Table 1. Baseline demographic and clinical characteristics of variables used for prediction.
Overall | No | Yes | P-value | |
---|---|---|---|---|
n | 1962 | 1642 | 320 | |
Sex = Male (%) | 778 (39.7) | 702 (42.8) | 76 (23.8) | <0.001 |
Age (mean (SD)) | 18.36 (3.57) | 18.52 (3.64) | 17.52 (3.11) | <0.001 |
History of self-harm = Yes (%) | 883 (45.0) | 616 (37.5) | 267 (83.4) | <0.001 |
History of suicide ideation = Yes (%) | 953 (48.6) | 731 (44.5) | 222 (69.4) | <0.001 |
SOFAS (mean (SD)) | 62.43 (9.02) | 62.75 (9.06) | 60.76 (8.62) | <0.001 |
NEET = Yes (%) | 307 (15.6) | 264 (16.1) | 43 (13.4) | 0.269 |
Clinical stage (%) | <0.001 | |||
Stage 1a | 604 (30.8) | 541 (32.9) | 63 (19.7) | |
Stage 1b | 1207 (61.5) | 973 (59.3) | 234 (73.1) | |
Stage 2 and above | 151 (7.7) | 128 (7.8) | 23 (7.2) | |
Depression (%) | <0.001 | |||
No | 598 (30.5) | 538 (32.8) | 60 (18.8) | |
Full-threshold | 633 (32.3) | 494 (30.1) | 139 (43.4) | |
Sub-threshold | 731 (37.3) | 610 (37.1) | 121 (37.8) | |
Anxiety (%) | 0.149 | |||
No | 587 (29.9) | 483 (29.4) | 104 (32.5) | |
Full-threshold | 624 (31.8) | 515 (31.4) | 109 (34.1) | |
Sub-threshold | 751 (38.3) | 644 (39.2) | 107 (33.4) | |
Bipolar disorder (%) | 0.957 | |||
No | 1780 (90.7) | 1490 (90.7) | 290 (90.6) | |
Full-threshold | 71 (3.6) | 60 (3.7) | 11 (3.4) | |
Sub-threshold | 111 (5.7) | 92 (5.6) | 19 (5.9) | |
Psychosis (%) | 0.204 | |||
No | 1862 (94.9) | 1553 (94.6) | 309 (96.6) | |
Full-threshold | 44 (2.2) | 41 (2.5) | 3 (0.9) | |
Sub-threshold | 56 (2.9) | 48 (2.9) | 8 (2.5) | |
Psychosis-like experiences = Yes (%) | 356 (18.1) | 282 (17.2) | 74 (23.1) | 0.014 |
Mania-like experiences = Yes (%) | 298 (15.2) | 240 (14.6) | 58 (18.1) | 0.13 |
Circadian disturbance = Yes (%) | 294 (15.0) | 244 (14.9) | 50 (15.6) | 0.791 |
History of hospitalisation = Yes (%) | 346 (17.6) | 265 (16.1) | 81 (25.3) | <0.001 |
Any childhood disorders = No (%) | 1691 (86.2) | 1404 (85.5) | 287 (89.7) | 0.058 |
Childhood autism spectrum = Yes (%) | 56 (2.9) | 52 (3.2) | 4 (1.2) | 0.089 |
Childhood ADHD = Yes (%) | 104 (5.3) | 94 (5.7) | 10 (3.1) | 0.078 |
Childhood generalised anxiety = Yes (%) | 41 (2.1) | 32 (1.9) | 9 (2.8) | 0.439 |
Childhood depression = Yes (%) | 30 (1.5) | 24 (1.5) | 6 (1.9) | 0.762 |
Family history—bipolar = Yes (%) | 154 (7.8) | 135 (8.2) | 19 (5.9) | 0.202 |
Family history—psychosis = Yes (%) | 82 (4.2) | 69 (4.2) | 13 (4.1) | 1 |
Family history—suicide = Yes (%) | 35 (1.8) | 28 (1.7) | 7 (2.2) | 0.715 |
Family history—depression = Yes (%) | 653 (33.3) | 542 (33.0) | 111 (34.7) | 0.604 |
Family history—anxiety = Yes (%) | 302 (15.4) | 248 (15.1) | 54 (16.9) | 0.472 |
Family history—alcohol = Yes (%) | 171 (8.7) | 134 (8.2) | 37 (11.6) | 0.062 |
Family history—substance misuse = Yes (%) | 140 (7.1) | 118 (7.2) | 22 (6.9) | 0.937 |
Any physical health problems = No (%) | 1598 (81.4) | 1332 (81.1) | 266 (83.1) | 0.444 |
Endocrine problem = Yes (%) | 64 (3.3) | 50 (3.0) | 14 (4.4) | 0.292 |
Metabolic problem = Yes (%) | 35 (1.8) | 31 (1.9) | 4 (1.2) | 0.577 |
Neurological problem = Yes (%) | 55 (2.8) | 49 (3.0) | 6 (1.9) | 0.36 |
Physical health problems—Other = Yes (%) | 257 (13.1) | 222 (13.5) | 35 (10.9) | 0.245 |
Treatment antidepressants = Yes (%) | 737 (37.6) | 606 (36.9) | 131 (40.9) | 0.194 |
Treatment antipsychotics = Yes (%) | 239 (12.2) | 195 (11.9) | 44 (13.8) | 0.398 |
Treatment mood stabilisers = Yes (%) | 124 (6.3) | 102 (6.2) | 22 (6.9) | 0.749 |
Treatment stimulants = Yes (%) | 166 (8.5) | 152 (9.3) | 14 (4.4) | 0.006 |
Treatment psychological therapy = Yes (%) | 1101 (56.1) | 927 (56.5) | 174 (54.4) | 0.532 |
Note. Columns present overall statistics and then statistics split by whether or not patients exhibited self-harm within 6 months of presentation (No vs Yes). P-value is reported for univariate comparisons between downstream self-harm. For categorical variables, the Chi-square test is performed and for continuous variables, the t-test is performed. Bolded text indicates best significant results at P<0.001.