Reading time: 2 minutes 30 seconds
Decisions about treatment, supervision, or detention often depend on how well forensic mental health professionals can predict the risk of reoffending. Although risk assessment tools have been developed to guide professionals in this important query, some professionals continue to assess these risks using unstructured professional judgment.
A meta-analysis shows that these tools can help predict various types of reoffending more accurately than unstructured professional judgments.
This is a plain language summary of:
Viljoen, J. L., Goossens, I., Monjazeb, S., Cochrane, D. M., Vargen, L. M., Jonnson, M. R., Blanchard, A. J. E., Li, S. M. Y., & Jackson, J. R. (2025). Are risk assessment tools more accurate than unstructured judgments in predicting violent, any, and sexual offending? A meta-analysis of direct comparison studies. Behavioral Sciences & the Law, 43(1), 75–113. https://doi.org/10.1002/bsl.2698
About the Meta-Analysis
Dr. Ilvy Goossens and her colleagues conducted the largest meta-analysis to date, directly comparing risk assessment tools with unstructured risk judgments. The authors included only studies that made direct comparisons between risk assessment tools and unstructured risk judgments (31 studies), evaluating both approaches using the same participants (more than 45,500 individual risk judgments). This approach addressed the limitations of earlier meta-analyses and provided a more reliable test of how well the tools predict reoffending.
The researchers examined three types of reoffending:
- Violent offending (actual, attempted, or threatened harm to another person),
- Sexual offending (actual, attempted, or threatened non-consensual sexual acts), and
- Any offending (a broader measure combining multiple offences, including violent or sexual offences, property crime, and violations of conditions).
To make these comparisons, Dr. Goossens and her colleagues considered two main types of risk assessment tools:
- Actuarial tools, which use fixed scoring formulas; and
- Structured professional judgment tools, which combine evidence-based guidelines with clinical expertise.
What They Found
On average, risk assessment tools predicted all three outcomes more accurately than unstructured risk judgments. While there was no meaningful difference between actuarial and structured professional judgment tools, both helped assessors focus on empirically validated risk factors, such as prior violent behaviour, substance use, mental health concerns, and criminal attitudes. For over three decades, research has consistently links these factors to reoffending.
The authors found that the results were robust, even after accounting for risk of bias, variations in population (e.g., ethnic composition, inpatient versus outpatient), assessment context, and outcome measurement. Risk assessment tools tended to perform better than unstructured judgments regardless of evaluator experience, assessment setting, or participant group.
Why It Matters
These findings support the broader use of risk assessment tools in forensic psychiatry and criminal justice systems. By guiding professionals toward evidence-based decision-making, these tools improve predictive accuracy and reduce errors caused by personal bias or professional intuition.
However, the authors caution that structured tools should complement, not replace, clinical expertise. The meta-analysis also highlights equity concerns. Researchers developed and validated most tools with Western samples, primarily involving males. This can limit understanding of how well they perform with non-Western populations, racialized groups, and females. More inclusive research is needed to ensure that risk assessment practices are fair and effective across different populations and settings.
Implications for Policy and Practice
Because these risk assessment tools are more accurate, justice and mental health agencies should support their use them more widely. They are already part of practice in parole offices, custodial facilities, forensic units, and treatment programs for sexual offending. But their use remains unevenly distributed. Large differences remain internationally and between professions (e.g., psychology versus psychiatry). Around the globe, almost half of risk assessments are still based on unstructured judgments. These gaps in implementation limit how much benefit the tools can provide.
Accuracy is important. But it will not improve outcomes unless the tools are used consistently and supported in practice. Agencies need to make sure the tools are built into everyday work, backed by training, and supported with the resources required to use them properly.
Researchers also have a part to play. They can help by investigating how tools are used in different contexts, addressing gaps in development and implementation, and explaining clearly what these tools can and cannot do. Tools do not explain themselves — people must decide how to interpret and apply the results.
Written by Kyle Roerick and approved by Dr. Ilvy Goossens
