Incidence and survival analyses for occult lung cancer between 2004 and 2015: a population-based study

BMC Cancer

Table 4 Univariable and multivariable Cox regression analyses for prognostic factors

	N	5-year CSS	Univariable analysis		Multivariable analysis
	N	5-year CSS	HR	P-Value	HR	95% CI	P-Value
T classification				< 0.001			< 0.001
T1a	1752	75%	1		1	reference
T1b	9439	67%	1.230	< 0.001	1.113	1.010–1.227	0.030
T1c	8412	56%	1.788	< 0.001	1.379	1.252–1.519	< 0.001
T2a	11,509	47%	2.345	< 0.001	1.889	1.719–2.076	< 0.001
T2b	3784	37%	3.079	< 0.001	2.172	1.964–2.402	< 0.001
T3	4180	31%	3.759	< 0.001	2.510	2.273–2.772	< 0.001
T4	11,043	19%	5.555	< 0.001	3.178	2.892–3.493	< 0.001
Tx (occult)	2353	15%	6.344	< 0.001	2.624	2.365–2.910	< 0.001
Subgroup comparison
Tx vs. T2b	–	–	2.020	< 0.001	1.186	1.104–1.273	< 0.001
Tx vs. T3	–	–	1.648	< 0.001	1.054	0.986–1.127	0.121
Tx vs. T4	–	–	1.127	< 0.001	0.845	0.801–0.891	< 0.001

HR hazard ratio, CI confidence interval
Cox regression’s method was Enter selection
The results of multivariable analysis were adjusted for other confounding factors, such as sex, age, tumor differentiation, radiotherapy, chemotherapy, surgical treatment, histological subtypes, marital status, tumor location and race/ ethnicity

ISSN: 1471-2407