Although the small sample size allows only preliminary conclusions about the validity of this instrument, early results show a reduction of the diagnosis of comorbidity compared with the SCID-II, together with an increased number of expected associations between independent measures of interpersonal functioning and categories of personality disorder.

Non-normative samples can pose major procedural and coding challenges to interviewers and raters of the Adult Attachment Interview (AAI). With reference to interview transcripts drawn from a population of personality disordered offenders detained in a high-security hospital, specific difficulties are identified and discussed. These difficulties have their roots in three separate but overlapping areas: extreme attachment-related experience; interviewees' psychological or psychiatric state; and factors relating to the context in which the interview is conducted. They raise questions about whether and when the use of the interview should be restricted, the rating rules elaborated and/or the rating system expanded. Suggestions are made as to how some of the difficulties might be addressed.

Current DSM-based instruments for personality disorders (PDs) limit the investigation of the course and outcome of treatment of these disorders. This study examined the validity of the Shedler-Westen Assessment and the Structured Clinical Interview for DSM-IV Axis II PD (SCID-II) in a sample of forensic PD patients. Results based on 66 participants indicated that the SWAP-200 Q-factors reduced the frequency of diagnostic comorbidity of PD categories by half compared with the SCID-II. Only the SWAP-200's Antisocial PD category showed good convergent and discriminant validity with respect to other instruments describing aspects of PD. The validity of the cutoff score for severe antisocial PD was confirmed, and this category predicted severe incidents in the hospital at 1 year of follow-up. A violence risk scale was constructed, which differentiated violent and nonviolent offenders. The results support the validity of the SWAP-200 and its potential clinical utility with forensic PD patients.The prevalence and complexity of personality disorders (PDs) are particularly high in forensic populations in both the criminal justice and health systems. In the UK, up to 73% of prisoners awaiting sentencing, 64% of sentenced males, and 50% of female prisoners are estimated to meet criteria for the diagnosis of PD (Meltzer, 2008). For violent prisoners and PD

Background Forensic professionals attach considerable importance to their patient's description of his or her index offence which is frequently used to inform the patient's management and predict future behaviour. However, despite the cardinal importance of the index offence there is no systematic approach to examining and formulating the patient's offence narrative. Aim To examine whether a clinical tool, which tapped into the patient's capacity to mentalize, could be developed from the index offence narratives of violent, personalitydisordered patients. To see whether this tool would capture how the patient represents his or her index offence and predict the patient's progress, in terms of institutional aggression, pro-social behaviour, inter-personal relationships and psychiatric symptomatology. Method This was a prospective, cohort study. The index offence narratives of 66 violent, personality disordered patients were obtained from a semi-structured interview and used to generate the Index Offence Representational Scales (IORS). The predictive validity of these scales was investigated across a range of outcome variables, controlling for the association between initial and final value of the dependent variable. Results The degree of interpersonal violence and malevolence as measured by the IORS predicted subsequent violent behaviour. In contrast to their actual aggressive behaviour these patients rated themselves as having fewer symptoms on the SCL-90-R and problems in interpersonal relationships on the IIP. A more empathic victim representation on the IORS predicted those patients who engaged better in the hospital's therapeutic regime. Conclusions The IORS may prove a useful tool to help clinicians predict both institutional aggression and pro-social engagement in these difficult to manage patients. Future replication studies would be useful to further validate the IORS.