Table 2.
Validation | Test | Combined | ||||
---|---|---|---|---|---|---|
C-index | p-value | C-index | p-value | C-index | p-value | |
CLN | 0.537 (0.406, 0.668) | 0.400 | 0.622 (0.500, 0.744) | 0.020 | 0.585 | 0.002 |
Rad-S | 0.692 (0.569, 0.815) | 0.010 | 0.608 (0.504, 0.712) | 0.050 | 0.645 | 0.013 |
RadCLN-S | 0.697 (0.574, 0.820) | < 0.001 | 0.638 (0.536, 0.740) | < 0.001 | 0.662 | 0.075 |
Rad-D | 0.748 (0.664, 0.832) | 0.002 | 0.681 (0.573, 0.789) | 0.020 | 0.729 | 0.310 |
RadCLN-D | 0.752 (0.660, 0.844) | < 0.001 | 0.695 (0.581, 0.809) | 0.004 | 0.734 | – |
Combined C-index is the weighted score calculated from the validation and the test set using the survcomp package in R
p-values of validation and test C-index indicate the significance of prognostic prediction (the log-rank test)
Combined p-value indicates the significance of the difference between the certain model and RadCLN-D (the student t-test)