A martingale approach to continuous-time marginal structural models
Author:
- Kjetil RøyslandKRKjetil Røysland
Open Access
- 1 August 2011
- journal article
- Published by Bernoulli Society for Mathematical Statistics and Probability in Bernoulli
- Vol. 17 (3) , 895-915
Abstract
Marginal structural models were introduced in order to provide estimates of causal effects from interventions based on observational studies in epidemiological research. The key point is that this can be understood in terms of Girsanov’s change of measure. This offers a mathematical interpretation of marginal structural models that has not been available before. We consider both a model of an observational study and a model of a hypothetical randomized trial. These models correspond to different martingale measures – the observational measure and the randomized trial measure – on some underlying space. We describe situations where the randomized trial measure is absolutely continuous with respect to the observational measure. The resulting continuous-time likelihood ratio process with respect to these two probability measures corresponds to the weights in discrete-time marginal structural models. In order to do inference for the hypothetical randomized trial, we can simulate samples using observational data weighted by this likelihood ratio.Keywords
All Related Versions
This publication has 17 references indexed in Scilit:
- Statistical modeling of causal effects in continuous timeThe Annals of Statistics, 2008
- Graphical Models for Marked Point Processes Based on Local IndependenceJournal of the Royal Statistical Society Series B: Statistical Methodology, 2008
- Long-term effectiveness of potent antiretroviral therapy in preventing AIDS and death: a prospective cohort studyThe Lancet, 2005
- A Structural Approach to Selection BiasEpidemiology, 2004
- Marginal Structural Models to Estimate the Causal Effect of Zidovudine on the Survival of HIV-Positive MenEpidemiology, 2000
- Noncausality in Continuous TimeEconometrica, 1996
- Estimation of the time-dependent accelerated failure time model in the presence of confounding factorsBiometrika, 1992
- Sur l'int grabilit uniforme des martingales exponentiellesProbability Theory and Related Fields, 1978
- Multivariate point processes: predictable projection, Radon-Nikodym derivatives, representation of martingalesProbability Theory and Related Fields, 1975
- Composable Markov processesJournal of Applied Probability, 1970