The Center for Law and the Biosciences recently hosted UVA Law Professor John Monahan.  The “leading thinker on the issue” of violence risk assessment (as referenced in the U.S. Supreme Court case Barefoot v. Estelle), Prof. Monahan spoke on how we assess and may control the risk of violence by the mentally ill.  Many critical policies rely on violence risk assessment including civil, forensic, and SVP commitments; the death penalty; bail; probation and parole. Although the general public maintains a strong belief in the relation between violence and mental disorder, a majority of patients with mental illnesses are not violent (Swanson, 1991) though patients with histories of substance abuse are more likely to commit violent acts (MacArthur Violence Risk Assessment Study, Steadman, 1998).

Prof. Monahan reviewed current methods of predicting violence and explained their strengths and weaknesses. While fairly accurate overall, unstructured clinical methods are poor at predicting violence by women. The HCR-20 and VRAG methods utilize standard sets of risk factors that the practitioner identifies, measures, and combines to produce an estimate of risk.  These methods suffer two significant flaws. First, these methods were validated by narrow patient populations at single research sites with a constricted range of risk factors. Additionally, these methods often use arrests for violent crime as indicators of violence, thereby missing intrafamilial violence.

Prof. Monahan proposed a more accurate, evidence-based method of assessing violence risk known as COVR^TM (short for “Classification of Violence Risk^TM“) that utilizes patient reports, collateral reports from people who knew the patients well, and police and mental health records to classify a patient. The COVR^TM is an interactive software program designed to estimate the risk of an acute psychiatric patient becoming violent to others within a few months of discharge. The program guides the evaluator through a brief chart review and a 10-minute interview with the patient. The COVR^TM then generates a report containing a statistically valid estimate of the patient’s violence risk. Because many variables might be potential risk factors for violence among patients with mental disorders, Prof. Monahan and his co-developers assessed personal factors (e.g., demographic and personality variables), historical factors (e.g., past violence, mental hospitalizations), contextual factors (e.g., social support, social networks), and clinical factors (e.g., diagnosis, specific symptoms). The COVR^TM is based on a “classification tree” method that allows many different combinations of risk factors to classify an individual at a given level of risk. Each assessment is individualized; the particular questions asked depend on the answers given to prior questions and only as many questions as are needed to classify the patient are asked.  This starkly contrasts with the traditional methods described above that are often lengthy and yield ambiguous results.

While violence risk assessment methodology is steadily improving, there is still significant room for improvement, especially given the magnitude of what is at stake for the patients. Future methodologies will undoubtedly yield accurate results across races/ethnicities/genders as well as across different regions.

— Vinita Kailasanath