Calibrated Transfer Learning for Survival Prediction and Treatment Inference
Abstract
Historical survival data can improve prediction in a cohort with few events, but prognostic differences can make borrowing harmful. To borrow while adapting, we propose calibrated transfer learning for the Cox model, in which a historical risk function enters the target analysis as a fixed offset with a prespecified correction for changing covariate effects and a separate baseline hazard. Calibration removes historical error in the correction space exactly, so only the remaining error governs the prediction rate, for any bounded learner. Accurate prediction does not, however, ensure valid treatment inference. In randomized trials, the remaining error biases the conditional treatment coefficient, because a treatment effect changes the prognostic composition of later risk sets. We characterize this bias and its sharp worst-case direction, and we introduce a cross-fitted projected score that removes it at first order and leaves only a second-order bias when a fixed error lies in its dictionary. A historical learning bound separates approximation, sampling, and optimization error. In simulations, calibration improves prediction when the historical risk is well learned, and the projected score restores interval coverage under drift outside the correction, also at a hazard ratio of 0.7. In a retrospective transfer between the POPLAR and OAK lung cancer trials, selected corrections reduce the integrated Brier score by 4.3\% relative to ridge-penalized clinical Cox with 100 training patients, without improving treatment precision.
Related
- Source projection for calibrated transfer learning in survival prediction
- Covariance-Calibrated Prediction-Powered Inference
- Calibrated Bayesian inference for overall and subgroup-specific causal effects in stepped-wedge cluster-randomized trials
- Locally calibrated and mesh-free inference for spatial point distributions: closed-form null, contamination law, and detectability threshold
- SCOPE Before You Sample: Conformal Scores Guide Posterior Exploration for Fast and Calibrated Bayesian Predictive Inference
BibTeX
@misc{wang2026calibrated,
title = {{Calibrated Transfer Learning for Survival Prediction and Treatment Inference}},
author = {Shuoyang Wang and Wanyu Zhang and Lingli Yang and Kexuan Li},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.21677},
eprint = {2610.21677},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.21677},
}
RIS
TY - UNPB
TI - Calibrated Transfer Learning for Survival Prediction and Treatment Inference
AU - Shuoyang Wang
AU - Wanyu Zhang
AU - Lingli Yang
AU - Kexuan Li
PY - 2026
DA - 2026/10/15
PB - arXiv
DO - 10.48550/arxiv.2610.21677
UR - https://arxiv.org/abs/2610.21677
AB - Historical survival data can improve prediction in a cohort with few events, but prognostic differences can make borrowing harmful. To borrow while adapting, we propose calibrated transfer learning for the Cox model, in which a historical risk function enters the target analysis as a fixed offset with a prespecified correction for changing covariate effects and a separate baseline hazard. Calibration removes historical error in the correction space exactly, so only the remaining error governs the prediction rate, for any bounded learner. Accurate prediction does not, however, ensure valid treatment inference. In randomized trials, the remaining error biases the conditional treatment coefficient, because a treatment effect changes the prognostic composition of later risk sets. We characterize this bias and its sharp worst-case direction, and we introduce a cross-fitted projected score that removes it at first order and leaves only a second-order bias when a fixed error lies in its dictionary. A historical learning bound separates approximation, sampling, and optimization error. In simulations, calibration improves prediction when the historical risk is well learned, and the projected score restores interval coverage under drift outside the correction, also at a hazard ratio of 0.7. In a retrospective transfer between the POPLAR and OAK lung cancer trials, selected corrections reduce the integrated Brier score by 4.3\% relative to ridge-penalized clinical Cox with 100 training patients, without improving treatment precision.
ER -