Prevalence of personality disorder in the case-load of an inner-city assertive outreach team

AIMS AND METHOD The aim was to record the prevalence, type and severity of personality 
disorder dealt with by an inner-city outreach team. Patients on the register 
of an assertive outreach team were approached and asked to give informed 
consent for an informant interview with their principal worker to determine 
their personality status, using the informant-based ICD–10 version of 
the Personality Assessment Schedule. RESULTS Of the 73 patients, 62 (85%) of whom had a psychotic diagnosis, 67 (92%) 
had at least one personality disorder, with 37 (51%) having complex or severe 
personality disorders. CLINICAL IMPLICATIONS The findings suggest that the National Service Framework requirements for 
assertive outreach teams tend to select many patients with comorbid 
personality disorder in addition to other severe psychiatric disorders.

The aim was to record the prevalence, type and severity of personality disorder dealt with by an inner-city outreach team. Patients on the register of an assertive outreach team were approached and asked to give informed consent for an informant interview with their principal worker to determine their personality status, using the informant-based ICD-10 version of the PersonalityAssessment Schedule.

CLINICAL IMPLICATIONS
The findings suggest that the National Service Framework requirements for assertive outreach teams tend to select many patients with comorbid personality disorder in addition to other severe psychiatric disorders. The National Institute for Mental Health in England (2003) has issued guidance for the development of services for people with personality disorders; we present here data that suggest that existing assertive outreach teams might have a part to play in such services.
Knowledge of the prevalence of personality disorders in the community mental health teams is important for identifying treatment needs and for provision of psychiatric services. Personality disorder in people with severe mental illness is associated with adverse consequences, complicates treatment and worsens prognosis. Personality disorder is common in medical services, with a prevalence of around 30% in primary care attenders (Moran et al, 2002), over 50% among in-patients (Casey, 2000) and around 50% in community mental health teams (Keown et al, 2002).
Research in individuals with both severe mental illness and personality disorder is problematic because it is often difficult to separate symptoms caused by mental illness from those due to a personality disorder. The other consideration is that the presentation of psychiatric symptoms could be transformed by the presence of personality disorder. However, these difficulties can be minimised if the patients are well known and there is knowledge of function independent of major mental disorder.

Method
All 74 patients who on 1 August 2002 were on the register of an assertive outreach team were approached for this project. The Paddington Outreach Rehabilitation Team was set up by following strictly the requirements of the National Service Framework (Department of Health, 1999). Patients were approached and asked to give informed consent for an interview to take place with their principal worker to find out more about their personality status, using the informant-based ICD-10 version of the Personality Assessment Schedule (PAS-I). The Personality Assessment Schedule (PAS; Tyrer et al, 1979) is a semistructured trait-based interview, and the PAS-I version (Merson et al, 1994;Tyrer, 2000) links trait description with the ICD-10 operational criteria for individual personality disorders. It is a two-stage interview: in stage 1 screening questions from the original PAS are asked, and if an answer is positive, the rater moves to stage 2 where questions address the diagnostic guidelines for each personality disorder. The scoring is based on the degree of social dysfunction created by the personality feature and includes the level of personality difficulty as a sub-threshold category. For the purposes of the assessment, all interviewees were required to have known the patient for at least 3 years so that a comprehensive picture of the patient's main personality characteristics was available.
In addition to the type of personality disorder, the severity of the disorder was determined using a standard method, with levels of 'no personality disorder', 'personality difficulty only', 'simple personality disorder' (one or more personality disorders in one of the three cluster equivalents of DSM (flamboyant/dramatic, odd/eccentric and anxious/fearful)), 'complex personality disorder' (personality disorders from more than one cluster) and 'severe personality disorder' (complex personality disorders which create widespread and severe disruption of relationships, represent a threat to society and usually are associated with risk of violence) all recorded (Tyrer & Johnson, 1996;Tyrer, 2000, p. 129). A separate assessment of whether the patients were willing to have treatment separated those with type S (treatment-seeking) from type R (treatment-rejecting) personality disorders (Tyrer et al, 2003).

Results
Out of 74 patients, one refused to take part in the study. The low refusal rate was associated with the reassurance that no direct interviewing of the patients was needed. Of the 73 patients included in the study, 41 had a primary diagnosis of schizophrenia, 11 of bipolar disorder, 10 of schizoaffective disorder, 10 of primary personality disorder and 2 of unipolar depression. The interviews took place between August and November 2002, with 53 principal workers interviewed by M.R., 15 by C.M. and 5 by D.R. Assessments with the PAS-I showed that 67 patients (92%) had at least one personality disorder, with 3 (4%) having personality difficulty and 3 (4%) having no personality disturbance. Of the 67 patients with the diagnosis of personality disorder, most (40) were men. Table 1 shows the distribution of personality disorder categories in the 67 patients identified as having a personality disorder using the PAS-I. The PAS-I may identify several personality disorders, but the one causing the greatest social dysfunction is given primacy. Table 2 gives the distribution of personality disorders by severity. Table 3 categorises the patients' personality disorders into type R (treatment-rejecting) and type S (treatmentseeking). All comorbid personality disorders are included. Analysis showed that the distribution of personality disorders is not random: cluster C personality disorders have an excess of type S personalities and clusters A and B have an excess of type R personalities (w 2 =25.5, d.f.=8, P=0.0013).

Discussion
To the best of our knowledge this is the first epidemiological study examining the prevalence of personality disorder in clients of an assertive outreach team. The high prevalence (92%) may not be generalisable, as the team concerned has an atypical inner-city clientele and an interest in personality disorders. The possibility that chronic illness might be interpreted as personality disorder cannot be excluded, as this could be a source of Ranger et al Prevalence of personality disorder original papers   (Tyrer et al, 1983). However, the requirement that all those being interviewed must have extensive knowledge of the patient reduced the possibility of false positives. It is also relevant that in a previous study of the case-load of a community mental health team (Keown et al, 2002), 52% of patients were found to have a personality disorder, and it might be expected that a higher proportion would be found in an assertive outreach team. Some assertive outreach teams exclude those with personality disorder from their case-load, but this would be difficult to do in an inner-city area. The findings suggest that the skills brought to assertive outreach might well be suitable in setting up the nucleus of a dedicated service for this group of patients when planning new developments.
The results also suggest that some of the special features required for eligibility for an assertive outreach team -difficulty in engagement, frequent admissions and crisis presentation, antipathy to intervention -might be at least as much a consequence of a personality disorder as of a resistant mental illness. The distinction between type R and type S disorders may be valuable in clinical practice, particularly for those contemplating treatment of patients who have a diagnosis of personality disorder, but the requirements of an assertive outreach team probably lead to a bias towards type R personalities.

Declaration of interest
None.