Defining and Estimating Intervention Effects for Groups that will Develop an Auxiliary Outcome

By M. Joffe, Dylan S. Small, Chi-yuan Hsu. 2006

95 citations recorded by the source corpus.

It has recently become popular to define treatment effects for subsets of the target population characterized by variables not observable at the time a treatment decision is made. Characterizing and estimating such treatment effects is tricky; the most popular but naive approach inappropriately adjusts for variables affected by treatment and so is biased. We consider several appropriate ways to formalize the effects: principal stratification, stratification on a single potential auxiliary variable, stratification on an observed auxiliary variable and stratification on expected levels of auxiliary variables. We then outline identifying assumptions for each type of estimand. We evaluate the utility of these estimands and estimation procedures for decision making and understanding causal processes, contrasting them with the concepts of direct and indirect effects. We motivate our development with examples from nephrology and cancer screening, and use simulated data and real data on cancer screening to illustrate the estimation methods.

Subfields: Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/cc5207c738bc77f15a67bdb1057912a57d271bff

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

Odds ratios for mediation analysis for a dichotomous outcome. (2010, background): “As discussed elsewhere, controlled direct effects are often of greater interest in policy evaluation (2, 10), whereas natural direct and indirect effects are often of greater interest in evaluating the action of various mechanisms (10, 11).”
Compared to what? Variation in the impacts of early childhood education by alternative care type (2016, background): “Finally, it is useful to briefly review some alternative strategies that leverage auxiliary covariates to disentangle the finite mixture model (Joffe et al., 2007).”
Conceptual issues concerning mediation, interventions and composition (2009, background): “…the probability of falling within the principal stratum M(a) = M(0) will be zero in many realistic applications (Robins et al., 2007), and because the notions of mediation based on principal direct effects do not correspond in any clear way to mechanisms of scientific interest (Joffe et al., 2007).”
Interpretation and identification of causal mediation. (2014, unknown): “These developments, coupled with the capability of expressing and visualizing causal assumptions in graphical forms, have given rise to an explosion of mediation papers that have taken natural effects as the gold standards for analysis (e.g., Albert and Nelson, 2011; Coffman and Zhong, 2012; Hafeman”
Causality, mediation and time: a dynamic viewpoint (2012, background): “The idea is that the mediator can be physically manipulated (Joffe et al. (2007), page 91) to this given level for both treatment levels.”
Augmented designs to assess principal strata direct effects (2011, background): “Principal stratification is one of several possible ways to conceptualize the mediatory role of an intermediate variable in the treatment–outcome relationship (Joffe et al., 2007).”
Simple relations between principal stratification and direct and indirect effects (2008, background): “Several ways to conceptualize the mediatory role of an intermediate variable in the treatment–outcome relationship have been proposed in the causal inference literature (Joffe et al., 2007).”
Principal stratification analysis using principal scores (2016, methodology): “, 2011; Stuart and Jo, 2015) used principal scores to analyse data subject to one-sided non-compliance, and theoretical researchers (Joffe et al., 2007) suggested the use of principal scores to identify general causal effects within principal strata.”
The Causal Mediation Formula—A Guide to the Assessment of Pathways and Mechanisms (2012, background): “For discussions regarding policy implications of NDE versus CDE, see (Albert and Nelson 2011; Hafeman and Schwartz 2009; Joffe et al. 2007; Kaufman 2010; Pearl 2001, 2009, p. 132; Robins 2003; Robins and Richardson 2011).”
Identification, Inference and Sensitivity Analysis for Causal Mediation Effects (2010, background): “The statistical literature began to formally study causal mediation analysis about fifteen years ago (Robins and Greenland, 1992), and a number of articles have appeared in more recent years (e.g., Pearl, 2001; Robins, 2003; Rubin, 2004; Petersen et al., 2006; Geneletti, 2007; Joffe et al., 2007; Te”
Controlled direct and mediated effects: definition, identification and bounds (2010, background): “Controlled direct effects are often of greater interest in policy evaluation (Pearl, 2001; Robins, 2003); natural direct and indirect effects are often of greater interest in evaluating the action of various mechanisms (Robins, 2003; Joffe et al., 2007; Hafeman & Schwartz, 2009).”
The Mediation Formula: A Guide to the Assessment of Causal Pathways in Nonlinear Models (2011, background): “As a result, it is becoming widely recognized that the controlled, natural and indirect effects discussed in this paper are of greater interest, both for the purposes of making treatment decisions and for the purposes of explanation and identifying causal mechanisms (Joffe et al., 2007; Albert and N”
Related Causal Frameworks for Surrogate Outcomes (2009, methodology): “(6) Direct and indirect effects in randomized trials may be estimated under two types of assumptions: initial randomization (Robins and Greenland, 1994; Ten Have et al., 2004; Joffe, Small, and Hsu, 2007) and sequential ignorability (Robins and Greenland, 1992).”