The performance of marginal structural models for estimating risk differences and relative risks using weighted univariate generalized linear models
Austin PC. Stat Methods Med Res. 2024; Apr 24 [Epub ahead of print].
Applied researchers frequently use automated model selection methods, such as backwards variable elimination, to develop parsimonious regression models. Statisticians have criticized the use of these methods for several reasons, amongst them are the facts that the estimated regression coefficients are biased and that the derived confidence intervals do not have the advertised coverage rates.
The authors developed a method to improve estimation of regression coefficients and confidence intervals which employs backwards variable elimination in multiple bootstrap samples. In a given bootstrap sample, predictor variables that are not selected for inclusion in the final regression model have their regression coefficient set to zero. Regression coefficients are averaged across the bootstrap samples, and non-parametric percentile bootstrap confidence intervals are then constructed for each regression coefficient.
The investigators conducted a series of Monte Carlo simulations to examine the performance of this method for estimating regression coefficients and constructing confidence intervals for variables selected using backwards variable elimination. They demonstrated that this method results in confidence intervals with superior coverage compared with those developed from conventional backwards variable elimination. The investigators illustrate the utility of their method by applying it to a large sample of subjects hospitalized with a heart attack.
Austin PC. Stat Med. 2008; 27(17):3286-300.
The ICES website uses cookies. If that’s okay with you, keep on browsing, or learn more about our Privacy Policy.