Exploring Gender Differences in the Utility of Strength-Based Risk Assessment Measures

ABSTRACT

The generalizability of risk assessment measures to female populations remains up for debate; in particular, few studies have made direct comparisons between male and female civil psychiatric patients on protective factors and risk factors relevant to violence risk assessments. To address this gap in the literature, we conducted a prospective study with 102 civil psychiatric patients (60.8% male) to investigate strength-based risk assessments. Outcome data (i.e., verbal, physical, and sexual aggression) was collected after 6 and 12 months. We found a number of potentially interesting gender differences in the predictive validity of the START, HCR-20^V2, and SAPROF. Findings are generally supportive of the use of established Structured Professional Judgement (SPJ) risk assessment measures with male civil psychiatric populations, and with the exception of the START, caution is warranted when using these measures with female civil psychiatric patients. Findings suggest that SPJ assessments that utilize both strengths and vulnerabilities generally performed better than SPJ assessments relying on either strengths alone or vulnerabilities alone.

KEYWORDS:

• Gender
• risk assessment
• violence
• strengths
• protective factors
• civil psychiatric patients

Acknowledgments

We are especially grateful to the many patients who shared their personal stories with us.

Funding

This study was funded by a Canadian Institute of Health Research grant. We are also grateful for the funding provided by British Columbia Mental Health & Addiction Services, the Michael Smith Foundation for Health Research, and the MIND Foundation. Tonia Nicholls would also like to recognize the support provided by her salary awards from the CIHR and the MSFHR. Simone Viljoen would like to recognize the support provided by her graduate funding from Social Science and Humanities Research Council of Canada.

Notes

¹ Charts were rich with information from multiple disciplines and had more than sufficient information to complete all measures. The same site has been used to complete file-based risk assessment studies using several of the same measures (i.e., HCR-20 and PCL:SV; Douglas, Ogloff, Nicholls, & Grant, 1999; Nicholls, Ogloff, & Douglas, 2004).

² The two raters were randomly selected from a possible sample of 12 raters.

³ Readers interested in the interviews used should please contact the corresponding author.

⁴ The HCR-20 Version 3 (Douglas, Hart, Webster, & Belfrage, 2013) was not available at the time this study was conducted.