Assessing the severity of borderline personality disorder

Assessing the severity of borderline personality disorder

Paul Moran, Mike J. Crawford
  • Declaration of interest

    P.M. has received honorarium payment from Roskilde University for speaking on the topic of personality disorder.


The identification of a reliable and valid severity index for borderline personality disorder has vexed researchers for decades. A simple, clinically intuitive severity index for borderline personality disorder with predictive validity has now been identified. This index could usefully guide treatment planning, but other contextual factors should also determine the need for specialist treatment.

The concept of clinical severity is intuitively appealing to clinicians and researchers. If robustly assessed, clinical severity can lead to more accurate predictions about the impact of a disorder on an individual’s quality of life, their level of functioning and ultimately their response, or lack of response, to treatment. In general medicine, measures of disease severity based on physiological indices are frequently used to inform treatment planning decisions. Unfortunately, psychiatry lags behind in this respect and measures of disease severity that can usefully guide treatment decisions have been identified with only limited success.1 In large part, this reflects fundamental failings in our understanding about the aetiology and classification of mental disorder. However much faith some may place in the identification of reliable biomarkers for mental disorder, increasingly these appear to be non-specific,2 and for the foreseeable future, we look set to remain reliant on a descriptive classification of mental disorder, with all the associated limitations. This is particularly problematic in the field of personality disorder treatment and research, where not only do we lack reliable biomarkers, but we also lack a coherent underpinning theoretical framework.

Severity and the classification of personality disorder

There is general consensus that although the construct of personality disorder is clinically indispensible, the current classification of personality disorder is not fit for purpose. Personality disorders represent the tail end of continuous dimensional personality traits that are normally distributed in the population. At the extreme end of this spectrum these traits are maladaptive and associated with severe problems with interpersonal functioning, poor health and increased early mortality.3 Diagnostic comorbidity is the rule rather than the exception; some categories of personality disorder are rarely used in clinical practice, whereas one of the most prevalent categories is ‘not otherwise specified’. In addition, the current ICD-10 and previous DSM-IV classification systems for personality disorder fail to capture the wide heterogeneity in social functioning experienced by individuals with personality disorder. All of these limitations have stimulated vigorous attempts to improve classification. Yet, after much heated debate, the classification of personality disorder in DSM-5 is largely unchanged. In contrast, it seems likely that personality disorders in ICD-11 will be classified primarily on the basis of severity.4

The severity of borderline personality disorder

Of all the personality disorder subtypes, borderline personality disorder is the most thoroughly researched category. The public health impact of this disorder is now firmly established. The condition occurs globally, with a community prevalence which is at least comparable to that for schizophrenia. Social impairment is an enduring feature of the disorder, with less than 50% of patients achieving stable recovery over 10 years and it is estimated that approximately 10% of individuals die by suicide. Over the past 10 years, a body of evidence has emerged to show that psychological treatments can effectively relieve some of the symptoms of the disorder.5 Despite these developments, we know very little about which people with borderline personality disorder may benefit most from specialist treatment; in short, we currently have no reliable evidence-based indicators to guide treatment planning. This is a significant problem given that there is limited access to psychological treatments in most countries, especially for the type of intensive and lengthy treatment programmes that are recommended for people with borderline personality disorder. The need for evidence about which patients with the disorder benefit most from specialist treatment is even more pressing in England, where major changes in commissioning arrangements for the management of personality disorder have been proposed.6 Treatment for personality disorder is likely to be increasingly focused in primary care and yet there is very little information to guide general practitioners as to which care pathway a patient with borderline personality disorder should be directed. Indeed, currently, personality status is not even routinely assessed in primary care, although the advent of rapid and effective screens for personality disorder7 means that this situation is eminently remediable.

