Dynamic risk assessment by landmark modeling of the restricted residual mean survival time.
- 1Division of Biostatistics, Department of Preventive Medicine, University of Tennessee Health Science Center, Memphis, TN, USA.
- 2Department of Biostatistics, Epidemiology and Informatics, University of Pennsylvania, Perelman School of Medicine, Philadelphia, PA, USA. [email protected].
Abstract
Dynamic risk assessment is an important tool for healthcare providers to inform treatment selection and optimize patient outcomes. Frequent evaluation of clinical status provides invaluable information on adverse event risk and/or treatment urgency, based on the changes in predictive biomarkers. Landmark analysis is a useful approach for dynamic risk assessment that obviates the complexity of jointly modeling longitudinal covariates and time-to-event outcome. Traditional landmark methods often rely on the proportional hazards assumption and assume independent censoring. We propose novel methods to broaden the scope of landmark analysis by directly modeling the restricted residual mean survival time (RRMST) at each landmark date while accommodating dependent censoring. Advantages of RMST models include removing the assumption that the (hazard) model is correct at every time point; RMST models instead consider a single capped survival time. Moreover, investigators may prefer the area under the survival curve to the hazard function as a metric in terms of interpretability. We derive the asymptotic properties of our proposed estimators and validate their performance in finite samples through extensive simulations. We apply the proposed methods to analyze outcomes under natural end-stage liver disease (ESLD) history (i.e., in the absence of liver transplantation).
Related
- Strong Simultaneous Inference for Restricted Mean Survival Time Processes in Multi-Arm Trials under Covariate-Adaptive Randomization
- Direct Learning of Treatment-Benefit Rankings for Restricted Mean Survival Time in Randomized Trials with Censoring
- Mixed-effects two-stage residual inclusion methods for individual patient data meta-analysis: A methodological framework for causal inference in survival analysis
- A Residual Tree Gaussian Process Modeling Framework for High-Dimensional Data
- SPARK: A General Goodness-of-Fit Assessment via Residual Projection
BibTeX
@article{zhang2026dynamic,
title = {{Dynamic risk assessment by landmark modeling of the restricted residual mean survival time.}},
author = {Yuan Zhang and Douglas E Schaubel},
journal = {Lifetime data analysis},
year = {2026},
month = oct,
doi = {10.1007/s10985-026-09731-z},
url = {https://europepmc.org/article/MED/42900447},
note = {CC BY},
}
RIS
TY - JOUR
TI - Dynamic risk assessment by landmark modeling of the restricted residual mean survival time.
AU - Yuan Zhang
AU - Douglas E Schaubel
PY - 2026
DA - 2026/10/20
JO - Lifetime data analysis
DO - 10.1007/s10985-026-09731-z
UR - https://europepmc.org/article/MED/42900447
AB - Dynamic risk assessment is an important tool for healthcare providers to inform treatment selection and optimize patient outcomes. Frequent evaluation of clinical status provides invaluable information on adverse event risk and/or treatment urgency, based on the changes in predictive biomarkers. Landmark analysis is a useful approach for dynamic risk assessment that obviates the complexity of jointly modeling longitudinal covariates and time-to-event outcome. Traditional landmark methods often rely on the proportional hazards assumption and assume independent censoring. We propose novel methods to broaden the scope of landmark analysis by directly modeling the restricted residual mean survival time (RRMST) at each landmark date while accommodating dependent censoring. Advantages of RMST models include removing the assumption that the (hazard) model is correct at every time point; RMST models instead consider a single capped survival time. Moreover, investigators may prefer the area under the survival curve to the hazard function as a metric in terms of interpretability. We derive the asymptotic properties of our proposed estimators and validate their performance in finite samples through extensive simulations. We apply the proposed methods to analyze outcomes under natural end-stage liver disease (ESLD) history (i.e., in the absence of liver transplantation).
SN - 1380-7870
SN - 1572-9249
ER -