Skip to main content
. 2021 Mar 6;3(1):vdab040. doi: 10.1093/noajnl/vdab040

Table 3.

Multivariable Logistic Regression Model for 30-Day Mortality, Glioblastoma, and National Cancer Database 2004–2014

Characteristic Event N OR 95% CI P value
Race ethnicity
 White non-Hispanic 1019
 Black non-Hispanic 64 0.88 0.66–1.17 .4
 Asian non-Hispanic 13 0.52 0.28–0.91 .031
 Hispanic 58 0.84 0.61–1.13 .3
Facility type
 Academic/Research Program 455
 Community Cancer Program 53 1.00 0.72–1.37 >.9
 Comprehensive Community Cancer Program 483 1.10 0.95–1.27 .2
 Integrated Network Cancer Program 163 1.23 1.00–1.51 .049
Age 1.02 1.01–1.02 <.001
Sex
 Female 447
 Male 707 1.27 1.11–1.45 <.001
Primary payer
 Medicaid 56
 Medicare 718 1.09 0.79–1.53 .6
 Not insured 46 1.20 0.77–1.86 .4
 Other government 16 0.76 0.40–1.39 .4
 Private insurance 318 1.08 0.79–1.50 .6
Urban/rural
 Metro 946
 Rural 21 0.87 0.52–1.39 .6
 Urban 187 0.93 0.77–1.11 .4
Charlson–Deyo score
 0 586
 1 310 1.62 1.38–1.91 <.001
 2 142 1.50 1.21–1.85 <.001
 3 116 2.33 1.82–2.96 <.001
Surgical resection
 Biopsy 417
 Gross total 327 0.67 0.57–0.80 <.001
 None 31 1.61 1.01–2.51 .039
 Subtotal 379 0.86 0.73–1.01 .073
Focality
 Multifocal 233
 Unifocal 921 0.73 0.62–0.87 <.001
Radiation
 Not received 1095
 Received 59 0.04 0.03–0.06 <.001
Chemotherapy
 Not received 1074
 Received 80 0.21 0.15–0.27 <.001

OR, odds ratio; CI, confidence interval.