Scholarlink

<p>This study proposes a novel semiparametric Bayesian joint estimation approach for mediation analysis under dependent right-censored data, combining latent variable structural equation modeling (SEM) with a frailty model. The method accounts for dependence between censoring mechanisms and survival times, reducing bias relative to approaches assuming independent censoring, and enhancing estimation accuracy and robustness for mediation analysis. Compared with existing methods, this approach may be particularly useful for populations with high incidence rates and substantial heterogeneity. Its performance has been evaluated through simulations and applications to breast cancer and Alzheimer’s disease datasets, demonstrating its practical utility and broad applicability.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Bayesian mediation analysis with latent mediators and survival outcome under dependent right-censored data

  • 期刊论文

  • 发表日期：

    2026年7月29日

查看全文
  • Shunxin Qi,
  • Wenshan Wang,
  • Jingjing Jiang,
  • Chunjie Wang

摘要

This study proposes a novel semiparametric Bayesian joint estimation approach for mediation analysis under dependent right-censored data, combining latent variable structural equation modeling (SEM) with a frailty model. The method accounts for dependence between censoring mechanisms and survival times, reducing bias relative to approaches assuming independent censoring, and enhancing estimation accuracy and robustness for mediation analysis. Compared with existing methods, this approach may be particularly useful for populations with high incidence rates and substantial heterogeneity. Its performance has been evaluated through simulations and applications to breast cancer and Alzheimer’s disease datasets, demonstrating its practical utility and broad applicability.