CMStatistics 2020: Start Registration
View Submission - CMStatistics
Title: Marginal structural models with joint exposure to assess variations to chemotherapy intensity Authors:  Marta Fiocco - Leiden University (Netherlands) [presenting]
Abstract: Marginal structural models are causal models designed to adjust for time-dependent confounders in observational studies with dynamically adjusted treatments. They are robust tools to assess causality in complex longitudinal data. A marginal structural model is proposed with an innovative dose-delay joint-exposure model for Inverse Probability of Treatment Weighted estimation of the causal effect of therapy modification. The model is motivated by a clinical question concerning the possibility of reducing dosages in a regimen. It is applied to data from a randomized trial of chemotherapy in osteosarcoma, an aggressive primary bone-tumor. This talk focuses on the clinical dynamical process of adjusting the therapy according to patients toxicity history, and the causal effect on the outcome of interest of such therapy modifications. Depending on patients toxicity levels, variations to therapy intensity may be achieved by physicians through a reduction or a delay of the next planned dose. Therefore, negative feedback is present between exposure to cytotoxic agents and toxicity levels, which acts as time-dependent confounders. The construction of the model is presented, and the high complexity and entanglement of chemotherapy data are discussed. Built to address dosage reductions, the model suggests that delays in therapy administration should be avoided.