An Artificial Intelligence Model for Predicting 1-Year Survival of Bone Metastases in Non-Small-Cell Lung Cancer Patients Based on XGBoost Algorithm

Table 2

Univariate analysis and multivariate logistic analysis based on all variables for 1-year survival (training cohort).

Characteristics	Univariate analysis	Multivariate logistic analysis
	value	HR (95% CI)	value
Age	<0.001	1.015 (1.008–1.022)	<0.001
Size	<0.001	1.008 (1.004–1.011)	<0.001
Race
Black	<0.001	Reference
Other		0.423 (0.311–0.575)	<0.001
White		1.073 (0.851–1.352)	0.552
Sex
Female	<0.001	Reference
Male		1.265 (1.088–1.470)	<0.05
Primary site
Main bronchus	<0.05
Overlapping lesion of lung
Lung, NOS
Lobe
Histologic type
ADC	<0.001	Reference
Others		1.746 (1.351–2.255)	<0.001
SCC		1.524 (1.246–1.865)	<0.001
Grade
I	<0.001	Reference
II		1.037 (0.751–1.433)	0.824
III		1.653 (1.206-2.266)	<0.05
IV		2.454 (1.292-4.659)	<0.05
Laterality
Left—origin of primary	0.752
Right—origin of primary
T stage
T1	<0.001
T2
T3
T4
N stage
N0	<0.001	Reference
N1		1.198 (0.901-1.592)	0.215
N2		1.636 (1.351-1.982)	<0.001
N3		1.816 (1.443-2.284)	<0.001
M stage
M1a	0.791
M1b
Radiotherapy
No	0.162
Yes
Surgery
No	<0.001	Reference
Yes		0.438 (0.311-0.617)	<0.001
Chemotherapy
No	<0.001	Reference
Yes		0.211 (0.174 -0.256)	<0.001
Brain metastasis
No	0.866
Yes
Liver metastasis
No	<0.001	Reference
Yes		1.948 (1.589-2.388)	<0.001
Lung metastasis
No	0.949
Yes
Insurance status
Insured	0.732
Uninsured
Marital status
Married	<0.001
Unmarried

ADC: adenocarcinoma; SCC: squamous cell carcinoma.