. 2009 Dec 3;37(4):800–810. doi: 10.1093/schbul/sbp135

Table 2.

Risk Estimates for Criminal Recidivism in Schizophrenia and Other Psychoses by Sample or Study Characteristics

Sample or Study Characteristics	Number of Studies	Number of Case Subjects With Psychosis	OR (95% CI)
Type of psychosis
Schizophrenia	17	871	1.0 (0.7–1.4)
Schizophrenia and other psychoses	10	196	1.1 (0.7–1.9)
Study period
Study conducted before 1990	12	186	0.8 (0.4–1.4)
Study conducted in 1990 or after	15	881	1.2 (0.9–1.7)
Diagnostic criteria^a
DSM criteria	15	891	0.9 (0.7–1.3)
ICD criteria	2	37	1.4 (0.7–2.9)
Design
Case-control	5	68	1.7 (0.9–3.0)
Cohort	22	999	1.0 (0.7–1.3)
Mean age^b
30 y or younger	6	59	0.6 (0.3–1.3)
Older than 30 y	11	348	0.8 (0.6–1.2)
Number of cases
<100 cases	17	242	1.2 (0.7–1.9)
≥100 cases	10	825	0.9 (0.6–1.3)
Outcome
Violent recidivism	6	177	1.0 (0.5–1.9)
Any criminal recidivism	21	890	1.1 (0.8–1.5)
Duration of follow-up (y)^c
<6	8	596	0.9 (0.7–1.2)
6–10	7	1923	1.0 (0.5–2.0)
>10	3	283	1.0 (0.5–2.7)

Note: OR, odds ratio; CI, confidence interval; DSM, Diagnostic and Statistical Manual of Mental Disorders; ICD, International Classification of Diseases.

Number of studies and case subjects differ in this analysis because 10 studies³⁹^,⁴¹^–⁴⁴^,⁴⁶^,⁴⁷^,⁴⁹^,⁵¹ did not provide information on diagnostic criteria.

Number of studies and case subjects differ in this analysis because 10 studies¹⁵^,³¹^,³⁹^,⁴²^,⁴⁶^,⁵⁰^,⁵¹^,⁵⁵^,⁵⁶ did not provide information on mean age of the sample.

Number of studies and case subjects differ in this analysis because 9 studies³⁸^,³⁹^,⁴²^,⁴⁵^–⁴⁷^,⁵⁶ did not provide information on years of follow-up.