Efficacy of a machine learning model integrating clinical-psychosocial factors in predicting posttraumatic bone nonunion and bone defects

Table 1 Baseline data comparison (nonunion vs healing), n (%)

Indicators	Nonunion group (n = 17)	Healing group (n = 128)	t/χ2	P value
Age (years)	58.24 ± 11.14	46.29 ± 14.76	3.213	0.002
Gender			0.024	0.878
Male	12 (70.59)	88 (68.75)
Female	5 (29.41)	40 (31.25)
Body mass index (kg/m2)	25.19 ± 4.10	23.70 ± 2.80	5.993	< 0.001
Comorbidity
Hypertension	8 (47.06)	55 (42.97)	0.102	0.749
Diabetes	7 (41.18)	21 (16.41)	5.909	0.015
Atherosclerosis	5 (29.41)	20 (15.63)	1.999	0.157
Open fracture	10 (58.8)	28 (21.9)	10.590	0.001
Bone defect length (mm)	51.82 ± 14.43	38.44 ± 12.01	7.362	< 0.001
Distraction start time (days)	12.59 ± 5.81	10.20 ± 3.49	2.420	0.017
Distraction period (days)	45.24 ± 10.65	42.78 ± 9.72	0.968	0.335
Bone mineralization time (days)	275.18 ± 102.43	191.45 ± 87.24	3.639	0.004
Depression (PHQ-9)	16.82 ± 4.25	11.28 ± 3.10	11.374	< 0.001
Anxiety (GAD-7)	15.24 ± 4.12	10.51 ± 2.40	6.908	< 0.001
Social support (MSPSS)	48.53 ± 9.98	58.83 ± 10.53	3.810	< 0.001
Current smoker	9 (52.9)	35 (27.3)	4.652	0.031
Low-income	11 (64.7)	48 (37.5)	4.603	0.032

PHQ-9: Patient Health Questionnaire-9; GAD-7: Generalized Anxiety Disorder-7; MSPSS: Multidimensional Scale of Perceived Social Support.