Predictive validity of the structured assessment of violence risk in youth (SAVRY) on the recidivism of juvenile offenders: a systematic review

This study systematically reviewed the available evidence on the predictive validity of the Structured Assessment of Violence Risk in Youth [SAVRY - Borum et al., 2006. Structured assessment of violence risk in youth (SAVRY). Psychological Assessment Resources, Inc] on the recidivism of juvenile offenders. In addition, we explored potential moderators related to characteristics of (a) the juvenile offender, (b) the administration of the instrument and (c) the measure of recidivism. In total, our search identified 1,845 references of which 13 reports fulfilled the criteria to be included in the final evidence-base. Our results are consistent with past research in underlining that the ability of the SAVRY to predict future violence varies extensively with effect sizes ranging from negligible or small (AUC = 0.44, SE = 0.08, 95% CI = 0.28-0.61) to large (AUC = 0.87, SE = not reported, 95% CI = 0.62-1.00), irrespective of the type of recidivism measure and type of SAVRY score used. Available data on potential moderators revealed no visible patterns to explain this variability. We discuss how the database included in this article might serve as a tool for practitioners to consider the variability concerning the predictive validity of the SAVRY by filtering according to different participant or administration setting characteristics that match their particular practical context.