This chapter describes the context of the 11th Revision of the International Classification of Diseases (ICD-11) related to mental health. It contains an explanation of the procedure adopted in making this revision, some background to the field trials and their results, and a brief account of the main changes, many of which are amplified in the later chapters. A detailed account of the changes in the ICD-11 as compared with the ICD-10 has been published elsewhere,Reference Reed, First and Kogan¹ as has a detailed comparison of the ICD-11 and the DSM-5.Reference First, Gaebel and Maj²

The context of the development of the ICD-11 is significant. This is the first major revision of the ICD in thirty years and has followed a thorough re-examination of each ICD-10 diagnosis in light of new scientific findings, best practices, and advances in information technology for health systems. The revision was approved by the World Health Assembly on 25 May 2019 and was formally implemented as a basis for health reporting by WHO member states from January 2022. Over the next few years, WHO member states will implement the ICD-11 within their clinical and health information systems. WHO has published a range of materials intended to be useful to countries in implementing the ICD-11.³ In some systems, implementation will happen quickly and in others clinical implementation will precede full data integration. For example, Scotland has already begun the implementation of the ICD-11 classification of mental, behavioural or neurodevelopmental disorders. This will make it possible for Scottish clinicians to benefit from the more than three decades of scientific and clinical advances reflected in the ICD-11, even if their systems are still collecting data using the ICD-10 as a framework.

The Development of the ICD-11 Classification of Mental Disorders

The ICD-11 has been developed incrementally over a Fifteen-year period. The basic structure of the ICD-11 chapter on Mental, Behavioural and Neurodevelopmental Disorders (MBND) and the brief descriptions for each disorder have been online and available for review, comment, and proposals for changesReference Fuss, Lemay and Stein⁴ since 2014 (https://icd.who.int/dev11/l-m/en). The World Health Organization Department of Mental Health and Substance Use (MSD) has also developed Clinical Descriptions and Diagnostic Requirements (CDDR) for ICD-11, which are available online at https://icd.who.int/browse11/l-m/en#/ and will also be published as a book. The structure of the ICD-11 MBND classification, the category names and brief descriptions for statistical use, and the detailed diagnostic guidance for clinical implementation contained in the CDDR were developed simultaneously by seventeen expert working groups in different areas appointed by MSD. Each group included experts from all WHO regions and substantial representation of low- and middle-income countries. Working groups were responsible for reviewing the available evidence related to their areas of responsibility, including the overlapping work on the development of the DSM-5.

In making recommendations for the ICD-11, working groups were asked to consider clinical utility and global applicabilityReference Reed, First and Kogan¹ in addition to the validity of proposed changes. Classification is the interface between health encounters and health information. If a new diagnostic system fails to provide clinicians with enough useful information that is feasible to implement given the time and resources available to them, it is unlikely to be applied consistently and faithfully. This will have implications for the overall data used for evaluation and decision making at the system, local, national, and global levels. A more clinically useful system therefore contributes to better health data. Because of the need for global applicability, the ICD-11 MBND revision was tested via a systematic programme of global field studies. The working groups included experts from all global regions, with particular attention to the representation of low- and middle-income countries. Hundreds of global experts were involved in developing the CDDR and thousands of global clinicians were involved in testing it across the world in multiple languages, as described below.

The CDDR is designed to provide sufficient and clinically useful information to enable psychiatrists and other diagnosing health professionals to consistently and accurately apply the ICD-11 MBND classification to make diagnoses in clinical settings.Reference Reed, First and Kogan¹ The sections of the CDDR follow a uniform structure,Reference First, Reed, Hyman and Saxena⁵ which has been a major improvement over the equivalent volume for ICD-10, the Clinical Descriptions and Diagnostic Guidelines (CDDG) for ICD-10 Mental and Behavioural Disorders.⁶ Each of the main disorder entries for the ICD-11 CDDR includes the following sections: 1) essential (required) features; 2) additional clinical features; 3) boundary with normality (threshold); 4) course features; 5) developmental presentations; 6) culture-related features; 7) sex and/or gender-related features; and 8) boundaries with other disorders and conditions (differential diagnosis).Reference First, Reed, Hyman and Saxena⁵

The essential features present briefly the characteristics of the disorder in descriptive terms.Reference First, Reed, Hyman and Saxena⁵ They represent the clinical features that a clinician could reasonably expect to see in all cases of the disorder. In this way, they resemble diagnostic criteria in the DSM. The ICD-11 differs from the DSM-5, however, in avoiding algorithmic pseudoprecision in terms of symptom counts or precise durations unless these are well established and empirically based. (For example, in ICD-11 five of ten possible symptoms of a depressive episode must be present, one of which must be depressed mood or anhedonia; two of seven psychotic symptoms are required for a diagnosis of schizophrenia, etc.) The ICD-11 essential features are stated more flexibly than DSM-5 diagnostic criteria in order to focus on the clinical essence of the syndrome and allow sufficient room for cultural variability and informed clinical judgement to enhance global applicability. But the idea that there are no diagnostic requirements in the ICD-11 CDDR is obviously false to anyone who has actually looked at them, and it is important to stress that those making such claims are misinformed.

