A form of relapse prevention for men in a high security hospital

Marlatt and Gordon's relapse prevention model has evolved, and been applied to many repetitive problem behaviours, including drug use, gambling and sexual offending. High security hospital patients often present with a number of such behaviours.

To evaluate a relapse prevention group in one high security hospital.

Nine patients were selected on clinical need for the group. In a naturalistic study design, each completed several self-report questionnaires pre- and post-group.

Analysis of group results demonstrated a significant increase in guilt and acceptance of responsibility for offences (GBAI ratings), and reduced endorsement of pro-offending thinking styles (PICTS). Movement towards increased empathy, and decreased impulsivity and venturesomeness, did not reach significance. All participants showed some changes in desired directions but effects were small.

Both relapse prevention and its evaluation are limited in high security hospitals by the restricted opportunities for testing out new skills. The feasibility of delivering an RP programme and evaluating it in high security has nevertheless been established. The small effects found may be promising, and suggest both justification and need for a randomized controlled trial of this approach in that setting.