In this issue, Bateman & Fonagy provide us with some much needed data about who among those with borderline personality disorder may benefit most from specialist treatment.8 In a secondary analysis of data from their randomised controlled trial of out-patient mentalisation-based treatment (MBT) v. structured clinical management (SCM) for individuals with borderline personality disorder, they examined the impact of clinical severity defined in four ways on clinical outcome. The indicators of severity that the authors selected concur with those identified in one review,9 and included the number of comorbid Axis 1 diagnoses, the number of borderline personality disorder criteria, the number of comorbid Axis II diagnoses and the severity of symptom distress. Surprisingly, there was little redundancy between these severity indicators, with only the number of Axis I and Axis II diagnoses being significantly correlated. None of the severity indicators predicted outcome at the end of treatment. However, patterns of recovery in the two treatment groups were significantly associated with both the baseline level of distress and the number of Axis II diagnoses. Only a quarter of patients with three or more Axis II diagnoses recovered in the SCM group, compared with nearly three-quarters of those in the MBT group. Patients with a larger number of Axis II diagnoses also recovered at a slower rate. Indeed, the rate of recovery diminished with each additional Axis II diagnosis for the SCM group, and remaining essentially unaltered for the MBT group. The deleterious impact of Axis II comorbidity on recovery emerged particularly in relation to self-harm, and MBT was more effective than SCM for individuals with a higher number of Axis II diagnoses in reducing self-harm. These effects did not appear to be explained by confounding. A relatively straightforward algorithm for rating the severity of personality disturbance in the general population, based on the DSM clustering system, has been shown to correlate strongly at a cross-sectional level with measures of social dysfunction.10 Now, based on their prospective data, Bateman & Fonagy appear to have identified an equally simple but clinically intuitive index for the severity of borderline personality disorder that could usefully guide treatment planning.

Strengths of the study are the prospective nature of the data, the measurement of a wide range of covariates assessed using standardised assessments and the use of a conservative alpha to minimise the risk of Type 1 error. Notwithstanding, the study relied on the significance of 3-way interaction terms (severity × group × time) for which the original trial was not powered to detect. It is therefore entirely possible that the other postulated severity indices may have predictive validity and these deserve further examination in a larger, appropriately powered longitudinal study.

Some cautionary notes

The identification of a severity index for borderline personality disorder is laudable not least because ‘stratified medicine is regarded as central to the progress of healthcare’.11 However, there is potential for the misuse of such an index. Bateman & Fonagy are wisely cautious in discussing their findings and it would be premature to currently advise referral to a specialist service based purely on the number of Axis II diagnoses that a patient meets. What about the patient with one Axis II diagnosis who has attempted murder or made a serious suicide attempt? No one would dispute the potential need for specialist input in such cases. Risk of harm to self and others has been used in an effort to develop a clinically helpful definition of the severity of other mental disorders,12 and we believe there is a case for using information about risk in judging the severity of personality disorder.

Finally, although these data appear to show that it is people with more complex personality problems that respond best to specialist treatment, it is important to note that the data are derived from a trial of patients treated in groups with a range of severity of disturbance. The composition of groups can have an important bearing on the outcome of group-based treatment,13 and it is possible that part of the benefit that people with severe personality disorder gain from groups is derived from interaction with people with similar but less severe problems. Restricting access to MBT or other group-based treatments for people with severe personality disorder may make groups more difficult to run and limit the opportunities that group members have to benefit from them. Greater understanding of the impact of context as well as content of treatments for people with personality disorder is needed to ensure that in the future, limited resources are used to maximum effect.


  • Funding

    P.M. is part-funded by the National Institute for Health Research (NIHR) Biomedical Research Centre for Mental Health at South London and Maudsley NHS Foundation Trust and Institute of Psychiatry, King’s College London. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.

  • Received April 18, 2013.
  • Accepted June 25, 2013.


Article Information

vol. 203 no. 3 163-164

Print ISSN
Online ISSN
  • Received 18 April 2013
  • Accepted 25 June 2013
  • Published online 2 September 2013.

Author Information

  1. Paul Moran and
  2. Mike J. Crawford

  1. Paul Moran, MD, Health Services and Population Research Department, King’s College London, Institute of Psychiatry, London; Mike J. Crawford, MD, Centre for Mental Health, Department of Medicine, Imperial College London, UK
  1. Correspondence: Dr Paul Moran, PO28, David Goldberg Building, Institute of Psychiatry, King’s College London, De Crespigny Park, London SE5 8AF, UK. Email:
  • Paul Moran (pictured) is Reader in Personality Disorder Research at the Institute of Psychiatry, King’s College London. Mike J. Crawford is Professor of Mental Health Research and Head of the Centre for Mental Health at Imperial College London.

  • See pp. 221–227, this issue.

  • Declaration of interest

    P.M. has received honorarium payment from Roskilde University for speaking on the topic of personality disorder.

Leave a Reply

%d bloggers like this: