To measure the risk that someone with a mental illness will commit a crime, Washington State University researchers have developed a new assessment tool. It could also speed up long-delayed competency evaluations for people awaiting trial.
The assessment provides a small but notable improvement over the current evaluation system, said Alex Kigerl, an assistant research professor in the WSU Department of Criminal Justice and Criminology. Writing in a recent edition of the International Journal of Offender Therapy and Comparative Criminology, Kigerl and co-author Zachary Hamilton note a growing need to evaluate patients based on their risks to others and themselves before assigning them to inpatient hospital care or outpatient treatment.
"Often a balance must be struck between patient rights and freedoms and public safety and risk," they write. State officials have struggled in recent years to ease overcrowding in mental-health facilities. They face legal action for delays in evaluating patients' competency to stand trial. Last year, a U.S. district judge said Washington "is violating the constitutional rights of some of its most vulnerable citizens" by failing to undertake evaluations within the legal limit of seven days.
Kigerl and Hamilton prepared the assessment tool for the nonpartisan Washington State Institute for Public Policy, which the state legislature has asked to develop a risk-assessment instrument for patients with mental health issues. The researchers modified a tool used for offenders under state department of corrections supervision, called the Static Risk Assessment or SRA. "Static" characteristics refer to characteristics that can't be changed, like sex, age and prior offenses.
The SRA only uses those items because they lend themselves to easy computation, saving the staff and evaluator time needed to manually assess dynamic factors like mental health and substance abuse, said Kigerl. The researchers adapted the SRA to create the SRA for Mental Health Patients, or SRA-MHP, using a sample of 16,289 involuntarily committed patients and 8,713 patients undergoing evaluation after being charged with a crime.
The researchers developed their assessment using data from half the pool of patients, then tested its predictive power by comparing it against the other half. "The next step on this project would be to add dynamic characteristics - ones that can be changed - such as clinical variables about the patient, their mental health status, maybe chemical dependency, their employment and educational attainment and history," said Kigerl.
"If you look at a lot of literature on risk assessments, dynamic factors add slow but incremental improvement." The strength of the tool is measured by a figure called the "area under the curve," or AUC. An AUC of .5 would mean a tool is about as good a predictor as a coin flip. An AUC of 1.0 is perfect.The old evaluation method has an AUC of .79; the new SRA-MHP gave an AUC of .81.
"When we're talking about violent offenses, a little bit can go a long way," Kigerl said. The research is in keeping with WSU's Grand Challenges, research initiatives aimed at large societal problems. It is particularly relevant to the Sustaining Health challenge and its theme of promoting healthy communities, as well as the challenge Advancing Opportunity and Equity.