounted for degrees of variance and covariance among major depression, generalized anxiety disorder, and panic disorder. Neuroimaging studies with positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) have begun to describe the functional neuroanatomy of emotion. Information processing biases in attention, interpretation, and memory have been examined across the internalizing disorders using a variety of paradigms. While the content of cognitions reliably distinguishes between the disorders, our knowledge of underlying cognitive processes cannot robustly inform classification. While a number of environmental stressors raise the probability of a fear disorder occurring, there is little specificity between a particular stressor and the onset of a particular disorder. Even in PTSD, the issue is why did this person develop the disorder, seeing that four out of five people who encounter the same stressor do not. Dr. Andrews concluded by noting that at the simplest level social phobia, PTSD, panic/agoraphobia, and specific phobia form a coherent group characterized by fear and avoidance of certain situations, attentional biases at the strategic level for processing of threat related words and amygdalocentric processing of this information.
Dilip Jeste, MD (San Diego, CA) presented a summary of the Dementia research conference. Presentations at the conference pointed out some problems with the existing DSM and ICD criteria for dementia and noted that although progress has been made in finding biomarkers for dementia, the field has yet not advanced to the point where this is currently diagnostically relevant. A number of general recommendations were made, including that revised diagnostic categories be validated, that epidemiologic studies be conducted with these validated criteria, that studies be expanded to include dementias other than Alzheimer’s disease (AD), that biotechnologies (e.g., functional neuroimaging and genetics/genomics) be applied to identify syndromes, subtypes, risk factors, and treatment mediators/moderators, that at-risk populations be identified for longitudinal studies (e.g., those with Mild Cognitive Impairment [MCI]), that potential prevention strategies be developed for these at-risk patients, that outcome measures be refined for clinical trials to include functional ability, quality of life, and economic measures; and that novel pharmacological targets specific to dementia etiology (e.g., beta-amyloid for AD) be identified. It was further recommended that MCI, or at least the amnestic type of MCI, should receive serious consideration for inclusion in DSM-V and ICD-11 and that the imaging portion of the text for dementia be expanded, for example, the role of FDG-PET (showing parieto-temporal or fronto-temporal hypometabolism) in the diagnosis of AD or MCI. Finally, Dr. Jeste recommended that specific behavioral syndromes associated with dementia (e.g., psychosis of AD) be added to allow better studies of their prevalence and incidence, to facilitate recognition and treatment of these conditions in clinical practice, to promote hypothesis-based research leading to a better understanding of the conditions and to enable treatment research in diagnostically homogeneous populations.
Mario Maj, MD, PhD (Naples, Italy), summarized the findings of the Deconstructing Psychosis conference, focusing on five major issues and recommendations. There is an apparently continuous distribution of delusions and hallucinations in the general population (e.g., the community prevalence of “any rating of psychotic or psychosis-like symptoms” was as high as 17.5% in the NEMESIS study). Consistent with this view is that the risk factors for psychosis in the community are the same as those for schizophrenia (e.g., low education, urban living, cannabis dependence). The group, however, recommended that it should not be taken for granted that delusions and hallucinations are appropriate indicators of the latent construct of schizophrenia and that population-based studies using appropriate statistical analyses (e.g., coherent cut kinetic methods) should test the assumption that schizophrenia exists as a categorical latent variable. A second issue is the detection of high-risk subjects and the prediction and prevention of their transition to full-blown psychosis. Projects aimed at the identification of high-risk subjects should mainly focus on patients contacting primary health services for psychological problems and further studies are needed on coping styles and perceived control as variables involved in the development of functional impairment and need for care. A third issue concerns the considerable degree of clinical and genetic overlap between schizophrenia and bipolar disorder. Evidence for common susceptibility genes for bipolar and schizophrenia include the fact that while a co-twin of an individual with schizophrenia has a 40% chance of developing schizophrenia, there is an 8% chance of developing mania and 8% of developing schizoaffective disorder. It should be noted that there is also considerable evidence supporting a difference between the two (e.g., different pattern of cognitive performance in schizophrenia and bipolar). The conference group recommended that although the schizophrenia/bipolar dichotomy should not be abandoned at this stage, the classification systems should allow for a supplementary dimensional characterization of each patient with psychosis, for both clinical and research purposes. A fourth issue relates to the evidence suggesting that cultural factors may affect the occurrence and manifestation of psychotic disorders. For example, African-Caribbean individuals with a diagnosis of psychosis are 40% less likely to suffer from a continuous illness than British whites, and are less likely to have a history of obstetric complications or neurological illness. The group recommended further research on the risk factors, incidence, phenomenology and course of psychotic disorders in different ethnic groups and that assessment instruments should be subjected to more sophisticated field trials in non-Western populations. Finally, given the extensive amount of data concerning neuroimaging and biobehavioural correlates of psychotic disorders, the group recommended that the subsections of laboratory findings in the text of DSM-V and ICD-11 be supplemented with research-oriented, evidence-based annotations on recommended endophenotype measures.
