A comparison of offenders with intellectual disability across three levels of security.

Background: A number of authors have described, with disparate results, the prevalence of people with intellectual disability and their characteristics, in a range of offender cohorts defined by service use. These have included high security, a range of criminal justice services and community services. There is a need for research comparing cohorts of offenders with intellectual disabilities across different settings. AIM AND HYPOTHESIS: To conduct such a comparison and test the hypothesis that severity of characteristics measured will be highest in highest levels of residential security.
Method: A clinical-record-based comparison a offenders with intellectual disability in high security (n = 73), medium/low security (n = 70), and a community service (n = 69).
Results: Groups were similar in age and tested IQ levels. Early psychiatric service contact had been more likely in the lower security groups. In line with the hypothesis, more complex presentations, in particular comorbid personality disorder, was more likely in the highest security group. Both fatal and non-fatal interpersonal violence convictions were significantly related to group, with more in the high security group sustaining a conviction both at the index offence and prior to that. Over 50% of all groups had at least one conviction for a sexual offence. A regression model accounting for 78% of the variance was made up largely of disposal variables (Mental Health Act status and probation) and indications of antisocial traits (criminal damage, lifetime conviction for murder and ICD-10 personality disorder classification).
Conclusions and Implications for Practice: The authors show that context of sampling affects most relationships between intellectual disability (ID) and offending when the methods for measuring ID are held constant. The results also present several questions on the relationship between risk, services available in an area and referral to higher security.