Table 2.
Tabulation of possible baseline predictors for the training and test data
| Predictor | Training data (n of spells, %) |
Test data (n of spells,%) |
| Age (years) | ||
| 18–30 | 5421 (12.1) | 2339 (12.2) |
| 31–40 | 9626 (21.5) | 4074 (21.2) |
| 41–50 | 13 145 (29.3) | 5624 (29.3) |
| 51–57 | 8762 (19.5) | 3811 (19.8) |
| 58–64 | 7876 (17.6) | 3367 (17.5) |
| Sex | ||
| Women | 24 791 (55.3) | 10 647 (55.4) |
| Men | 20 042 (44.7) | 8568 (44.6) |
| Geographical region | ||
| Northern | 6306 (14.1) | 2688 (14.0) |
| Middle | 6651 (14.8) | 2741 (14.3) |
| Stockholm/Gotland | 8750 (19.5) | 3716 (19.3) |
| Western | 13 264 (29.6) | 5776 (30.1) |
| Southern | 9862 (22.0) | 4294 (22.3) |
| Educational level | ||
| Elementary (≤9) | 8779 (19.6) | 3783 (19.7) |
| High school (10–12) | 9110 (20.3) | 3856 (20.1) |
| University/college (>12) | 26 944 (60.0) | 11 576 (60.2) |
| Family composition | ||
| Living with partner and children | 18 015 (40.2) | 7927 (41.3) |
| Single with children at home | 4776 (10.7) | 1946 (10.1) |
| Living with partner without children | 7302 (16.3) | 3112 (16.2) |
| Single without children | 14 740 (32.9) | 6230 (32.4) |
| Country of birth | ||
| Sweden | 34 379 (76.7) | 14 730 (76.7) |
| Other Nordic | 1551 (3.5) | 680 (3.5) |
| Other EU25 | 1262 (2.8) | 525 (2.8) |
| Rest of the word | 7641 (17.0) | 3260 (17.0) |
| Occupational status | ||
| Employed | 41 711 (93.0) | 17 831 (92.8) |
| Unemployed | 2463 (5.5) | 1097 (5.7) |
| Parental leave/student | 659 (1.5) | 287 (1.5) |
| Part-time disability pension at start of the SA spell | ||
| Yes | 2781 (6.2) | 1144 (6.0) |
| No | 42 052 (93.8) | 18 071 (94.0) |
| SA extent at the start of the SA spell | ||
| 100% | 36 491 (81.4) | 15 698 (81.7) |
| 75% | 1154 (25.7) | 476 (2.5) |
| 50% | 5943 (13.2) | 2504 (13.0) |
| 25% | 1245 (27.8) | 537 (2.8) |
| Multimorbidity* | ||
| Yes | 11 440 (25.5) | 4861 (25.3) |
| No | 33 393 (74.5) | 14 354 (74.7) |
| SA days in the 365 days before the SA spell | ||
| 0 | 31 161 (70.5) | 13 412 (69.8) |
| >0 and ≤90 | 10 612 (23.7) | 4516 (23.5) |
| >90 and ≤180 | 1718 (3.8) | 704 (3.7) |
| >180 | 1342 (3.0) | 583 (3.0) |
| Number of days in inpatient care during the 365 days before the SA spell† | ||
| 0 | 40 935 (91.3) | 17 586 (91.5) |
| 1–2 | 16 463 (3.7) | 711 (3.7) |
| >2 | 2255 (5.1) | 918 (4.8) |
| Number of visits in specialised outpatient care during the 365 days before the SA spell | ||
| 0 | 24 072 (53.7) | 10 370 (54.0) |
| 1–2 | 13 795 (30.8) | 5821 (30.3) |
| >2 | 6966 (15.5) | 3024 (15.7) |
| The SA spell was initiated in primary healthcare | ||
| Yes | 38 308 (85.4) | 16 508 (85.9) |
| No | 6525 (14.6) | 2707 (14.1) |
*Any inpatient or specialised outpatient healthcare during the period of 4 days before and 8 days after the start of the SA spell.
†Excluding codes O80 and Z00-Z99 except Z73.0 at least three different types of medication (different ATC codes) at least once.
ATC, Anatomical Therapeutic Chemical; SA, sickness absence.