Risk threshold reporting across cohorts and the use of reclassification metrics in the BLISS prognostic score

AI Briefing

• – The authors suggest that explicit reporting of threshold-specific performance in external validation cohorts would enhance the transparency and clinical utility of the BLISS score, and that reporting net reclassification improvement (NRI) and integrated discrimination improvement (IDI) would provide additional insight into the relative performance of the BLISS score versus the Bertens' score.
• – The authors request that the authors provide the additional analyses or comment on their feasibility, highlighting the importance of assessing the model's transportability.
• – The authors emphasize the need for explicit reporting of threshold-specific performance in external validation cohorts and the use of reclassification metrics, such as NRI and IDI, to evaluate the BLISS score's performance in different settings.