Structured Transfer Learning for Survival Risk Stratification in Data-Sparse Clinical Cohorts

2026-05-15

AI Summary

This study addresses the instability of existing survival risk prediction models in small-sample or event-sparse clinical cohorts, where direct transfer of models trained on large datasets is hindered by heterogeneity in risk factor effects across populations. To overcome this challenge, the authors propose the CORE-Cox framework, which first learns a shared coefficient structure across multiple outcomes in a source cohort using a low-rank multitask Cox model, then adapts to a target cohort through regularized residual correction for personalized calibration. By integrating shared structural information with cohort-specific adjustments, CORE-Cox achieves average C-indices of 0.766 and 0.658 on the UK Biobank and MIMIC-IV datasets, respectively, outperforming or closely matching the best alternatives in eight of nine evaluated outcomes. The method demonstrates substantially improved identification of high-risk individuals while maintaining interpretability, generalizability, and adaptability.

Abstract

Background: Survival prediction models are often less reliable in clinical groups with limited sample sizes or few outcome events. Target-only models may be unstable, whereas models from larger cohorts may transfer poorly when risk-factor effects differ across populations. We evaluated whether structured transfer learning can improve survival risk stratification in data-sparse cohorts while allowing cohort-specific adaptation. Methods: We developed the COhort-shared Rank-rEduced Cox model (CORE-Cox), a two-stage framework for multi-outcome survival prediction. CORE-Cox learns shared risk-factor patterns across related outcomes in a larger source cohort via a low-rank Cox coefficient structure, then adapts these patterns to a smaller target cohort through regularized residual correction. We evaluated CORE-Cox in UK Biobank (White source, n=150,093; Asian target, n=2,534) and MIMIC-IV (White ICU source, n=15,997; Asian ICU target, n=672), comparing against target-only Cox, penalized Cox, low-rank multi-task, naive pooling, direct transfer, and single-outcome residual transfer under repeated nested cross-validation. Results: CORE-Cox achieved best or near-best discrimination across most outcomes. Mean C-index improved from 0.733 to 0.766 in UK Biobank and from 0.628 to 0.658 in MIMIC-IV, with gains in eight of nine outcomes. CORE-Cox also improved top-15% risk enrichment, with hazard-ratio estimates typically intermediate between source-only and target-only models. Discussion: CORE-Cox offers an interpretable transfer-learning framework for survival risk stratification in data-sparse cohorts, combining shared cross-outcome structure with cohort-specific adaptation. Further validation is needed before use in calibrated absolute-risk prediction or clinical decision-making.

Problem

• survival prediction
• data-sparse cohorts
• transfer learning
• risk stratification
• clinical cohorts

Innovation

• structured transfer learning
• low-rank Cox model
• survival risk stratification
• data-sparse cohorts
• multi-outcome adaptation