A neurophysiological predictor of reoffending in special hospital patients
The aim of this study was to assess a possible role for a slow brain potential measure, the Go/No Go contingent negative variation (CNV), as an indicator of subsequent reoffending by mentally disordered offenders released from a special high-security hospital. The rationale was twofold: (1) Go/No Go...
| Authors: | ; |
|---|---|
| Format: | Electronic/Print Article |
| Language: | English |
| Published: |
1996
|
| In: |
Criminal behaviour and mental health
Year: 1996, Volume: 6, Issue: 2, Pages: 147-156 |
| Online Access: |
Volltext (doi) |
| Journals Online & Print: | Drawer...
|
| Check availability: | HBZ Gateway |
| Keywords: |
| Summary: | The aim of this study was to assess a possible role for a slow brain potential measure, the Go/No Go contingent negative variation (CNV), as an indicator of subsequent reoffending by mentally disordered offenders released from a special high-security hospital. The rationale was twofold: (1) Go/No Go CNV has been found to correlate with psychometric measures of impulsivity; (2) impulsivity has been found to predict recidivism in sexual offenders and in violent offenders generally. The subjects were 44 men compulsorily admitted to Broadmoor Hospital from 1977 to 1980 under the Mental Health Act 1959 (MHA 1959). They fell within the legal categories of psychopathic disorder (PD: n = 16) or mental illness (MI: n = 28). CNV recordings were carried out within 3 months of admission using a Go/No Go avoidance paradigm. Subjects were assigned to a "high risk" (HR: n = 21) group if their Go/No Go difference fell at or beyond a standard deviation from the mean for a normal control group comprising 19 student nurses, matched for age and sex, in whom CNVs were recorded at the same time; they were otherwise classed as "low risk" (LR: n = 23). Patients were followed post-discharge for up to 15 years. Information was obtained pertaining to reconvictions, rehospitalisation and death from the special hospitals case register. Reconvictions for any offence occurred in six of 21 HR patients, compared with one of 23 LR patients, giving an overall accuracy for the predictor of 63.6%, and a relative improvement over chance of 72%. These results offer some promise that Go/No Go CNV may be a useful predictor of recidivism in mentally disordered offenders. Copyright Ř 1996 Whurr Publishers Ltd |
|---|---|
| ISSN: | 0957-9664 |
| DOI: | 10.1002/cbm.82 |