A Proposal of a Personalized Surveillance Strategy for Gastric Cancer: A Retrospective Analysis of 9191 Patients

Table 2

Univariate and multivariate Cox hazards regression analyses of independent prognosis factors.

	Univariate analysis		value	Multivariate analysis		value
	HR	95% CI		HR	95% CI
Race
White	Ref
Black	0.981	0.912-1.0154	0.599
Other	0.684	0.640-0.730	<0.001
Unknown	0.722	0.375-1.389	0.329
Sex
Male	Ref
Female	0.967	0.917-1.020	0.217
Age
≤60	Ref
>60	1.414	1.341-1.490	<0.001	1.724	1.621-1.833	<0.001
Grade
Well differentiated	Ref
Moderately differentiated	1.688	1.400-2.035	<0.001	1.083	0.895-1.308	0.409
Poorly differentiated	2.299	1.918-2.755	<0.001	1.254	1.044-1.508	0.016
Undifferentiated	2.303	1.814-2.924	<0.001	1.202	0.944-1.530	0.135
Unknown	1.633	1.286-2.074	<0.001	1.062	0.835-1.351	0.625
Tumor size
≤.5 cm	Ref
>4.5 cm	1.836	1.741-1.936	<0.001	1.097	1.037-1.160	0.001
TNM stage
IA	Ref
IB	2.005	1.691-2.377	<0.001	1.918	1.617-2.276	<0.001
IIA	3.039	2.627-3.516	<0.001	2.908	2.507-3.374	<0.001
IIB	4.942	4.290-5.693	<0.001	4,660	4.031-5.386	<0.001
IIIA	6.613	5.754-7.602	<0.001	6.455	5.592-7.451	<0.001
IIIB	8.414	7.357-9.624	<0.001	8.846	7.685-10.183	<0.001
IIIC	10.985	9.572-12.607	<0.001	11.944	10.324-13.819	<0.001
Number of LNs examined
≤15	Ref
>15	0.863	0.818-0.911	<0.01	0.664	0.628-0.702	<0.001

CI: confidence interval; HR: hazard risk; Ref: reference category; LNs: lymph nodes. American Indian/AK native or Asian/Pacific Islander.