Multiple imputation for handling missing outcome data in randomized trials involving a mixture of independent and paired data

Abstract

Randomized trials involving independent and paired observations occur in many areas of health research, for example in paediatrics, where studies can include infants from both single and twin births. Multiple imputation (MI) is often used to address missing outcome data in randomized trials, yet its performance in trials with independent and paired observations, where design effects can be less than or greater than one, remains to be explored. Using simulated data and through application to a trial dataset, we investigated the performance of different methods of MI for a continuous or binary outcome when followed by analysis using generalized estimating equations to account for clustering due to the pairs. We found that imputing data separately for independent and paired data, with paired data imputed in wide format, was the best performing MI method, producing unbiased point and standard error estimates for the treatment effect throughout. Ignoring clustering in the imputation model performed well in settings where the design effect due to the inclusion of paired data was close to one, but otherwise led to moderately biased variance estimates. Including a random cluster effect in the imputation model led to slightly biased point estimates for binary outcome data and variance estimates that were too small in some settings. Based on these results, we recommend researchers impute independent and paired data separately where feasible to do so. The exception is if the design effect due to the inclusion of paired data is close to one, where ignoring clustering may be appropriate.

Mediation Analysis to Untangle Opposing Associations of High-Dose Docosahexaenoic Acid With IQ and Bronchopulmonary Dysplasia in Children Born Preterm

“…A rich set of auxiliary variables was incorporated into models for IQ and BPD to minimize the potential for bias under a missing-at-random assumption. Given that children from a multiple birth were randomized individually, all observations were considered independent in the imputation model . A total of 100 complete data sets were generated, with results combined across data sets using Rubin rules.…”

“…Given that children from a multiple birth were randomized individually, all observations were considered independent in the imputation model. 28 A total of 100 complete data sets were generated, with results combined across data sets using Rubin rules. Additional details on imputation methods (including auxiliary variables) are provided in the eAppendix in Supplement 1 .…”

Partially clustered designs for clinical trials: Unifying existing designs using consistent terminology

“…30,36,73 The statistical issues that arise in partially clustered trials involving pre-existing clusters have been discussed in specialty fields including ophthalmology, 21 orthopaedics 71,72 and surgery, 74 but without the use of a specific term for this design. Recent methodological work has been conducted on sample size 23 and multiple imputation 75 for trials involving a combination of independent and paired data, but without reference to partial clustering. Sample size methods exist for some types of partially clustered trials.…”