To take one specific example, the essential features of the ICD-11 diagnosis of PTSD can be summarized as followsReference Maercker, Brewin and Bryant⁷ (see CDDR, https://icd.who.int/browse11/l-m/en#/, for the complete version):

• 
  

  Exposure to an event or situation (either short- or long-lasting) of an extremely threatening or horrific nature.

  
  
• 
  

  Following the traumatic event or situation, the development of a characteristic syndrome lasting for at least several weeks, consisting of three core elements:

  
  
  
  
  1. 
     

     1. Re-experiencing the traumatic event in the present, in which the event(s) is not just remembered but is experienced as occurring again in the here and now.

     
     
  2. 
     

     2. Deliberate avoidance of reminders likely to produce re-experiencing of the traumatic event(s).

     
     
  3. 
     

     3. Persistent perceptions of heightened current threat.

  
  
  
  
• 
  

  The disturbance causes significant impairment in functioning.

Diagnostic requirements are clearly stated, and all must be present. At the same time, the ICD-11 essential features for PTSD are vastly simpler than the diagnostic criteria in the DSM-5, which include 20 different symptoms in four different groups, as well a list of specific experiences that ‘qualify’ for a diagnosis that appear to be largely based on US liability concerns. It has been calculated that there are 636,120 different combinations of symptomsReference Galatzer-Levy and Bryant⁸ that would qualify for a PTSD diagnosis under DSM-5.

Some have expressed concern that ICD-11’s more flexible approach to diagnostic requirements would result in overdiagnosis and inflated prevalence rates, but there is no evidence to support this claim. Using World Mental Health Survey data, Stein et al.Reference Stein, McLaughlin and Koenen⁹ found that the ICD-11 diagnostic requirements resulted in fewer diagnoses of PTSD compared with ICD-10, and comparable rates compared with DSM-5. Lago et al.Reference Lago, Bruno and Degenhardt¹⁰ also applied diagnostic requirement in the major classifications to Mental Health Survey data for disorders due to use of alcohol and disorders due to use of cannabis. They found almost perfect concordance among ICD-11, DSM-IV, and ICD-10, but much lower concordance with DSM-5. Evans et al.Reference Evans, Roberts and Keeley¹¹ found that, compared with ICD-10 and DSM-5, the ICD-11 CDDR led to more accurate identification of severe irritability and better differentiation from boundary presentations. Participants using the DSM-5 were more likely to assign psychopathological diagnoses to developmentally normative irritability. Although relatively few head-to-head comparisons of the ICD-11 and DSM-5 have been conducted, of those that have, none has found higher rates of diagnoses using the ICD-11.

Field Testing of ICD-11 MBND

Another major area of innovation has been the extensive and systematic programme of field studies supporting the ICD-11 MBND classification and its associated CDDR.Reference Reed, First and Kogan^1, Reference Keeley, Reed and Roberts^12, Reference Evans, Roberts and Keeley¹³ Twenty internet-based case-controlled studies have been conducted using the Global Clinical Practice Network (GCPN; https://gcp.network). The GCPN is a network of over 18,500 mental health and primary care professionals from 163 countries who took part in the development of the ICD-11 through participation in field studies. Slightly more than half of GCPN members are physicians – almost all of these psychiatrists – with a third being psychologists, and the rest a mixture of other mental health disciplines. Thirty-seven per cent are working in low- and middle-income countries. GCPN studies have been conducted in a minimum of three and up to six languages: Chinese, English, French, Japanese, Russian, and Spanish. Specific studies were also conducted in German.

Case-controlled studies for ICD-11 most commonly involved participants being randomly assigned to use the ICD-11 CDDR or the ICD-10 CDDG to assign diagnoses to standardized, validated clinical case vignettes that had been manipulated to highlight key diagnostic issues.Reference Keeley, Reed and Roberts^12, Reference Evans, Roberts and Keeley¹³ The studies compared the accuracy and consistency of diagnostic judgements based on the two systems. Across these studies, ICD-11 consistently outperformed the ICD-10.Reference Claudino, Pike and Hay^14, Reference Kogan, Stein and Rebello¹⁵ The methodology also permitted an examination of which specific diagnostic elements were accounting for any observed confusion, which in turn permitted refinements in the CDDR before they were finalized.Reference Keeley, Reed and Roberts^16, Reference Reed, Sharan and Rebello¹⁷

