Skip to main content
. 2015 Jun 29;4(4):213–224. doi: 10.2217/cns.15.17

Table 3. . Multivariate cox proportional hazard model of overall survival using demographic and medical factors (n = 666).

Demographic and medical factors Coefficient p-value Hazard ratio 95% CI of hazard ratio
Age ≥65 years 0.309 0.0002 1.362 1.155–1.607

KPS categories: <0.0001
– 0–40 vs 80–100 1.326 <0.0001 3.767 2.899–4.894
– 50–70 vs 80–100 0.494 <0.0001 1.638 1.356–1.979
– 0–40 vs 50–70 <0.0001 2.299 1.821–2.887

Postoperative -0.675 0.0005 0.509 0.347–0.747

Primary cancer site: 0.0025
– GI vs breast 0.728 0.0001 2.070 1.426–3.005
– GU vs breast 0.506 0.0081 1.659 1.141–2.413
– Lung vs breast 0.304 0.0087 1.356 1.080–1.702
– Other vs breast 0.188 0.3216 1.207 0.832–1.751
– Unknown vs breast 0.483 0.0339 1.620 1.037–2.531

Generalized R2 = 20.21%

In the Cox proportional hazard regression analysis, a covariate with a positive coefficient is associated with worse survival prognosis while covariate with a negative coefficient is associated with improved survival prognosis.

p < 0.05.

GI: Gastrointestinal; GU: Genitourinary; KPS: Karnofsky performance status.