Survival Analysis Under the Aalen's Additive Hazards Model With Covariate Measurement Error: Application to Causal...

Area of Science:

• Biostatistics
• Survival Analysis
• Causal Inference

Background:

• Covariate measurement error is a known issue in Cox proportional hazards models.
• The impact of measurement error on Aalen's additive hazards model is understudied.
• Aalen's additive hazards model is increasingly used in survival mediation analysis, where mediators often have measurement error.

Purpose of the Study:

• To address covariate measurement error within Aalen's additive hazards model.
• To develop and extend a measurement error correction strategy for causal mediation analysis in survival data with error-prone longitudinal mediators.

Main Methods:

• Proposed a novel measurement error correction strategy for Aalen's additive hazards model.
• Extended the correction method to handle error-prone longitudinal mediators in survival causal mediation analysis.
• Employed numerical studies to evaluate the performance of the proposed methods.

Main Results:

• Developed a method to correct for covariate measurement error in Aalen's additive hazards model.
• Achieved corrected estimation of direct and indirect effects in survival mediation analysis.
• Demonstrated the effectiveness of the proposed strategy through simulation studies.

Conclusions:

• The proposed method effectively corrects for measurement error in Aalen's additive hazards model.
• The extension provides a valuable tool for causal mediation analysis with longitudinal, error-prone mediators in survival settings.
• The findings advance statistical methodologies for handling measurement error in complex survival data analyses.