Progress in developing a classification of personality disorders for ICD‐11

In appointing a Working Group charged with developing recommendations in the area of personality disorders (PDs) for the ICD‐11, the World Health Organization (WHO) Department of Mental Health and Substance Abuse highlighted several problems with the classification of PDs in the ICD‐10.

First, PDs appeared to be substantially underdiagnosed relative to their prevalence among individuals with other mental disorders. Second, of the ten specific PDs, only two (emotionally unstable personality disorder, borderline type and dissocial personality disorder) were recorded with any frequency in publicly available databases. Third, rates of co‐occurrence were extremely high, with most individuals with severe disorders meeting the requirements for multiple PDs. Fourth, the typical description of PD as persistent across many years was inconsistent with available evidence about its lack of temporal stability.

The WHO, therefore, asked the Working Group to consider changes in the basic conceptualization of PDs and specifically to explore the utility and feasibility of a dimensional approach. At the same time, the WHO emphasized that any classification system of PDs for the ICD‐11 must be usable and useful for health care workers in lower‐resource settings who are not highly trained specialist mental health professionals1.

The Working Group, under the leadership of P. Tyrer, took the WHO's requests very seriously in developing its proposal for ICD‐11. PD was conceptualized in terms of a general dimension of severity, continuous with normal personality variation and sub‐threshold personality difficulty. After meeting the general requirements for a diagnosis of PD, an individual would be assigned a mild, moderate or severe PD diagnosis, based primarily on the extent of interpersonal dysfunction and the risk of harm to self or others. The ICD‐10 specific PDs were abandoned entirely in favour of five broad trait domains grounded in the scientific literature on personality2: negative affectivity, disinhibition, detachment, dissociality and anankastia.

Descriptions of the Working Group proposal were subsequently published in specialty and more general scientific journals3, 4. It should be noted that, although the essence of the ICD‐11 proposal was conceptually compatible with what came to be the “alternative model” of PD diagnosis in the DSM‐5, the Working Group recommended against adoption of that model for ICD‐11 because it was seen as too complicated for implementation in most clinical settings around the world.

The WHO became aware of significant concerns among some members of the practice community and some PD researchers about various aspects of the proposal. This led to a meeting of the WHO with representatives from the European Society for the Study of Personality Disorders (ESSPD), the International Society for the Study of Personality Disorders (ISPPD), and the North American Society for the Study of Personality Disorders (NASSPD). A description of the concerns of members of the leadership of these organizations about the original Working Group proposal has recently been published5, although these concerns were not universal6. Nevertheless, the WHO believed it was important to attempt to engage a process that would help to avoid further divisiveness and acrimony in this area.

The WHO thus convened a Task Group consisting of members appointed by ISSPD/ESSPD/NASSPD and members of the original Working Group, which was asked to develop recommendations for responding to the concerns. Through discussions over several months, it became clear that the ISSPD/ESSPD/NASSPD representatives were willing to accept a dimensional model of PDs, but felt that the one that had been proposed provided insufficient information about the nature of individual personality disturbance to support case conceptualization, treatment selection, and management.

The other major issue to be addressed was the diagnostic status of borderline PD. Some research suggests that borderline PD is not an independently valid category, but rather a heterogeneous marker for PD severity7, 8. Other researchers view borderline PD as a valid and distinct clinical entity, and claim that 50 years of research support the validity of the category9. Many – though by no means all – clinicians appear to be aligned with the latter position. In the absence of more definitive data, there seemed to be little hope of accommodating these opposing views. However, the WHO took seriously the concerns being expressed that access to services for patients with borderline PD, which has increasingly been achieved in some countries based on arguments of treatment efficacy, might be seriously undermined.

In September 2017, the Task Group held a face‐to‐face meeting in Heidelberg, Germany, with the leadership and support of S.C. Herpertz, then ISSPD President. The purpose of the meeting was to develop specific proposals for modifications to the ICD‐11 guidelines that would address the issues of concern. The main recommended changes were as follows:

• Systematic incorporation of self functioning in the core diagnostic guidelines for PD. PD is conceptualized as an enduring disturbance characterized by problems in functioning of aspects of the self (e.g., identity, self‐worth, accuracy of self‐view, self‐direction) and/or interpersonal dysfunction.

• A substantially richer and more clinically informative operationalization of PD severity. The degree and pervasiveness of disturbances in functioning of aspects of the self; of interpersonal dysfunction across various contexts and relationships (e.g., romantic relationships, school/work, parent‐child, family, friendships, peer contexts); of emotional, cognitive and behavioural manifestations of the personality dysfunction; as well as of associated distress or functional impairment should be considered in making a severity determination for individuals who meet the general diagnostic requirements for PD.

• A substantially richer and more clinically informative operationalization of trait qualifiers. Each should describe the core feature of the trait domain, followed by a description of the common manifestations of that domain in individuals with PD.

• A complete description of PD includes the severity rating and the applicable trait domain qualifiers. The WHO acknowledges that it will not be feasible to conduct such a complete evaluation in all settings.

• Provision of an optional qualifier for “borderline pattern”. This qualifier may enhance clinical utility by facilitating the identification of individuals who may respond to certain psychotherapeutic treatments. Whether it will provide information that is non‐redundant with the trait domain qualifiers is an empirical question.

A revision of the diagnostic guidelines for PDs based on the above recommendations has been approved by the ICD‐11 Working Group and the ISSPD/ESSPD/NASSPD representatives. These guidelines are available for review and comment at http://gcp.network, and are now being used in field testing.

Geoffrey M. Reed
Department of Mental Health and Substance Abuse, World Health Organization, Geneva, Switzerland