scholarly journals Introduction to Time-dependent Confounders and Marginal Structural Models

2021 ◽  
Vol 3 (2) ◽  
pp. 37-45
Author(s):  
Asuka Tsuchiya
Keyword(s):  
Structural Models ◽  
Time Dependent ◽  
Marginal Structural Models

scholarly journals Simulating from marginal structural models with time-dependent confounding

2012 ◽  
Vol 31 (30) ◽  
pp. 4190-4206 ◽  
Author(s):  
W.G. Havercroft ◽  
V. Didelez
Keyword(s):  
Structural Models ◽  
Time Dependent ◽  
Marginal Structural Models

scholarly journals P0846ESA HYPORESPOSIVENESS AND CLINICAL OUTCOMES DURING TREATMENT WITH EPOETIN BETA PEGOL IN HEMODIALYSIS PATIENTS: A MULTICENTER PROSPECTIVE STUDY; PARAMOUNT-HD STUDY

2020 ◽  
Vol 35 (Supplement_3) ◽  
Author(s):  
Takayuki Hamano ◽  
Hideki Fujii ◽  
Ken Tsuchiya ◽  
Kuragano Takahiro ◽  
Nobuhiko Joki ◽  
...  
Keyword(s):  
Prospective Study ◽  
Cardiovascular Events ◽  
Structural Models ◽  
Mixed Effects ◽  
Mixed Effects Model ◽  
Time Dependent ◽  
Marginal Structural Models ◽  
On Line ◽  
Landmark Analyses ◽  
Multicenter Prospective Study

Abstract Background and Aims Hemodialysis (HD) patients hyporesponsive to erythropoiesis stimulating agents (ESAs) were reported to have poor prognosis. However, little is known regarding the association between the hyporesponsiveness to CERA and the types of outcome in HD patients. Moreover, the effect of on-line HDF on hyporesponsiveness to CERA has not been evaluated so far. Method In this multicenter prospective study, we enrolled 4034 maintenance HD patients receiving any kinds of ESA. Prior ESA was changed to CERA in all patients. We studied the association between erythropoietin resistance index (ERI) at 6-month after the change to CERA (baseline ERI) and such outcomes as cardiovascular events and/or mortality using Cox proportional hazards models (landmark analyses). ERI was defined as monthly CERA dose divided by hemoglobin and dry weight. Just before the enrollment of the patients, iron-based phosphate binders became available and on-line hemodiafiltration (HDF) began to be reimbursed in Japan, therefore, we examined whether oral iron-containing drugs and on-line HDF had some effects on the serial trend of ERI by mixed effects model with time-dependent ERI as a dependent variable. When ERI is found to be improved by these changes in practice patterns, we further studied the effect of time-dependent ERI on such outcomes as cardiovascular events, mortality, death due to cancer, and death due to infection by using marginal structural models to eradicate time-dependent confounding by iron parameters, C-reactive protein, iron-containing drugs, and HDF. Missing values were imputed by multiple imputations. Results Mean age was 65.9 years and 43.1% of patients had diabetes. The median dialysis vintage and observation period was 5.0 years and 22.1 months, respectively. The percentage of patients receiving oral iron-containing drugs increased from 11.1% at baseline to 25.0% at 24-month. As a result, mean TSAT level increased from 24.5% to 27.4% at 24-month. The percentage of patients on on-line HDF also increased from 13.5% to 22.6%. ERI gradually decreased as time went by with great improvement especially in patients with highest quintile of ERI (Q5). Mixed effects model with time-dependent ERI as a dependent variable showed that introduction of iron-containing drugs and on-line HDF had improved ERI significantly. The landmark analyses including 3001 patients failed to show significant associations between baseline ERI quintile and cardiovascular events or mortality. We only found that highest quintile of baseline ERI (Q5) was associated with significantly higher composite events of mortality and cardiovascular events as compared to the lowest quintile (Q1) (Hazard ratio [HR], 1.56; 95% CI; 1.04-2.32). However, marginal structural models showed that time-dependent ERI Q5 was significantly associated with higher cardiovascular event rate as compared to Q1 (HR, 2.11; 95% CI; 1.31-3.38). Trend toward higher rate of mortality with the increase in time-dependent ERI quintile was also observed (HR of Q5, 3.07; 95% CI; 1.95-4.83). Similar but stronger relationships were observed for death due to infection (HR of Q5, 6.70; 95% CI; 1.89-23.77) and death due to cancer (HR of Q5, 15.08; 95% CI; 4.08-55.74). Conclusion The prevailing use of iron-containing drugs and on-line HDF has improved hyporesponsiveness to CERA in Japan. Therefore, baseline ERI at 6-month did not predict subsequent cardiovascular events or death. However, high time-dependent ERI was a predictor of cardiovascular events, death due to infection, and death due to cancer as well as all-cause mortality. Strong association of time-dependent ERI was observed especially with death due to cancer.

scholarly journals Controlling for Time-dependent Confounding using Marginal Structural Models

2004 ◽  
Vol 4 (4) ◽  
pp. 402-420 ◽  
Author(s):  
Zoe Fewell ◽  
Miguel A. Hernán ◽  
Frederick Wolfe ◽  
Kate Tilling ◽  
Hyon Choi ◽  
...  
Keyword(s):  
Structural Models ◽  
Time Dependent ◽  
Marginal Structural Models

scholarly journals Optimal balancing of time-dependent confounders for marginal structural models

