Community Treatment Programs

Violence tends to be concentrated in a small subgroup of the population (Blumstein, Cohen, Roth, & Visher, 1986). This pattern applies to violence involving individuals with mental disorders as well. For example, Gardner, Mulvey, Lidz, and Shaw (1996) found that 5 % of psychiatric patients followed in their community samples were involved in 45% of all reported violent incidents in the sample. More recently, Banks et al. (in press) found that 15 % of psychiatric patients enrolled in their study were involved in 67 % of identified violent incidents. There is some evidence that these frequently violent patients ’ incidents are more serious (e.g., assaults requiring medical treatment) than incidents in which “occasionally violent ” patients are involved (Gardner et al., 1996). Relatively little is known about methods for identifying patients involved in frequent and serious violence. Research on violence risk typically focuses on distinguishing patients who are likely to be involved in one or more violent incident from those with no involvement in violence (Borum, 1996). Although this work has advanced risk assessmen