Skip to main content

Table 4 Hazard relationships for dichotomous age and z

From: Statistical learning methods as a preprocessing step for survival analysis: evaluation of concept using lung cancer data

Model

Age Hazard Ratio

Az

Accepted

  

Dichotomous Age

1.72 (1.02, 2.90)

0.5792

Grade adjusted

1.78 (1.06, 3.02)

0.606

Gender adjusted

1.64 (0.96, 2.78)

0.669

Grade Gender adjusted

1.68 (0.99, 2.85)

0.677

Model

z Hazard Ratio

Az

Hybrid

  

Dichotomous z

0.25 (0.14, 0.47)

0.691

Gender adjusted

0.28 (0.15, 0.53)

0.738

  1. For the age and z variables, two groups were formed using the respective distribution median as the cut-point and compared. The hazard ratios (HRs) are provided parenthetically with 95% confidence intervals. The area under the receiver operating characteristic curves (Azs) derived from Cox regression models are also provided. Because age and z translate inversely with respect to hazard, increased age confers a greater hazard while decreased z confers a greater hazard. To make HR comparisons of z with age, the reciprocal of the z HR is required.