From: Predicting hospitalization following psychiatric crisis care using machine learning
Variable | All participants (n = 2084) | Hospitalized (n = 710) | Not hospitalized (n = 1374) | X2 (df) | p |
---|---|---|---|---|---|
M (SD)|n (%) | M (SD)|n (%) | M (SD)|n (%) | |||
Age | |||||
 Years | 40.8 (15.1) | 41.0 (13.8) | 40.7 (15.7) | 0.94 (1) | 0.33 |
Sex | |||||
 Male | 1083 (52.0%) | 405 (57.0%) | 678 (49.3%) | 10.81 (1) | 0.001 |
 Female | 1001 (48.0%) | 305 (43.0%) | 696 (50.7%) |  |  |
Diagnosis |  |  |  | 120.2 (5) | < 0.0001 |
 Psychotic | 807 (38.7%) | 373 (52.5%) | 434 (31.6%) |  |  |
 Depressive | 285 (13.7%) | 98 (13.8%) | 187 (13.6%) |  |  |
 Substance related | 239 (11.5%) | 84 (11.8%) | 155 (11.3%) |  |  |
 Manic/bipolar | 34 (1.6%) | 12 (1.7%) | 22 (1.6%) |  |  |
 Other | 561 (26.9%) | 103 (14.5%) | 458 (33.3%) |  |  |
 No or deferred | 158 (7.6%) | 40 (5.6%) | 118 (8.6%) |  |  |
Living situation |  |  |  | 35.35 (5) | < 0.0001 |
 Alone | 1018 (48.8%) | 385 (54.2%) | 633 (46.1%) |  |  |
 With partner/other(s) | 564 (27.1%) | 142 (20.0%) | 422 (30.7%) |  |  |
 With parents | 235 (11.3%) | 73 (10.3%) | 162 (11.8%) |  |  |
 Homeless | 96 (4.6%) | 42 (5.9%) | 54 (3.9%) |  |  |
 Institutionalized | 68 (3.3%) | 31 (4.4%) | 37 (2.7%) |  |  |
 Other | 103 (4.9%) | 37 (5.2%) | 66 (4.8%) |  |  |
Cultural background | Â | Â | Â | 12.16 (5) | 0.033 |
 Dutch | 1151 (55.2%) | 409 (57.6%) | 742 (54.0%) |  |  |
 Surinamese/Antilles | 303 (14.5%) | 124 (17.5%) | 189 (13.8%) |  |  |
 Moroccan | 145 (7.0%) | 44 (6.2%) | 101 (7.4%) |  |  |
 Turkish | 82 (4.0%) | 22 (3.1%) | 60 (4.4%) |  |  |
 Other non-western | 243 (11.7%) | 78 (11.0%) | 165 (12.0%) |  |  |
 Other western | 160 (7.7%) | 43 (6.1%) | 117 (8.5%) |  |  |