Auxiliary variables were not included in the imputation model due to convergence problems of the EM algorithm, which was used to estimate the initial values of imputation model parameters, of procedure MI.
To assess the added value of multiple imputation (MI) of missing repeated outcomes measures in longitudinal data sets analyzed with linear mixed-effects (LME) models.
To our best knowledge, this is the first study to evaluate the value of an imputation method alternative to the WOMAC method.
When the logP value of the imputation test is lower than that of the haplotype analysis, the SNPs at the QTL peak are unlikely to be causal; in contrast when the merge logP is higher, it suggests that the tested variant is consistent with being a quantitative trait nucleotide (QTN), with the caveat that there may be multiple variants with the same SDP that are equally likely to be causal [ 34].
For missing data at the study level (i.e. information not documented in a study), we will also consider imputation to allow multivariable analyses on the most complete dataset, although the added value of such major imputation efforts may be limited and will be explored in the perspective of IPD meta-analysis [ 50].
The purpose of this step is to set a baseline to which we compared the added value of the 2 imputation methods.
Finally, a complete-case analysis will be performed to assess the influence of missing data and the value of adopting multiple imputation methods.
This paper evaluated the added value of the EM imputation method (which is based on probabilistic models) as compared to the WOMAC method.
35 Ten imputed datasets were created by replacing missing values with simulated values from a set of imputation models built from all potential prognostic variables (smoking prevalence, household crowding and ILI rate) and the outcome variable (incidence of IMD).
However, owing to the moderate amounts of missing data in the cohort, primarily for occupation and occupational activity (both missing 6.3% of values), multiple imputation of missing observations was performed to evaluate whether complete-case analyses introduced bias.
