Drs. Fazel and Grann Reply

To the Editor: In our article, we investigated the population impact of patients with severe mental illness on violent crime by analyzing national hospital and crime registers in Sweden over a 13-year period. Overall, we estimated a population attributable risk fraction of 5.2%, suggesting that one in 20 violent crimes were committed by patients with hospital diagnoses of severe mental illness. In addition, we reported an increased risk of patients with severe mental illness to commit violent crimes (crude odds ratio=3.8, 95% confidence interval [CI] 3.7–3.9).

Dr. Pinta is right to point out that one of the reasons for this increased risk ratio may be that psychiatric patients are more likely to be criminalized than individuals without such illnesses. For example, psychiatric patients who offend may be more likely to be caught by the police or have poor legal representation when charged with a crime. In Sweden, courts routinely provide mentally disordered offenders with qualified legal representation, so having poor legal advice may not be an important factor. However, it is also possible that there is a greater likelihood that individuals with mental illnesses who commit violent acts are acquitted before charges are pressed or have their charges dropped before trial than the non-mentally disordered. This may be because the police or prosecuting authorities consider that it does not serve the public interest to pursue such criminal investigations. Thus, to summarize, we feel it is unlikely that the higher rate of violent convictions among the mentally ill is simply because of biased criminal justice processes.

Dr. Pinta suggests that it is not possible to draw conclusions about the causative relationship between mental illness and violent crime from our study. We agree—the main focus of our study was to estimate the population impact of patients with severe mental illness to violent crime. We stated in our article that the relationship between severe mental illness and violence is more complex than simple unidirectional causality and that a host of demographic, criminal history, and clinical risk factors will affect the relationship. This is also relevant in studies of the population impact of other mental disorders, such as substance abuse and dependence (1) . The relative contribution of these various risk factors remains an important area of future research.