Methods of covariate selection: directed acyclic graphs and the change-in-estimate procedure.

@article{Weng2009MethodsOC,
  title={Methods of covariate selection: directed acyclic graphs and the change-in-estimate procedure.},
  author={Hsin-Yi Weng and Ya-Hui Hsueh and Locksley L. Mcv. Messam and Irva Hertz-Picciotto},
  journal={American journal of epidemiology},
  year={2009},
  volume={169 10},
  pages={
          1182-90
        }
}
Four covariate selection approaches were compared: a directed acyclic graph (DAG) full model and 3 DAG and change-in-estimate combined procedures. Twenty-five scenarios with case-control samples were generated from 10 simulated populations in order to address the performance of these covariate selection procedures in the presence of confounders of various strengths and under DAG misspecification with omission of confounders or inclusion of nonconfounders. Performance was evaluated by standard… 
View on PubMed
academic.oup.com
173 Citations
Highly Influential Citations
3
Background Citations
13
Methods Citations
47
Results Citations
2

Tables from this paper

173 Citations

Citation Type

Citation Type

Combining directed acyclic graphs and the change-in-estimate procedure as a novel approach to adjustment-variable selection in epidemiology
TLDR
This approach to adjustment-variable selection combines background-knowledge and statistics-based approaches using methods already common in epidemiology and communicates assumptions and uncertainties in a standardized graphical format.
Re: "Methods of covariate selection: directed acyclic graphs and the change-in-estimate procedure".
  • L. Nie
  • Environmental Science, Mathematics
    American journal of epidemiology
  • 2009
TLDR
The sampling distribution-based standard error reported by Weng et al. (1) for the combined procedures is not the same as the standard error considered by Robinson and Jewell (2), therefore, the results on standard errors do not contradict the theoretical results derived by Robinsonand Jewell.
Outcome modelling strategies in epidemiology: traditional methods and basic alternatives
TLDR
This work reviews several traditional modelling strategies, including stepwise regression and the ‘change-in-estimate’ (CIE) approach to deciding which potential confounders to include in an outcome-regression model for estimating effects of a targeted exposure, and provides some basic alternatives and refinements that do not require special macros or programming.
Covariate selection strategies for causal inference: Classification and comparison.
TLDR
There are considerable differences between the approaches to covariate selection but none of them is uniformly best, with performance depending on the chosen adjustment method as well as the true confounding structure.
Statistical foundations for model-based adjustments.
TLDR
This work reviews several established strategies for selecting model covariates, describes their shortcomings, and point to refinements, assuming that the main goal is to derive the most accurate effect estimates obtainable from the data and available resources.
The change in estimate method for selecting confounders: A simulation study
TLDR
It is questionable whether change in estimate methods are beneficial in general, considering their low ability to improve the precision of estimates without introducing bias and inability to yield valid confidence intervals or to identify true confounders.
Selecting Shrinkage Parameters for Effect Estimation: The Multi-Ethnic Study of Atherosclerosis
TLDR
This work connects the theory of causal inference, which describes how variable adjustment should be performed with large samples, with shrinkage estimators such as ridge regression and the least absolute shrinkage and selection operator (LASSO), which can perform better in sample sizes seen in epidemiologic practice.
Should we adjust for a confounder if empirical and theoretical criteria yield contradictory results? A simulation study
  • P. Lee
  • Psychology
    Scientific reports
  • 2014
TLDR
In the presence of a true confounding effect, unbiased estimates were obtained only by using the CIE criterion with a simulated cutoff, and all criteria performed well regardless of adjustment, but the adjusted estimates were biased.
Variance estimation for stratified propensity score estimators.
TLDR
This work deriving the asymptotic marginal variance of the stratified estimate of treatment effect, correctly taking into account the estimation of the propensity score, shows that routinely used variance estimators are likely to produce confidence intervals that are too conservative when the propensity Score model includes variables that predict (cause) the outcome, but only weakly predict the treatment.
A descriptive review of variable selection methods in four epidemiologic journals: there is still room for improvement
TLDR
There is still room for improvement in variable selection methods used by epidemiologists and in their reporting, despite being less frequent than in the previous review, stepwise and univariable analyses, which are susceptible to introduce bias and produce inadequate inferences, were still prevalent.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 37 REFERENCES
Simulation study of confounder-selection strategies.
TLDR
The authors compared the performance of several such strategies for fitting multiplicative Poisson regression models to cohort data, finding that the change-in-estimate and equivalence-test-of-the-difference strategies performed best when the cut-point for deciding whether crude and adjusted estimates differed by an important amount was set to a low value.
The effect of omitted covariates on confidence interval and study power in binary outcome analysis: a simulation study.
Biased estimates of treatment effect in randomized experiments with nonlinear regressions and omitted covariates
SUMMARY Certain important nonlinear regression models lead to biased estimates of treatment effect, even in randomized experiments, if needed covariates are omitted. The asymptotic bias is determined…
The effects of covariate adjustment in generalized linear models
Results from classical linear regression regarding the effects of covariate adjustment, with respect to the issues of confounding, the precision with which an exposure effect can be estimated, and…
The role of model selection in causal inference from nonexperimental data.
TLDR
It is argued that all modelling strategies contain implicit prior beliefs about nature, and one can justify a modelling strategy only if the strategy is justified.
Modeling and variable selection in epidemiologic analysis.
  • S. Greenland
  • Computer Science
    American journal of public health
  • 1989
TLDR
An overview of problems in multivariate modeling of epidemiologic data is provided, and some proposed solutions are examined, including model and variable forms should be selected based on regression diagnostic procedures, in addition to goodness-of-fit tests.
A Backwards-Manual Selection Macro for Binary Logistic Regression in the SAS ® v . 8 . 02 PROC LOGISTIC Procedure
Multivariate logistic regression is often used within the field of epidemiology to describe the relationship between disease occurrence and an exposure suspected to be associated with the disease.…
Confounder selection in environmental epidemiology: assessment of health effects of prenatal mercury exposure.
The impact of confounder selection criteria on effect estimation.
TLDR
The results of a Monte Carlo simulation of several confounder selection criteria, including change-in-estimate and collapsibility test criteria, are presented, compared with respect to their impact on inferences regarding the study factor's effect.
Some surprising results about covariate adjustment in logistic regression models
Summary Results from classic linear regression regarding the effect of adjusting for covariates upon the precision of an estimator of exposure effect are often assumed to apply more generally to…
...
1
2
3
4
...