Violence in schizophrenia has garnered much recent attention in the wake of the July 2012 Aurora, Colo., shooting, when a gunman allegedly diagnosed with schizophrenia opened fire in a movie theater.1 But despite highly publicized cases of this nature, the percentage of individuals with schizophrenia who are violent is actually rather low,2,3 with estimates ranging from 8–19%.2,4-7 Nevertheless, individuals with schizophrenia are known to be at increased risk of engaging in violence, compared to the general population.8 For this reason, current treatment guidelines, both in the U.S. and in the U.K., recommend that risk of violence be assessed in all individuals diagnosed with schizophrenia,9,10 and more than 100 instruments have been developed for this purpose.11,12

Currently used instruments designed to identify patients who are at highest risk of violence are time-consuming and expensive to administer. For example, certain structured risk assessment tools can take an average of 15 hours to complete and might cost service providers as much as $100 per hour.13 This labor-intensive process overburdens the resources of mental health services,8 so developing a more efficient risk assessment process would be extremely helpful.

In their article titled “A National Study of Violent Behavior in Persons with Schizophrenia,” Singh et al propose a novel approach to determining risk of violence.8 Rather than beginning by assessing individuals at high risk for violence, which they call a “one-step approach,” they suggest a “stepped” approach that begins with screening out individuals at low risk for violence before proceeding to in-depth assessments of potential high risk.

The authors compare their approach to that of other diagnostic screening tools used in physical medicine—for example, using mammography screening for breast cancer, or prostatic specific antigen (PSA) screening for prostate cancer—in which results of these screenings inform the decision whether or not to biopsy. They note that this approach would conserve mental health resources by screening out those at lowest risk and reserving the more detailed, costly, and time-consuming risk assessments for those at higher risk, who are more likely to need preventive interventions.

The authors developed and cross-validated a simple screening tool to identify individuals with schizophrenia who are at lowest risk of violence after hospital discharge. The proposed screening tool assesses five factors known to increase the risk of violence in persons with schizophrenia: male sex, previous criminal conviction, young age at assessment, comorbid alcohol abuse, and comorbid drug abuse.8 Using data on 13,806 patients with schizophrenia extracted from Swedish national registers, the researchers followed post-discharge patients for up to 33 years (between January 1, 1973 and December 31, 2004). “Violence” was defined as a conviction for homicide, assault, robbery, arson, sexual offense, illegal threats, or intimidation.

The researches utilized Cox regression to determine risk factors for violent crime and construct the screening tool. The instrument was calibrated on 6,903 participants and cross-validated using three independent replication samples of 2,301 participants each.

At five years after discharge, the instrument had a negative predictive value of 0.99 (95% CI = 0.98–0.99), meaning that only a small number of individuals who had been screened out by the tool (n = 2,359 out of original sample of 6,903) were subsequently convicted of a violent offense. The authors state that “the tool could accurately be used to make ‘rule out’ decisions (ie, identifying who will not go on to violently offend), suggesting utility as a screening tool that could be used as part of a stepped approach to risk assessment.”