The first α1 parameter, which we denote as α1,1 is the (adjusted) treatment effect, but here the focus of interest is on the covariates that are important in the missing data model, ie β1 and β2 are paramount.
Percentages were calculated based on valid data (ie, missing data were excluded).
Baseline drinking status (ever vs never tried alcohol) did not predict attrition, but to account for attrition bias related to other variables, estimation was carried out after multiple imputation using the standard missing at random assumption (ie, missing data are assumed missing at random conditional on observed predictors included in the model).
Furthermore, we should be able to explore more general issues such as how best to deal with missing data in large clinical samples (ie, complete case, dummy variables for missing data, imputation).
They should consider the amount of missing data from their meta-analysis (ie, the percentage of the sample sizes of the studies that were included compared with those that would have been eligible to be included in the meta-analysis but no outcome data were reported), and this information should be included along with the pooled effect estimate.
In case of missing data a non-responder imputation will be applied (ie, baseline observation carried forward).
The results for complete cost and quality-of-life data (ie, those with no missing data) were provided to identify the impact of missing data on the analysis.
25 Of self-reported non-smokers, 48 did not provide a saliva sample and in two cases the cotinine analysis was not possible (ie, insufficient sample); these cases were considered missing data.
Similar to table 1, the values in each row of table 2 do not sum to the total number of studies in the first column (ie, 'Conducted ITT') due to 42 studies utilising multiple missing data strategies.
Besides the explicit coding of unobserved walking ability, missing data for all other items of the phenotype were treated as missing (ie, not given an explicit code).
