SAVRY

The Structured Assessment of Violence Risk in Youth (SAVRY), developed by Randy Borum, Patrick Bartel, and Adelle Forth, is a risk assessment instrument designed to structure appraisals of violence risk and risk management plans for adolescents. Such assessments are routinely required by juvenile and criminal courts and at nearly every juncture of the juvenile justice system. In the SAVRY, one’s risk for serious violence is viewed as the result of dynamic and reciprocal interplay between factors that increase and factors that decrease the likelihood of violence in the developing juvenile over time. Its central objective is to facilitate assessments that are systematic, empirically grounded, developmentally informed, treatment oriented, flexible, and practical.

The SAVRY is based on the “structured professional judgment” (SPJ) risk assessment framework, and is designed for use with adolescents between the approximate ages of 12 and 18 who have been detained or referred for an assessment of violence risk. Evaluators systematically assess predetermined risk factors that are empirically associated with violence, consider the applicability of each risk factor to a particular examinee, and classify each factor’s severity. The ultimate determination of an examinee’s overall level of violence risk is based on the examiner’s professional judg-ment as informed by a systematic appraisal of relevant factors. In this way, the SPJ model draws on the strengths of both the clinical and actuarial (formula-driven) approaches to decision making and attempts to minimize their respective drawbacks.

The SAVRY protocol is composed of 6 items defining protective factors and 24 items defining risk factors. Items are coded on the basis of reliable, available information. Information should be obtained from multiple sources, including an interview with the youth and a review of records (e.g., police or probation reports and mental health and social service records). Risk items are divided into three categories: historical, individual, and social/contextual, and each is coded for severity according to a three-level coding structure (high, moderate, or low). Protective factors are coded as present or absent. The coding form also includes a section for listing “additional risk factors” and “additional protective factors” because the SAVRY is not exhaustive in identifying all potential risk and protective factors for any given individual. If these additional factors are present, they should be documented and weighed in final decisions of risk.

Though the SAVRY is sufficiently flexible to accommodate varying styles of risk communication, the coding form prompts evaluators to make a final summary risk rating of low, moderate, or high. The ultimate risk rating is not based on cutting scores (SAVRY items are not assigned numeric values) or a specific formula. Evaluators exercise their professional judgment to determine the nature and degree of the juvenile’s risk for violence after carefully weighing the relevant risk and protective factors.

Psychometric Properties

One of the primary indices used in SAVRY research (not in practice) is the “SAVRY Risk Total.” This is calculated by transposing item ratings of low, moderate, and high to numerical values of 0,1, and 2, respectively, and summing the values. The summary risk rating is similarly transposed for statistical analysis.

Reliability and Internal Consistency. The internal consistency (alpha coefficient) of the SAVRY Risk Total has been shown to fall in the range of .82 to .84. Interrater reliability for the SAVRY Risk Total has ranged between .74 and .97, with an unweighted average of .84. Interrater reliability coefficients for the Summary Risk Rating have ranged between .72 and .85, with an unweighted average of .78.

Validity. The concurrent validity of the SAVRY has been examined in relation to the Youth Level of Service/Case Management Inventory (YLS/CMI) and the Hare Psychopathy Checklist: Youth Version (PCL:YV). In the initial validation study, the SAVRY Risk Total correlated significantly with both instruments among offenders and in community samples (e.g., .89 with YLS/CMI and .78 with PCL:YV among offenders). The SAVRY protective domain was negatively correlated with both the other measures. Across five studies, correlations between the SAVRY Risk Total and the PCL:YV Total Score have ranged from .48 to .74, with an unweighted average of .67. Although the correlations are significant, indicating that the SAVRY shares variance with other measures, it also possesses independent predictive power.

With regard to criterion validity, studies have found significant correlations between SAVRY scores and various measures of violence in juvenile justice and high-risk community-dwelling populations. In the initial validation sample, SAVRY Total Risk scores were all significantly related to behavioral measures of institutional aggressive behavior (.40) and aggressive conduct disorder symptoms (.52). The SAVRY has also demonstrated incremental (criterion) validity (or predictive power) beyond the YLSI and the PCL:YV. Results of hierarchical regression analyses showed that adding the SAVRY improved the power of the YLSI and the PCL:YV in predicting both institutional aggressive behavior and serious aggressive conduct disorder symptoms. The SAVRY also accounted for a large proportion of the explained variance in each type of violence.

Using receiver operating characteristic (ROC) analysis, which measures predictive accuracy in terms of relative improvement over chance, we found that areas under the curve (AUCs) for the total score average about .74 to .80 across studies. Interestingly, the examiner judgments (summary risk rating), not made on the basis of any cutting score, consistently perform as well as, and often better than, the linear combination of the scores themselves. For example, using ROC analysis, an unpublished master’s thesis at Simon Fraser University reported an AUC of .70 for the SAVRY total score, but the AUC for the SAVRY Summary Risk Rating was .89. This finding has been evident in research on other SPJ tools as well and provides some of the first empirical evidence that clinical judgments—properly structured and based on sound assessments—can achieve levels of accuracy that rival that of any other known predictors, while maintaining latitude for case-specific analysis.

Additional research is needed and is under way to clarify the SAVRY’s applicability in different countries, across genders, and with various ethnic groups.

References:

1. Borum, R., Bartel, P., & Forth, A. (2006). Manual for the Structured Assessment for Violence Risk in Youth (SAVRY). Odessa, FL: Psychological Assessment Resources.
2. Borum, R., & Douglas, K. (2003). New directions in violence risk assessment. Psychiatric Times, 20(3), 102-103.
3. Borum, R., & Verhaagen, D. (2006). A practical guide to assessing and managing violence risk in juveniles. New York: Guilford Press.

Return to the overview of Violence Risk Assessment in Forensic Psychology.