2021 ◽  
Vol 9 (1) ◽  
pp. 345-369
Author(s):  
Nathan Kallus ◽  
Michele Santacatterina
Keyword(s):  
Causal Effect ◽  
Model Misspecification ◽  
Structural Models ◽  
The United States ◽  
Time Dependent ◽  
Marginal Structural Models ◽  
Time To Death ◽  
Immunodeficiency Virus ◽  
Effect Of Treatment ◽  
Variance Estimates

Abstract Marginal structural models (MSMs) can be used to estimate the causal effect of a potentially time-varying treatment in the presence of time-dependent confounding via weighted regression. The standard approach of using inverse probability of treatment weighting (IPTW) can be sensitive to model misspecification and lead to high-variance estimates due to extreme weights. Various methods have been proposed to partially address this, including covariate balancing propensity score (CBPS) to mitigate treatment model misspecification, and truncation and stabilized-IPTW (sIPTW) to temper extreme weights. In this article, we present kernel optimal weighting (KOW), a convex-optimization-based approach that finds weights for fitting the MSMs that flexibly balance time-dependent confounders while simultaneously penalizing extreme weights, directly addressing the above limitations. We further extend KOW to control for informative censoring. We evaluate the performance of KOW in a simulation study, comparing it with IPTW, sIPTW, and CBPS. We demonstrate the use of KOW in studying the effect of treatment initiation on time-to-death among people living with human immunodeficiency virus and the effect of negative advertising on elections in the United States.

scholarly journals Targeted Maximum Likelihood Estimation for Dynamic and Static Longitudinal Marginal Structural Working Models

2014 ◽  
Vol 2 (2) ◽  
pp. 147-185 ◽  
Author(s):  
Maya Petersen ◽  
Joshua Schwab ◽  
Susan Gruber ◽  
Nello Blaser ◽  
Michael Schomaker ◽  
...  
Keyword(s):  
Maximum Likelihood ◽  
Structural Model ◽  
Structural Models ◽  
Time Dependent ◽  
Marginal Structural Models ◽  
Marginal Structural Model ◽  
Working Models ◽  
Targeted Maximum Likelihood ◽  
Baseline Covariates ◽  
Time Point

AbstractThis paper describes a targeted maximum likelihood estimator (TMLE) for the parameters of longitudinal static and dynamic marginal structural models. We consider a longitudinal data structure consisting of baseline covariates, time-dependent intervention nodes, intermediate time-dependent covariates, and a possibly time-dependent outcome. The intervention nodes at each time point can include a binary treatment as well as a right-censoring indicator. Given a class of dynamic or static interventions, a marginal structural model is used to model the mean of the intervention-specific counterfactual outcome as a function of the intervention, time point, and possibly a subset of baseline covariates. Because the true shape of this function is rarely known, the marginal structural model is used as a working model. The causal quantity of interest is defined as the projection of the true function onto this working model. Iterated conditional expectation double robust estimators for marginal structural model parameters were previously proposed by Robins (2000, 2002) and Bang and Robins (2005). Here we build on this work and present a pooled TMLE for the parameters of marginal structural working models. We compare this pooled estimator to a stratified TMLE (Schnitzer et al. 2014) that is based on estimating the intervention-specific mean separately for each intervention of interest. The performance of the pooled TMLE is compared to the performance of the stratified TMLE and the performance of inverse probability weighted (IPW) estimators using simulations. Concepts are illustrated using an example in which the aim is to estimate the causal effect of delayed switch following immunological failure of first line antiretroviral therapy among HIV-infected patients. Data from the International Epidemiological Databases to Evaluate AIDS, Southern Africa are analyzed to investigate this question using both TML and IPW estimators. Our results demonstrate practical advantages of the pooled TMLE over an IPW estimator for working marginal structural models for survival, as well as cases in which the pooled TMLE is superior to its stratified counterpart.

scholarly journals Marginal structural models with dose-delay joint-exposure for assessing variations to chemotherapy intensity

2018 ◽  
Vol 28 (9) ◽  
pp. 2787-2801
Author(s):  
Carlo Lancia ◽  
Cristian Spitoni ◽  
Jakob Anninga ◽  
Jeremy Whelan ◽  
Matthew R Sydes ◽  
...  
Keyword(s):  
Causal Effect ◽  
Randomised Trial ◽  
Bone Tumour ◽  
Structural Models ◽  
Cytotoxic Agents ◽  
Time Dependent ◽  
Exposure Model ◽  
Marginal Structural Models ◽  
Dynamical Process ◽  
Dose Delay

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. In this paper, 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 alterations to the therapy intensity. The model is motivated by a precise clinical question concerning the possibility of reducing dosages in a regimen. It is applied to data from a randomised trial of chemotherapy in osteosarcoma, an aggressive primary bone-tumour. Chemotherapy data are complex because their longitudinal nature encompasses many clinical details like composition and organisation of multi-drug regimens, or dynamical therapy adjustments. This manuscript focuses on the clinical dynamical process of adjusting the therapy according to the patient’s 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 the allocation of either a reduction or a delay of the next planned dose. Thus, a negative feedback is present between exposure to cytotoxic agents and toxicity levels, which acts as time-dependent confounders. The construction of the model is illustrated highlighting the high complexity and entanglement of chemotherapy data. Built to address dosage reductions, the model also shows that delays in therapy administration should be avoided. The last aspect makes sense from the cytological point of view, but it is seldom addressed in the literature.

Sign in / Sign up

Export Citation Format

Share Document