Psychiatric diagnosis in ICD-11: Lessons learned (or not) from the mood disorders section in DSM-5

ICD Insights

Even before its publication in May 2013, the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) had been criticized by a large number of leaders in our field for a myriad of reasons: using phenotypic categories with no basis in biology (virtually the entire manual); medicalizing human conditions that should not be lumped with psychiatric pathologies (such as eliminating the bereavement criteria in diagnosing major depression); and forcing into categories what might be better conceptualized along dimensional lines (e.g. personality disorders). Others have bemoaned the missed opportunity to shift boundaries between disorders (expanding the boundaries of bipolar disorder at the expense of major depression). There is basis for all of these criticisms. Yet, finding a better diagnostic system – with greater validity, not just reliability – may prove just as difficult for the International Classification of Diseases, 11th Revision (ICD-11) as it was for the authors of DSM-5. We are limited by our ignorance on two major issues: (1) the biological underpinnings and proper boundaries of psychiatric disorders; and (2) how to set the proper balance between being inclusive enough to ensure the proper treatment of suffering individuals while not overpathologizing human conditions. A few examples from the mood disorders section will be illustrative.

Proposals for expanding the boundaries of bipolar disorder

Proposals for bipolar disorder focused on: decreasing the time criterion for hypomania from 4 to 2 days; and/or defining hypomania using increased energy as the core A criterion (Angst et al., 2013) (as opposed to a mood criterion alone or mood plus energy as in DSM-5). The justification for these proposals was to ensure that patients with bipolar spectrum disorders – those who would not have been diagnosed as bipolar in DSM-IV – are both accurately diagnosed and then properly treated. DSM-5 rejected both of these proposals for a number of reasons. (1) Shifting the epidemiology of mood disorders towards bipolar versus unipolar diagnoses will simply change the number of false-positive bipolar diagnoses at the expense of false-negatives (Zimmerman, 2012). Inherent in this logic is the notion that there is no intrinsic advantage in shifting the diagnoses of those with unipolar depression and/or borderline personality disorder to a bipolar diagnosis in the absence of validation of the proper diagnosis. (2) There is a lack of controlled studies demonstrating that patients in this broader bipolar spectrum are more effectively treated by mood stabilizers versus other agents used for major depression or borderline personality disorder, thereby making the treatment implications of expanding the bipolar spectrum unknown.

In contrast, DSM-5 properly expanded the boundaries of mixed mood pathology, both by making it easier to diagnose mixed mania and by introducing depression with a mixed mania specifier. The latter category was introduced on the basis of many studies demonstrating that this mixed depression subgroup had other features (such as family history of mania and a greater likelihood of antidepressant-induced switches) compared to those with classic unipolar depression. Yet, the specific manic features used for the mixed depression specifier have been criticized (Malhi, 2013), especially the exclusion of irritability and agitation. As with the diagnostic boundaries of bipolar disorder noted above, however, the lack of controlled studies examining the treatment responses of patients with mixed depression makes the clinical utility of this new DSM-5 diagnosis unknown.

DSM-5 includes a new, controversial category, disruptive mood dysregulation disorder (DMDD), to characterize children with severe and recurrent temper outbursts along with persistent irritability who do not meet criteria for a manic or hypomanic episode. The intent of this category is to reduce false-positive diagnoses of bipolar disorder in children; chronically irritable children later develop depression, but not mania (Leibenluft, 2011). Others argue that this category introduces new problems (e.g. Axelson et al., 2011). DMDD has a high overlap with oppositional defiant disorder, and there are no known treatments for it. Perhaps most importantly, the relevant research supporting inclusion of DMDD is based on severe mood dysregulation (Leibenluft, 2011), a childhood disorder that includes hyperarousal and other attention deficit hyperactivity disorder (ADHD)-like symptoms, as well as temper outbursts (e.g. Axelson et al., 2011). Here, too, whether this new category will advance diagnostic clarity and/or more appropriate treatment is unknown.

Conclusions

In the absence of clear evidence for the validity of diagnostic categories, all diagnostic systems including DSM-5 and ICD-11 will be inherently imperfect creations with compromises based on our field’s ignorance, not necessarily willful thoughtlessness. Alternatively, the US National Institute of Mental Health has proposed an entirely different system of classification, the Research Domain Criteria, consisting of five behavioral dimensions (e.g. arousal/modulatory systems) for which the underlying neural circuitry has been articulated (Cuthbert and Insel, 2013). Whether this will be an improvement upon the symptom-based systems of DSM-5 or ICD-11 will be unclear for many years. Until we have firmer answers to these questions, the key in optimally using DSM-5 or ICD-11 is for clinicians and researchers alike to treat these diagnostic systems as crude approximations based on best current guesses, and not to reify and concretize these criteria or to treat them as revealed truths.