Letter

In Reply: We thank Dr. Lamberti for his clarification. In his 2007 article he noted that the MacArthur study reported a "lack of an association between psychosis and crime," and his was one of many similar references that we had noted over the years. He is correct that his article did not claim that the MacArthur results can be generalized; however, once a finding is repeated often enough in the literature, it develops a life of its own and tends to be generalized.

The main purpose of our undertaking the debate in the February issue was to make clear that on the basis of data from the MacArthur study that became available after the original publication, it is erroneous to cite the study as having shown a "lack of an association between psychosis and crime." Dr. Lamberti is also correct that several more recent studies have clearly shown that "active psychosis is an additional risk factor for violent crime." In his 2007 article, Dr. Lamberti cited the 2006 study by Swanson and colleagues ( 1 ), and we would add three others ( 2, 3, 4 ).

The most important point made by Dr. Lamberti, and also by subsequent publications on the MacArthur study, is that the treatment of individuals with psychoses significantly reduces the incidence of their violent behavior. The question he raises about "how to engage high-risk individuals in treatments" must take into account the fact that approximately half of such individuals are unaware of their own illness and of the need for treatment. In these cases treatment must sometimes be mandated, by using such mechanisms as assisted outpatient treatment, which has been demonstrated to significantly reduce violent behavior among individuals with severe psychiatric disorders ( 5 ).