Skip to main content
. 2020 Sep 9;3(9):e2015444. doi: 10.1001/jamanetworkopen.2020.15444

Table 2. Univariable and Multivariable Cox Proportional Regression Analysis of Factors Associated With Overall Survival.

Variable Univariable analysis Multivariable analysis
HR (95% CI) P value HR (95% CI) P value
Age at diagnosis, continuous 1.02 (1.02-1.03) .001 1.02 (1.01-1.02) .001
Sex
Men 1 [Reference] NA 1 [Reference] NA
Women 0.88 (0.81-0.96) .003 0.94 (0.86-1.04) .24
Race
White 1 [Reference] NA 1 [Reference] NA
Black 1.07 (0.93-1.24) .36 1.09 (0.93-1.27) .28
Other 0.75 (0.59-0.94) .01 0.81 (0.64-1.04) .09
County-level residents without high school diploma, %a
≥13 1.06 (0.97-1.15) .19 NA NA
<13 1 [Reference] NA NA NA
Median county-level income, $a
≥35 000 1 [Reference] NA 1 [Reference] NA
<35 000 1.14 (1.05-1.23) .003 1.07 (0.98-1.17) .12
Place of residencea
Urban 1 [Reference] NA NA NA
Rural 1.10 (0.85-1.43) .46 NA NA
Hospital type
Academic 1 [Reference] NA 1 [Reference] NA
Community 1.36 (1.25-1.49) .001 1.27 (1.16-1.38) .001
Insurance statusa
Insured 1 [Reference] NA NA NA
Not insured 1.08 (0.89-1.32) .42 NA NA
Charlson/Deyo score
0 1 [Reference] NA 1 [Reference] NA
≥1 1.23 (1.12-1.34) .001 1.15 (1.05-1.26) .004
Year of diagnosisa
2014-2016 1 [Reference] NA NA NA
2010-2013 1.03 (0.94-1.13) .54 NA NA
Type of cancer
Kidney 1 [Reference] NA 1 [Reference] NA
Breast 0.92 (0.81-1.04) .18 0.96 (0.83-1.12) .60
NSCLC 0.74 (0.66-0.83) .001 0.80 (0.71-0.90) .001
CRC 1.33 (1.14-1.56) .001 1.43 (1.21-1.69) .001
Melanoma 1.86 (1.62-2.14) .001 1.97 (1.70-2.29) .001
Chemotherapy
Yes 0.61 (0.56-0.66) <.001 0.68 (0.62-0.75) .001
No 1 [Reference] NA 1 [Reference] NA
Radiation therapy
Yes 0.79 (0.72-0.87) <.001 0.94 (0.85-1.04) .21
No 1 [Reference] NA 1 [Reference] NA
Immunotherapy
Yes 0.73 (0.60-0.88) <.001 0.62 (0.51-0.76) .001
No 1 [Reference] NA 1 [Reference] NA

Abbreviations: CRC, colorectal cancer; HR, hazard ratio; NA, not applicable; NSCLC, non–small cell lung cancer.

a

These variables were not included in the multivariable analysis because they had P > .15 in the univariable analysis.