Smooth Concordance Metrics for Survival Models
Abstract
Concordance indices are widely popular metrics for assessing the ability of predictive survival models to discriminate underlying risk levels. However, these statistics have also been criticized for using only the rank orderings of the model's predicted risk scores and being insensitive to important model features, such as the addition of strong predictor variables into the model. In this paper, we address these limitations by developing smooth concordance metrics that model the underlying risk discrimination probabilities as continuous functions of the predicted risk score differences, where the shapes of these functions are estimated from the observed data. As a result, these smooth concordance metrics assess model performance across the entire range of possible risk score differences, allowing one to identify specific scenarios where the candidate model performs especially well or better than other models. Simulations show that the proposed smooth concordance metrics provide more detailed information about risk discrimination performance and are much more sensitive to the addition of meaningful predictors. We apply these methods to compare predictive survival models for cancer recurrence.
Related
- The Conway knot has infinite order in smooth concordance
- Psh variation of singular Kähler-Einstein metrics and Invariance of plurigenera for smooth Kähler families
- On the Definition and Smooth Examples of Weakly Kähler--Finsler Metrics
- Failure of local extension of semipositive metrics in a smooth projective family
- A Model-threshold Dimension Bound and Sharp Critical Ends for Smooth Singular Sets of Constant Positive σk-curvature Metrics
BibTeX
@misc{hartman2026smooth,
title = {{Smooth Concordance Metrics for Survival Models}},
author = {Nicholas Hartman and Grace Richards},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2606.06411},
eprint = {2606.06411},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2606.06411},
}
RIS
TY - UNPB
TI - Smooth Concordance Metrics for Survival Models
AU - Nicholas Hartman
AU - Grace Richards
PY - 2026
DA - 2026/11/05
PB - arXiv
DO - 10.48550/arxiv.2606.06411
UR - https://arxiv.org/abs/2606.06411
AB - Concordance indices are widely popular metrics for assessing the ability of predictive survival models to discriminate underlying risk levels. However, these statistics have also been criticized for using only the rank orderings of the model's predicted risk scores and being insensitive to important model features, such as the addition of strong predictor variables into the model. In this paper, we address these limitations by developing smooth concordance metrics that model the underlying risk discrimination probabilities as continuous functions of the predicted risk score differences, where the shapes of these functions are estimated from the observed data. As a result, these smooth concordance metrics assess model performance across the entire range of possible risk score differences, allowing one to identify specific scenarios where the candidate model performs especially well or better than other models. Simulations show that the proposed smooth concordance metrics provide more detailed information about risk discrimination performance and are much more sensitive to the addition of meaningful predictors. We apply these methods to compare predictive survival models for cancer recurrence.
ER -