Skip to main content
. 2019 Aug 30;19:863. doi: 10.1186/s12885-019-6074-6

Table 2.

Multivariable covariate-adjusted Cox models for colorectal cancer-specific survival

Patient characteristics Colorectal cancer–specific HRa (95% CI)b Colorectal cancer–specific P valuec
Prior cancer diagnosis (vs none)
 Prior non-leukemic malignancies .930 (.909 to .952) <.001
 Prior leukemias 1.263 (1.116 to 1.430) <.001
Age (vs 20–65), y
 < 20 .931 (.748 to 1.158) .531
 > 65 1.376 (1.361 to 1.392) <.001
Sex (vs female)
 Male 1.124 (1.112 to 1.136) <.001
Race (vs white)
 Black 1.135 (1.116 to 1.154) <.001
 Others .872 (.854 to .890) <.001
Marital status (vs single)
 Married .895 (.880 to .910) <.001
 Separated 1.116 (1.066 to 1.168) <.001
 Divorced .942 (.920 to .965) <.001
 Widowed 1.108 (1.087 to 1.129) <.001
Stage (VS Localized)
 Regional 2.983 (2.937 to 3.029) <.001
 Distant 13.548 (13.331 to 13.769) <.001
Grade (Vs Well Differentiated)
 Moderately Differentiated 1.153 (1.132 to 1.174) <.001
 Poorly Differentiated 1.668 (1.634 to 1.701) <.001
 Undifferentiated and Anaplastic 1.904 (1.828 to 1.983) <.001
Radiation (Vs No)
 Yes 1.019 (1.004 to 1.034) .011
Surgery (Vs No)
 Yes .434 (.427 to .442) <.001
 Ablation 1.058 (.918 to 1.219) .435

athis number represent the hazard ratio for all cause and colorectal cancer specific death for the above coverables. All statistical tests were tow-sided

bthis represents confidence interval

cTwo-sided P value was calculated from multivariable covariate-adjusted Cox models