DSM-V perspectives on disentangling disability from clinical significance

Üstün and Kennedy provide a cogent argument for the DSM to reflect a clearer conceptualization of mental disorders and associated impairments, activity limitations, and participation restrictions.

Setting thresholds for caseness has long been important to researchers and policy makers who have interests in distinguishing people with mental disorders from those who are suffering transient symptoms and expectable reactions to the stresses of everyday life. The ubiquity of the latter conditions makes this distinction of crucial importance for the identification of homogeneous research samples and for the allocation of scarce resources for mental health services. Clinicians, particularly those in primary care settings, are also frequently confronted with routine emotional and behavioral complaints that raise the question “treat or don’t treat?”. The importance of accounting for distress and impaired functioning in daily activities when determining caseness has always been an underlying concern for clinical and health policy decisions, although not one clearly articulated.

The DSM definition of “mental disorder” has contained the concepts of distress and limitations in activities since the release of DSM-III. DSM-III also implicitly dictated that careful specification of symptom criteria for each disorder would suffice in establishing a disorder threshold; that is, the combinations of symptoms specified for each disorder would be inherently distressing or disabling. However, higher than expected rates of people with disorders were found in community studies, leading to concerns about a “false positive” problem with the symptom criteria 1.

The solution arrived at by the developers of DSM-IV was to add a “clinical significance criterion” to many DSM mental disorder criterion sets. This criterion specified that the person with a mental disorder had to display clinically significant distress or impairment in social, occupational, or other important spheres of daily functioning. As noted by Üstün and Kennedy, this path was not taken by the developers of the mental disorders chapter of ICD-10, who produced the International Classification of Functioning, Disability and Health (ICF, 2) to classify disabilities, with the intention of keeping the symptom syndrome and the associated activity limitations separate.

Subsequent research did show that clinical significance-like specifiers reduced the community rates of DSM mental disorders and identified persons more likely to be using mental health services or having more severe symptoms 3. However, the problems with the DSM’s handling of disabilities, including the clinical significance criterion and the Global Assessment of Functioning (GAF), were many. These problems are well stated by Üstün and Kennedy. First, the concept of “clinically significant distress or impairment in functioning” is not defined and the terminology is not consistent with the World Health Organization (WHO) standard. The GAF is not a “clean” scale, mixing as it does symptom severity, social functioning, and assessments of dangerousness. The DSM symptom criteria themselves are confounded with activity limitations, apart from the clinical significance criterion. Finally, the operationalization of syndrome severity is inadequate. These limitations have substantial consequences for research and clinical assessment of mental disorders.

In our own criticisms of DSM-IV, we are in substantial agreement with Üstün and Kennedy. We are also in agreement that the problems need to be fixed, and the DSM-V Work Groups are taking steps to do so. A study group has been formed to address the problems and to identify possibilities for restructuring the forthcoming DSM-V to create greater consistency with ICD-11 and the ICF. We are moving to standardize our terminology and better define and operationalize the concepts of severity, disability, distress, and so forth. This could require a significant re-orienting of clinicians – American clinicians, specifically – to a different way of thinking about mental disorders.

Although the ICF is an official WHO classification, its usage in the United States is limited, and it has not been adopted as an official code set. Apart from the widespread unfamiliarity with ICF terminology, the complexity of the classification system may prove daunting. Specification of key domains of activity limitations for persons with mental disorders, with a corresponding global assessment tool, as suggested by the authors, would go a long way in promoting acceptance of disability assessment in the DSM.

We are also taking steps to separate activity limitations from symptom descriptions. Consistent with the authors’ suggested solutions, the incorporation of dimensional measures of symptoms into the diagnostic assessment process will help differentiate symptom/syndrome severity from disability by focusing specifically on symptom ratings of frequency, intensity, and/or duration. The Patient Health Questionnaire (PHQ-9) 4 for depression, for example, is a brief empirically-validated, DSM-IV-derived measure that has demonstrated good acceptability in primary care and psychiatric settings for assessing diagnostic threshold, planning treatment, and tracking outcomes 5. Similar measures have been developed for anxiety and somatic complaints. Planned development of DSM-V clinical diagnostic interviews and lay interviews for epidemiologic surveys will also need to attend to improving assessments of symptom severity and disability.

However, this exercise is somewhat dependent on the extent to which the symptom criteria themselves are already liberated from elements that are better seen as activity limitations. Contrary to Üstün and Kennedy’s assertion, this is not a minor issue limited to occasional disorders such as residual schizophrenia. The DSM-IV and the ICD-10 criteria for research both have many examples of activity limitations serving as symptoms. In the ICD-10, personality disorders, substance dependence, hyperkinetic disorders, and conduct disorder all have substantial components of activity limitations in their symptom criteria. It is unclear at this point whether symptom criteria for these disorders can be fully “cleansed” of their activity limitations components. Ideally, more specific assessments would target the impaired mental processes that underlie these activity limitations. The development and implementation of such assessments will depend largely on the state of the existing science and technology as well as the practical limitations of implemention in routine clinical settings.

The views of Üstün and Kennedy are neither radical nor revolutionary, but reasonable. They reflect the progress in our field and continued efforts to unify psychiatry in the United States with all of medicine and with the rest of the world. We look forward to further cooperation with WHO as the next steps in the development process are taken.