Table 2.
Characteristics of the patients in the training, validation, and test datasets for function
| Parameter | Complete dataset (n = 1469) | % missing | Training dataset (n = 883) | Validation dataset (n = 293) | Test dataset (n = 293) |
| Age in years | 61 ± 8 | 0 | 61 ± 7 | 60 ± 8 | 61 ± 8 |
| Gender, women | 79 (1167) | 0 | 78 (689) | 81 (236) | 83 (242) |
| Duration of symptoms in months | 37 ± 46 | 2 | 38 ± 36 | 38 ± 36 | 34 ± 32 |
| Second opinion | 89 (1305) | 0 | 89 (789) | 88 (258) | 88 (258) |
| Hand dominance | 0 | ||||
| Right | 85 (1251) | 85 (748) | 89 (260) | 83 (243) | |
| Left | 10 (141) | 10 (90) | 7 (21) | 10 (30) | |
| Both | 5 (77) | 5 (45) | 4 (12) | 7 (20) | |
| Dominant hand treated | 47 (686) | 0 | 49 (428) | 46 (136) | 42 (122) |
| Occupational intensity | 0 | ||||
| Not employed | 47 (691) | 49 (428) | 44 (130) | 45 (133) | |
| Light | 19 (277) | 18 (156) | 20 (58) | 22 (63) | |
| Moderate | 22 (329) | 23 (199) | 20 (58) | 25 (72) | |
| Heavy | 12 (172) | 11 (100) | 16 (47) | 9 (25) | |
| BMI in kg/m2 | 26.6 ± 3.9 | 36 | 26.6 ± 4.0 | 26.4 ± 3.6 | 26.9 ± 3.9 |
| Smoking | 45 | ||||
| Never | 24 (351) | 25 (218) | 20 (58) | 26 (75) | |
| Disease severity | |||||
| Preoperative MHQ pain score | 34.2 ± 14.0 | 0.3 | 33.8 ± 14.1 | 34.5 ± 14.0 | 35.3 ± 13.8 |
| Preoperative MHQ function score | 48.9 ± 17.0 | 0 | 49.1 ± 17.2 | 47.9 ± 16.1 | 49.2 ± 17.1 |
| Medical history | 36 | ||||
| Diabetes | 3 (51) | 3 (30) | 3 (9) | 4 (12) | |
| Cardiovascular system | 7 (104) | 7 (62) | 7 (20) | 8 (22) | |
| Thrombosis/vasculitis | 1 (13) | 1 (10) | 0.3 (1) | 1 (2) | |
| Respiratory system | 8 (118) | 8 (70) | 7 (21) | 9 (27) | |
| Liver/kidneys | 1 (12) | 1 (8) | 1 (2) | 1 (2) | |
| Cranial nerves | 2 (22) | 2 (15) | 2 (5) | 1 (2) | |
| Locomotor system | 23 (337) | 24 (208) | 19 (55) | 25 (74) | |
| Rheumatic disorders | 17 (243) | 18 (158) | 13 (37) | 16 (48) | |
| Hemorrhoids/varicosities | 11 (162) | 11 (101) | 11 (33) | 10 (28) | |
| Allergies | 17 (249) | 17 (149) | 16 (47) | 18 (53) | |
| Hematomas | 3 (49) | 3 (30) | 4 (13) | 2 (6) |
Data presented as mean ± SD or % (n); the training dataset was used to select the predictive variables and to train our models; the performance of all models was evaluated in the validation dataset, after which one model was selected for further evaluation; performance of this final model was evaluated on the test data set.