Joseph Zohar, MD (Tel-Hashomer, Israel) summarized the recommendations of the Obsessive Compulsive Behavior Spectra conference. The group focused on four questions: 1) are obsessive compulsive spectrum disorders part of Anxiety Disorders; 2) if they are separate, then what disorders should be included in such a spectrum; 3) what system could be used to diagnose these disorders?; and 4) is there a case for a schizo-obsessive subtype of schizophrenia? Data supporting removing OCD from the anxiety disorders included different comorbidity patterns (more likely to be mood disorders rather than other anxiety disorders in OCD), OCD-related disorders (e.g., Tourette’s Syndrome) are particularly common in OCD, there is an increased prevalence of OCD as compared to other anxiety disorders in family members, there is a preferential response to serotonin reuptake inhibitors in OCD as compared to other anxiety disorders in which serotonin reuptake inhibitors, noradrenergic reuptake inhibitors, and serotonin-noradrenergic reuptake inhibitors were found to be equally effective, there are clear differences in cognitive functioning (i.e., decreased cognitive flexibility and inhibition in OCD but not in other anxiety disorders), OCD shows unique involvement of fronto-striatal circuitry unlike other anxiety disorders, and the male-female ratio in OCD is 1:1 as compared to 2-3:1 in other anxiety disorders. Removing OCD from the anxiety disorders raises the question of which other disorders in the DSM should be considered part of the same behavioral spectrum? The group reviewed other disorders in the DSM that had repetitive behaviors as a core feature in terms of phenomenology, co-morbidity, family history, frontal striatal circuitry, and treatment response. Using the requirement of positive evidences in at least 3 of these areas (with at least one being family history or brain circuitry), the group found the strongest evidence for including body dysmorphic disorder, hypochondriasis, and Tourette’s disorder in the spectrum, with less strong but still positive evidence for Obsessive-Compulsive Personality Disorder, Repetitive grooming disorders (e.g., trichotillomania), and Sydenham’s chorea. The group also suggested the consideration of a number of possible subtypes for OCD, including tic-related, childhood-onset, symmetry, hoarding, post-partum, poor insight, and contamination. Finally, the group also suggested the possibility of adopting a dimensional approach to OCD that involves two steps; the first step would be to make a categorical diagnosis of OCD and a second step would be to make dimensional ratings in terms of insight, impulsivity, tics/motor/sensory, reward sensitivity, attention, mood, anxiety, and social functioning.
Robert Krueger PhD (Minneapolis, MN) presented a summary of the dimensional approaches to psychiatric classification conference. The group recommended that the DSM-V would benefit from offering explicit criteria for both categories and dimensions, with the agenda being to supplement the categorical definitions with dimensions, and not to replace them. For any specific disorder, a number of specific aspects of the disorder could be conceptualized and assessed dimensionally. For example, dimensions could indicate the number of symptoms, the severity or duration of symptoms or the level of illness impairment and could potentially be uniform across disorders. Including explicit dimensional aspects of disorders in the DSM-V could result in standardization with benefits similar to those for DSM categories. Assessment instruments could accompany DSM-V for getting at the dimensional aspects. Recognition of cross-cutting spectrum dimensions in the DSM could also yield important benefits for research and practice (for example, internalizing, externalizing, psychotic spectra) which could help to conceptualize multi-morbid cases that carry much of the burden of mental illness. Dimensional definitions could encourage sensitivity to development, gender, and ethnicity by allowing thresholds for clinical significance to be adjusted for demographics. Clinicians already think dimensionally and will likely welcome explicit dimensional concepts and measurement-based care requires dimensional conceptualization and assessment. The DSM classifications have traditionally been constructed in a “top-down” fashion, i.e., experts define the disorders. There are many potential advantages to a bottom-up approach in which the structure of the classification is bootstrapped from the data. Analysis of fine-grained dimensional data from DSM-V could inform committee work on DSM-VI.
Joel Dimsdale MD (La Jolla, CA), presented a summary of the Somatic Presentation conference. There is a large somatic symptom burden which is commonly seen in medicine but less so in psychiatry. Although pure somatization disorder is rare, other somatoform disorders are more common but not well studied. Somatic symptoms are frequently associated with anxiety and depression and may possibly be associated with abuse. Recall of somatic symptoms is poor, complicating the making of a diagnosis. Presentations of somatoform disorders may be different in non-Western countries and diagnostic stability of somatoform disorders is similar to other diagnoses in the classification. There have been multiple studies of CBT and pharmacological interventions for somatoform disorders although there are almost no studies from the developing world. The participants in the conference expressed little enthusiasm for the way that somatoform disorders are currently diagnosed in DSM-IV and ICD-10 and suggested that a major revision is needed. Recommendations included obtaining stakeholder input when considering changes to the classification, particularly those of primary care practitioners. The group also recommended possibly reconsidering neurasthenia for inclusion in DSM-V, that hypochondriasis be considered in terms of its relationship to anxiety disorders and OCD, and finally that basing the diagnosis of somatoform symptoms on the presence of “medically unexplained symptoms” be abandoned in favor of reconceptualizing these disorders as part of a “somatic distress syndrome.”
