Evaluating the impact of imputations for missing participant outcome data in a network meta-analysis
Loukia M Spineli, Julian PT Higgins, Andrea Cipriani, Stefan Leucht, Georgia Salanti
Missing outcome data frequently occur in randomized controlled trials (RCTs), either because of participants leaving the study early or due to post-randomization exclusions. Missing data are a potential threat to the validity of inferences from RCTs. They lead to increased uncertainty over the effects of an intervention and, if ignored, may lead to biased estimates. It is therefore important that missing data are appropriately considered when RCTs are included in meta-analyses. Informative missing outcome data can lead to biased meta-analysis estimates as well as to high heterogeneity. Read more... |
|
CRD databases updates: 09/05/2013 to 15/05/2013
171 systematic reviews have been added to DARE
42 economic evaluations have been added to NHSEED
8 summaries of health technology assessments have been added to the HTA database
|