Constructing inverse probability weights for marginal structural models.

By Stephen R. Cole, Miguel A. Hern'an. American Journal of Epidemiology, 2008

2,645 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/ddd660f2a83ec98ebb7f0b96499822b8ff35d24b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Statistics and Probability

Citations in the corpus

The Rotterdam Study: 2010 objectives and design update (2007, background): “First, based on earlier publications on marginal structural modeling [374], we are currently analyzing the association between statins and cardiovascular mortality with such models in which we try to use more precise exposure definitions than use/non-use.”
Doubly robust estimation of causal effects. (2011, background): “estimators, it is known that weights for individuals with unusual combinations of characteristics and exposures can lead to unstable estimates with relatively large standard errors (19).”
Bayesian Regression Tree Models for Causal Inference: Regularization, Confounding, and Heterogeneous Effects (with Discussion) (2020, neutral): “This problem is aggravated by the tension between satisfying ignorability and positivity (Cole and Hernán, 2008).”
A Bayesian view of doubly robust causal inference (2016, methodology): “The probabilities pE(zi) are given by the chosen regime E that is the object of inference; in practice the estimation is most efficient when we choose the target regime to be as close as possible to the observed regime O; this can be achieved by fixing pE(zi) to the marginal treatment assignment pro”
The consistency statement in causal inference: a definition or an assumption? (2009, background): “The positivity assumption states that there is a nonzero (ie, positive) probability of receiving every level of exposure for every combination of values of exposure and confounders that occur among individuals in the population.(2,3) It remains unclear why the consistency and positivity assumptions ”
Beverage purchases from stores in Mexico under the excise tax on sugar sweetened beverages: observational study (2016, background): “effects) probit models to address the potential selection bias associated with the probability of purchasing (5).”
Observational data for comparative effectiveness research: an emulation of randomised trials to estimate the effect of statins on primary prevention of coronary heart disease (2011, methodology): “the positivity condition may not hold.(24) In the THIN data positivity approximately holds as shown by the extent of the overlap between the distribution of the estimated propensity scores for initiators and non-initiators (Figure 4).”
A Primer on Inverse Probability of Treatment Weighting and Marginal Structural Models (2016, background): “The following key assumptions must be met (Cole & Hernán, 2008): 1.”
Estimating the effects of multiple time-varying exposures using joint marginal structural models: alcohol consumption, injection drug use, and HIV acquisition (2012, background): “The lack of extreme weights in the ALIVE cohort also provides some evidence against model misspecification.(28,29) In addition, positivity and correct model specification were facilitated by careful consideration of all categorizations, particularly for alcohol consumption and injection drug use.”
Stable Weights that Balance Covariates for Estimation With Incomplete Outcome Data (2015, methodology): “…studies of treatment effects they can be used as an alternative to inverse probability weighting (where weights are multiplied successively tending to achieve very large values; Hogan and Lancaster 2004), and as a complement to marginal structural models (Joffe et al. 2004; Cole and Hernán 2008).”
Generalizing evidence from randomized clinical trials to target populations: The ACTG 320 trial. (2010, methodology): “Last, a Bernoulli random variable was generated for selection into the trial S, conditional on the realized value of the demographic characteristic Z, as 1=f1þ exp1⁄2 b0 b1Zi g, with b1 set at log(4) to reflect the size of selection effects observed in the ACTG 320 trial for age groups and b0 chosen”
An introduction to inverse probability of treatment weighting in observational research (2021, background): “An important methodological consideration of the calculated weights is that of extreme weights [26].”
Doubly Robust Estimation of Optimal Dynamic Treatment Regimes (2013, methodology): “It enables us to estimate causal effects from observational data (see Hernán and Robins [11], for a discussion of the exchangeability assumption).”