Efficient Inference for Time-to-Event Outcomes by Integrating Right-Censored and Current Status Data

2025-08-14

AI Summary

Existing methods struggle to efficiently integrate two heterogeneous types of survival data: right-censored data—typically modeled separately—and current-status data—which suffers from slow convergence and non-normal limiting distributions, precluding direct use in standard meta-analysis. This paper proposes the first semiparametric data fusion framework, constructing a joint model under exchangeability assumptions, deriving the canonical gradient for survival probability, and designing a first-order estimator with double robustness. We prove that this estimator achieves the semiparametric efficiency bound. The method substantially improves estimation efficiency even when current-status data are information-poor, and naturally extends to settings with covariate shift. Simulation studies demonstrate superior estimation accuracy and inferential performance compared to existing approaches. By unifying analysis of mixed censoring schemes under a theoretically rigorous and computationally feasible paradigm, our framework advances the methodology for integrative survival analysis.

Technology Category

• Search and Optimization: Mixed Discrete/Continuous Search
• Reasoning under Uncertainty: Relational Probabilistic Models
• Machine Learning: Probabilistic Circuits and Graphical Models

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

We propose a semiparametric data fusion framework for efficient inference on survival probabilities by integrating right-censored and current status data. Existing data fusion methods focus largely on fusing right-censored data only, while standard meta-analysis approaches are inadequate for combining right-censored and current status data, as estimators based on current status data alone typically converge at slower rates and have non-normal limiting distributions. In this work, we consider a semiparametric model under exchangeable event time distribution across data sources. We derive the canonical gradient of the survival probability at a given time, and develop a one-step estimator along with the corresponding inference procedure. Our proposed estimator is doubly robust and attains the semiparametric efficiency bound under mild conditions. Importantly, we show that incorporating current status data can lead to meaningful efficiency gains despite the slower convergence rate of current status-only estimators. We demonstrate the performance of our proposed method in simulations and discuss extensions to settings with covariate shift. We believe that this work has the potential to open new directions in data fusion methodology, particularly for settings involving mixed censoring types.

Problem

• Efficient survival probability inference from mixed censored data
• Overcoming slower convergence of current status-only estimators
• Developing doubly robust semiparametric fusion framework

Innovation

• Semiparametric data fusion framework
• One-step doubly robust estimator
• Integrates right-censored and current status data