A real-world validation of the Risk for Sexual Violence Protocol (RSVP)

This study assessed the concurrent and predictive validity of clinician judgements made using the Risk for Sexual Violence Protocol (RSVP) in routine practice, and examined how they used the guideline. Risk assessments were accessed from the files of 100 men referred to an Australian community forensic mental health service due to sexual violence. Predictive validity was examined at four time points based on police charges, though only 5 men reoffended during the follow-up period. RSVP Case Prioritization judgements were strongly correlated with the Static-99 and had excellent predictive validity at all follow-up times. Those judged High priority reoffended significantly more often than cases judged Moderate or Low priority. Generally, clinician judgements had better sensitivity than specificity, though the low number and rate of recidivism mean these findings must be interpreted cautiously and require replication. In practice clinicians rarely completed the Future Relevance ratings, Psychopathy risk factor, or the Serious Physical Harm and Immediate Action Required Summary Judgements. Aspects of the practice setting in which the RSVP was used may have helped to maintain predictive validity despite clinicians’ departure from its instructions.