A stacked approach for chained equations multiple imputation incorporating the substantive model

TL;DR

This paper introduces a novel stacked imputation method that integrates the analysis model directly into multiple imputation for missing data, ensuring model compatibility and providing a new way to analyze stacked imputations.

Contribution

The paper proposes a new approach that stacks multiple imputations and incorporates the analysis model via weights, along with a novel standard error estimator, enhancing flexibility and applicability.

Findings

Produced unbiased estimates when the analysis model was correct

Developed an R package, StackImpute, for implementation

Applicable to a wide range of analysis models and missing data scenarios

Abstract

Multiple imputation by chained equations (MICE) has emerged as a popular approach for handling missing data. A central challenge for applying MICE is determining how to incorporate outcome information into covariate imputation models, particularly for complicated outcomes. Often, we have a particular analysis model in mind, and we would like to ensure congeniality between the imputation and analysis models. We propose a novel strategy for directly incorporating the analysis model into the handling of missing data. In our proposed approach, multiple imputations of missing covariates are obtained without using outcome information. We then utilize the strategy of imputation stacking, where multiple imputations are stacked on top of each other to create a large dataset. The analysis model is then incorporated through weights. Instead of applying multiple imputation combining rules, we obtain parameter estimates by fitting a weighted version of the analysis model on the stacked dataset. We propose a novel estimator for obtaining standard errors for this stacked and weighted analysis. Our estimator is based on the observed data information principle in Louis (1982) and can be applied for analyzing stacked multiple imputations more generally. Our approach for analyzing stacked multiple imputations is the first well-motivated method that can be easily applied for a wide variety of standard analysis models and missing data settings. In simulations, the proposed strategy produced unbiased parameter estimates when the analysis model was correctly specified. We developed an R package, StackImpute, allowing this imputation approach to be easily implemented for many standard analysis models.