Log-rank test survival time and relapse or no-relapse

[plbenveniste]

In the file data_preprocessing/6_survival_primitive_vs_metastasis.py we perform the following log-rank test.

• Null Hypothesis (H₀): There is no difference in survival distributions between the groups (e.g., subjects with relapse vs. without relapse).
• Alternative Hypothesis (H₁): There is a significant difference in survival distributions between the groups.

Two experiements were done :

1. First experiment: if the survival time is NaN then we set it equal to +∞ as we consider that the persone dies after the study
2. Second experiment: if the survival time is NaN then we remove the person from the study

The results of the experiments are very different. Here is the output of the code:

First Strategy: if 'delai_fin_DC' is nan then death occurs at time = + infinite
Log-Rank Test Statistic: 8.211735435498051
Log-Rank Test p-value: 0.004162032598667009

Second Strategy: remove the patients for which 'delai_fin_DC' is nan
Log-Rank Test Statistic: 1.2871655808759024
Log-Rank Test p-value: 0.2565709557686867

[plbenveniste]

To Do:

• Plot the Kaplan Meyer survival plot