Table 2.
AUC with 95% confidence intervals for the validation and external test set for each prediction model, benchmarked against models based on TNM-stage and performance status.
| Outcome | Validation Set |
External Test Set |
|||
|---|---|---|---|---|---|
| AUC | 95% CI | AUC | 95% CI | ||
| RFS | Our prediction model | 0·682 | 0·575–0·788 | 0·681 | 0·597–0·766 |
| TNM based model | 0·650 | 0·541–0·760 | 0·695 | 0·616–0·774 | |
| PS based model | 0·464 | 0·363–0·565 | 0·499 | 0·418–0·58 | |
| Recurrence | Our prediction model | 0·687 | 0·582–0·793 | 0·722 | 0·635–0·810 |
| TNM based model | 0·670 | 0·563–0·777 | 0·707 | 0·622–0·791 | |
| PS based model | 0·506 | 0·402–0·609 | 0·584 | 0·503–0·665 | |
| OS | Our prediction model | 0·759 | 0·663–0·855 | 0·717 | 0·634–0·800 |
| TNM based model | 0·649 | 0·541–0·756 | 0·665 | 0·579–0·751 | |
| PS based model | 0·459 | 0·357–0·561 | 0·531 | 0·447–0·615 | |