In Reply: Dr. Torrey correctly points out that our sample was involuntarily hospitalized and subsequently court-ordered to outpatient commitment, orders generally associated with noncompliance, and, furthermore, that our sample had a 71 percent noncompliance rate, as we reported in another article (reference 2 above). He concludes that these factors indicate "a direct relationship between medication noncompliance and criminal victimization."
It is possible that taking antipsychotic medication may reduce the risk of victimization by ameliorating symptoms such as disorientation, which might cause a severely mentally ill person to wander into dangerous areas, appear more vulnerable, and thus be criminally victimized, or by ameliorating symptoms such as annoying behavior, which might be provocative of abuse from acquaintances and friends. But we believe that the relationship between medication noncompliance and criminal victimization is more complicated.
Victimization reflects exposure to crime and violence in the social environment. Our analysis, along with the analyses of others who have looked at the problem, shows that victimization is associated with a constellation of problems such as alcohol abuse, drug dependence, homelessness, and mental disorder. Additionally, in our sample and in other samples of persons without mental disorders, victimization is associated with criminal and violent behavior.
Our second paper to which Dr. Torrey refers reported that medication noncompliance has a significant effect on violence only in interaction with substance abuse. As we suggest, medication noncompliance may lead to self-medicating with alcohol or drugs, and the subsequent impairment may impede medication adherence so that focusing on treatment of one without the other is unlikely to lead to effective reduction of violence and victimization. Problems of the social environment, such as homelessness, need to be addressed as well.