David Shaffer, MD (New York, NY) presented a summary of the Externalizing Disorders of Childhood conference which covered Attention-deficit/Hyperactivity Disorder (ADHD), Conduct Disorder (CD), Oppositional-defiant Disorder ODD), and Juvenile Bipolar Disorder. ADHD issues included the validity of the current ADHD subtypes in the context of a tradition of internally reliable unidimensional scales; whether the high comorbidity rates between ODD and ADHD (80%) mean that ADHD and ODD should not be considered independent disorders, the status of adult ADHD (60-70% of child get better by adolescence or early adulthood, leaving a disproportionate amount of women with adult ADHD); concerns about ADHD criteria themselves (e.g., about the degree of cross-situationality that is required), and questions about the whether possible indicators of the underlying pathological processes can be useful in making a diagnosis (e.g., although ADHD is said to represent impairment of executive functioning, the base rate of executive functioning of unimpaired sibs is so high it cannot really be used for making a diagnosis). CD and ODD issues included clarifying the relationship between CD and ODD and between CD and antisocial personality disorder, reconsidering the use of early onset substance abuse as a CD criterion, revisiting the subtyping (age of onset vs. unsocialized), examining the extent to which the criteria are gender-biased (e.g., relational aggression and precocious sexuality were added to DSM-IV to boost prevalence in females but these may not have improved validity). Regarding juvenile presentations of bipolar disorder, there are a number of distinctive features, including that children are more likely to be irritable than grandiose, that retarded depression and psychosis is rare in children, that the course is chronic with few euthymic episodes (ultradian cycling is the rule), increased goal-directedness is uncommon (unlike in adults) and there is high comorbidity with CD, ADHD, and social phobia. The main nosological question raised at the conference was whether presentations of juvenile bipolar represented an age-specific and symptom-specific form of adult bipolar disorder, whether it is really a very severe form of ADHD, or whether it is a discrete and hitherto unrecognized disorder (e.g., severe mood dysregulation [SMD]). Future research suggested by the group includes conducting more basic phenomenological work regarding the definitions of irritability, episodicity, grandiosity and euphoria, piggy-backing studies (e.g., looking at the predictive value of early irritability) and the promotion of new research using provocation paradigms.
David Goldberg, DM (London, UK) summarized the recent conference on the overlap of depression and GAD and began by noting that the conference focused on two questions: are GAD and MDD different forms of the same disorder, closely related disorders, or distally related disorders, and should MDD and GAD be separate categories or in the same category as sub-types of the same disorder? There is considerable evidence regarding shared causal factors, including a shared correlation of genetic factors of 0.96 which is shared equally between “neuroticism” and other genes; families of each have a higher incidence of anxiety disorders; they share “negative emotionality” (i.e., neuroticism); parents of both show low care and high over-protection; and both are associated with parental neglect, sexual and physical abuse during childhood, periods of parental separation, elevated rates of personality disorders, and mothers with internalizing disorders. Furthermore, compared with either disorder on its own, comorbid cases of MDD and GAD have lower self-esteem, higher neuroticism in adolescence, periods of chronic illness before aged 5, parental divorce/separation before age 15, and have the highest rates of risk factors. Some causal and biological factors were found to be specific to either depression or anxiety; causal factors specific to depression include an additional familial aggregation component, the fact that episodes are more likely to be triggered by loss events and in some studies by “humiliation and entrapment” whereas causal factors specific to anxiety include higher rates of maternal internalizing, inhibited temperament, and more maltreatment during childhood and that episodes are likely to be triggered by threat and danger. Biological factors specific to depression include activation of the ventral cingulate and the insula, and de-activation of dorsal cortical regions, differences in the hypothalamo-pituitary-adrenal and the hypothalamo-thyroid axes in depression (but not in anxiety) and the fact that neuropeptide Y is decreased. Biological factors specific to anxiety include activation of the orbito-frontal cortex and de-activation of ventral regions and the fact that cholecystokinin provokes anxiety (but not depression). Dr. Goldberg summarized the opinions of the conference participants by noting that a minority thought that MDD and GAD were only distally-related, being impressed by the biological differences between depression and anxiety, and the (rather small) causal differences. The majority, however, felt that MDD and GAD were closely related, noting that the causal factors are almost identical, descriptive data are very close, symptom dimensions are closely related, and response to medication are very similar.
Overall recommendations of the conference were a strong reinforcement of the need to align and harmonize DSM-V and ICD-11. There was also unanimous recognition of the need to develop applications for primary care and other medical specialties that would be useful both for clinicians and other stakeholders. There is a need for the future classification to be user-friendly and for assessment instruments and appropriate glossaries to be developed. Consistent with this overriding recommendation, there is a need for the developers of ICD-11 and DSM-V to work with other medical specialties at certain diagnostic interface areas, such working with neurologists on the definition of dementia.
Papers based on these presentations will appear in a future monograph to be published by American Psychiatric Publishing, Inc.