Clinic-based studies of the reliability and clinical utility of the ICD-11 CDDR have been conducted in 14 countries covering all global regions.Reference Reed, Sharan and Rebello^18, Reference Chmielewski, Clark and Bagby¹⁹ These studies focused on mental disorders accounting for the greatest proportion of global disease burden and the highest levels of service utilization – schizophrenia or other primary psychotic disorders, mood disorders, anxiety or fear-related disorders, and disorders specifically associated with stress – among adult patients presenting for treatment at 29 participating centres. A concurrent joint-rater design was used, examining whether two clinicians, relying on the same clinical information, agreed on the diagnosis when separately applying the ICD-11 CDDR. Intraclass kappa coefficients for diagnoses weighted by site and study prevalence ranged from 0.45 (dysthymic disorder) to 0.88 (social anxiety disorder) and would be considered moderate to almost perfect for all diagnoses.Reference Reed, Sharan and Rebello¹⁷ Overall, the reliability of the ICD-11 CDDR was superior to that previously reported for equivalent ICD-10 guidelines. Clinician ratings of the clinical utility of the proposed ICD-11 diagnostic guidelines were very positive overall.Reference Reed, Sharan and Rebello¹⁸ The CDDR were perceived as easy to use, corresponding accurately to patients’ presentations (i.e., goodness of fit), clear and understandable, providing an appropriate level of detail, taking about the same or less time than clinicians’ usual practice, and providing useful guidance about distinguishing disorder from normality and from other disorders.

The reliability results from the clinic-based studies challenge the claim that some have put forward that the more clinician-friendly, less concretely algorithmic, and less highly operationalized approach adopted for the ICD-11 CDDR is inherently less reliable. The concern that the ICD-11 CDDR has sacrificed reliability is not based on any data, but rather based on assumptions that have been built into the DSM since DSM-III, including the assumptions that clinicians apply the criteria as they are written, which we do not believe is the case. In our clinic-based studies, clinicians with diverse training and experience used the ICD-11 CDDR following a relatively brief training (about 4 hours) to conduct routine clinical assessments (lasting about 1 hour) using open form interviews. They obtained reliability coefficients similar to those achieved using more complex and time-consuming structured instruments.Reference Chmielewski, Clark and Bagby^19–Reference Lobbestael, Leurgans and Arntz²¹ It is possible that further gains in reliability among clinicians could be obtained by focusing greater attention on appropriate training in diagnostic skills and interviewing techniques, rather than on continuing to devote attention and resources to introducing greater precision in operationalization as a part of successive refinements in diagnostic criteria.

New Disorder Categories

Twenty-three new disorders have been added to the ICD-11 MBND chapter (see Table 1.1), reflecting either a distinct disorder that was not classifiable in the ICD-10 (e.g., Hoarding Disorder), or a disorder that is a result of extending, expanding, or subdividing an existing disorder in such a way that has resulted in a new disorder rather than a subtype (e.g., Binge Eating Disorder).Reference Reed, First and Billieux²² Most of these were either already in the DSM-IV or added to the DSM-5. The effect of adding these categories has therefore generally been to enhance compatibility between the ICD-11 and the DSM-5.

The most consequential additions are arguably four new disorders in the ICD-11 that represent different decisions than were taken for the DSM-5. These are Complex PTSD, Prolonged Grief Disorder, Gaming Disorder, and Compulsive Sexual Behaviour Disorder, although Prolonged Grief Disorder has since been added to the DSM 5.1 and Internet Gaming Disorder appears in the DSM-5 and DSM 5.1 research appendix. We have published a detailed review of the rationale and consequences of adding these four disorders,Reference Reed, First and Billieux²² concluding that each describes an important and distinctive clinical population that is an appropriate focus of health services and with specific treatment needs that would otherwise likely go unmet. WHO’s announced intention to include these categories has clearly facilitated an expansion of research in each area, which has generally supported their validity and utility, as well as increased availability of appropriate services.

Complex PTSD

The essential or required features of complex PTSD include all three core symptoms of PTSD (re-experiencing in the present, avoidance, and ongoing sense of threat). Additional features of complex PTSD include three characteristic types of disturbances in self-organization: severe and persistent problems in affect regulation; beliefs about the self as diminished, defeated, or worthless; and difficulties in sustaining relationships and in feeling close to others.Reference Maercker, Brewin and Bryant^7, Reference Reed, First and Billieux²² Complex PTSD is more likely to be the product of certain types of traumas, such as prolonged or repetitive events from which escape is difficult or impossible (e.g., torture, slavery, prolonged domestic violence, repeated childhood sexual or physical abuse), and typically requires longer and more complex treatments than does PTSD. However, treatments for complex PTSD are not typically as long or as complex as evidence-based treatments for borderline personality disorder. Emerging evidence indicates that complex PTSD and borderline personality disorder are quite distinct, having only the feature of affect dysregulation in common.Reference Reed, First and Billieux^22–Reference Frost, Murphy and Hyland²⁴