Abstract

Background -- Over the last 10 years there has been greater interest in the diagnosis of
personality disorder (PD) in people with intellectual disabilities (ID). One important characteristic of
a diagnostic system is that it should have validity as a contribution to utility. PD has been found to
have a predictive relationship with violence and the present study reviews two methods for the
diagnosis of PD in offenders with ID in order to evaluate the utility of the diagnoses.
Method --212 offenders with ID were recruited from three settings -- maximum-security,
medium/low security and community services. Three research assistants were trained over a period of
two weeks in order to increase reliability for the extraction of information from the case notes and the
diagnosis of PD. Diagnoses of PD in the case files were compared with a structured system of
diagnosis based on DSM IV traits.
Results – There were significant differences between the two systems with a significantly
higher frequency of PD diagnosis in the community forensic setting in the structured assessment
system. There was no relationship between the case files diagnosis of PD and future violence but there
was a significant predictive relationship between the structured diagnosis of PD and future violence
with an AUC=.62.
Conclusions -- Only the structured assessment of PD had utility for the prediction of violence.
Reasons for the differences between the systems are discussed and suggestions made on how a
diagnosis of PD can be structured for the busy clinician.