A0832
Title: Estimation of average treatment effect for survival outcomes with continuous treatment in observational studies
Authors: Qi Zheng - University of Louisville (United States) [presenting]
Abstract: In healthcare research, where extensive observational data such as claims data and electronic records are readily available, researchers often seek to investigate both the treatment effect and the pathway of that effect. While recent literature on causal effects in survival analyses primarily focuses on binary or multiple treatment settings, studies involving continuous treatment settings are rarely explored. The estimation of the average treatment effect (ATE) of continuous treatment is explored on time-to-event outcomes by addressing multiple confounding factors and considering censoring observations. The ATE is proposed to estimate using the accelerated failure time marginal structural model (AFT-MSM), incorporating the inverse probability of treatment weighting (IPTW) method along with censoring weights. The IPTW method is designed to mitigate the influence of confounding variables on treatment assignment while censoring weights and addressing potential biases arising from censored observations. Extensive simulation studies have demonstrated the effectiveness of the proposed method. The proposed methodology is applied to investigate the impact of blood lead levels on the time to death among older individuals in the United States, utilizing data from the NHANES III survey dataset.
