How effective are risk assessments/measures for predicting future aggressive behaviour in adults with intellectual disabilities (ID): A systematic review and meta-analysis.

Background: Risk assessments assist professionals in the identification and management of risk of aggression. The present study aimed to systematically review evidence on the efficacy of assessments for managing the risk of physical aggression in adults with intellectual disabilities (ID).

Methods: A literature search was conducted using the databases PsycINFO, EMBASE, MEDLINE, Web of Science, and Google Scholar. Electronic and hand searches identified 14 studies that met the inclusion criteria. Standardised mean difference effect sizes Area Under Curve (AUC) were calculated for studies. Random effects subgroup analysis was used to compare different types of risk measures (Actuarial, Structured Professional Judgment and dynamic), and prospective vs. catch-up longitudinal study designs.

Results: Overall, evidence of predictive validity was found for risk measures with ID populations: (AUC)=0.724, 95% CI [0.681, 0.768]. There was no variation in the performance of different types of risk measures, or different study design.

Conclusions: Risk assessment measures predict the likelihood of aggression in ID population and are comparable to those in mainstream populations. Further meta-analysis is necessary when risk measures are more established in this population.