Table 4.
Twelve-Month Incidence of Crime Victimization Among Men and Women with Severe Mental Illness, Incidence Ratios (Relative to the National Crime Victimization Survey), and 95% Confidence Intervals*
|
Men (n = 519 Incidents) |
Women (n = 651 Incidents) |
||||
|---|---|---|---|---|---|
| Type of Crime† | Annual Incidence per 1000 (95% CI) | Incidence Ratio (95% CI)‡ | Annual Incidence per 1000 (95% CI) | Incidence Ratio (95% CI)‡ | PValues for Tests of Sex Differences in Annual Incidence |
| Violent personal crimes | |||||
| Any crimes of violence | 165.8 (136.9–194.6) | 3.5 (2.8–4.4)§ | 170.5 (144.6–196.3) | 4.9 (4.0–6.1)§ | .81 |
| Completed violence | 76.6 (57.4–95.9) | 3.9 (2.8–5.3)§ | 87.2 (67.4–107.0) | 5.4 (4.1–7.2)§ | .45 |
| Attempted/threatened violence | 89.1 (66.2–112.0) | 3.3 (2.3–4.7)§ | 83.3 (65.1–101.4) | 4.5 (3.6–5.7)§ | .69 |
| Rape/sexual assault | NA|| | NA | 29.4(17.8–41.0) | 7.7 (5.2–11.5)§ | NA |
| Robbery | 36.3 (23.7–48.9) | 3.6 (2.5–5.4)§ | 12.3 (4.6–20.1) | 2.8 (1.4–5.4)¶ | .002 |
| Assault | 125.9 (100.1–151.6) | 3.4 (2.4–4.8)§ | 128.7 (106.0–151.4) | 4.9 (3.8–6.4)§ | .87 |
| Aggravated | 32.7 (22.2–43.3) | 3.8 (2.1–6.9)§ | 21.8 (11.9–31.6) | 5.2 (3.3–6.0)§ | .14 |
| Simple | 93.1 (69.3–116.9) | 3.3 (2.4–4.7)§ | 107.0 (86.1–127.8) | 4.8 (3.8–6.2)§ | .39 |
| Personal theft (theft of property from person) | 135.5 (107.3–163.7) | 63.1 (38.1–104.6)§ | 149.2 (122.1–176.3) | 58.0 (39.7–84.8)§ | .49 |
| Property crimes | 390.0 (334.7–445.3) | 1.6 (1.3–1.8)§ | 360.6 (311.5–409.6) | 1.6 (1.4–1.9)§ | .44 |
| Household burglary | 150.1 (107.7–192.6) | 2.8 (2.4–3.3)§ | 134.3 (101.9–166.6) | 2.5 (1.9–3.1 )§ | .56 |
| Motor vehicle theft | 6.7 (0.9–12.4) | 0.4 (0.1–1.2) | 15.4 (4.5–26.3) | 1.2 (0.5–2.7) | .16 |
| Property theft | 233.2 (185.3–281.0) | 1.3 (1.1–1.6)# | 210.9 (167.6–254.2) | 1.4 (1.0–1.8)# | .50 |
Abbreviations: CI, confidence interval; NA, not applicable; NCVS, National Crime Victimization Survey.
The NCVS incidences include all persons living in the “central cities” (ie, the largest cities of each standard metropolitan area) of metropolitan statistical areas weighted to the age, sex, racial/ethnic, and income distribution of our sample of persons with severe mental illness.
Categories and subcategories of type of crime are those used in NCVS publications.
Ratio of incidence for persons with severe mental illness to incidence reported in NCVS.
The 2-tailed probability that the confidence bound of the incidence ratio overlaps with 1.0 (no effect) is <.001.
There are too few cases of rape/sexual assault in the sample of men with severe mental illness to report incidence rates, incidence ratios, or tests by sex.
The 2-tailed probability that the confidence bound of the incidence ratio overlaps with 1.0 (no effect) is <.01.
The 2-tailed probability that the confidence bound of the incidence ratio overlaps with 1.0 (no effect) is <.05.