Skip to main content

Table 2 Multivariate Cox-regression model and bootstrapping

From: Does hyperthermic intraoperative chemotherapy lead to improved outcomes in patients with ovarian cancer? A single center cohort study in 111 consecutive patients

 

Cox regression model A)

Bootstrap B)

 

P C)

HR

95.0% CI for HR

Selection (%)

HR >  1 (%)

Age [years]

0.541

1.016

0.966

1.069

94.75

100.00

Systemic chemotherapy

0.038

0.279

0.084

0.932

90.32

0.00

Recurrent disease

0.349

0.558

0.163

1.908

30.83

15.21

FIGO IV vs. III/II

0.964

0.964

0.199

4.676

21.60

63.29

HIPEC

0.830

0.881

0.275

2.820

89.19

0.03

  1. Analysis limited to 52 patients with CC0 score
  2. A) Full Cox-regression model. Additionally, backward and forward variable selections identified systemic chemotherapy as the only independent significant predictor for survival
  3. B) Bootstrapping procedure for backward variable selection process with 9999 samples. Column selection expresses the fraction indicating how often the variable was selected in the backward variable selection process and column HR > 1 expresses the fraction of a hazard ratio above 1.0
  4. C) Likelihood ratio test