Given that these U.S. studies examined individuals referred for clinical assessment, their results reflect patterns seen in individuals for whom a judge or attorney had raised a concern about competence to stand trial or criminal responsibility. Thus, these studies inform us of the diagnoses of a selected sample of homicide offenders but do not provide data on homicide offenders more generally. It is therefore not surprising that the rates of mental illness observed in these studies vary considerably, probably because of differing referral patterns across jurisdictions. These studies also do not provide an accurate assessment
of the relationships. between case characteristics and features of the homicide offense, since selection bias could affect any patterns observed. To our knowledge, there has been no previous study of the psychiatric characteristics of an unselected group of homicide offenders in the United States or an examination of the relationship of mental disorders to case outcomes in such a sample (19).