Handbook of Generalized Pairwise Comparisons
Marc Buyse, Johan Verbeeck, Everardo D. Saad, Mickaël De Backer, Vaiva Deltuvaite‐Thomas, Geert Molenberghs
Abstract
Missing data are ubiquitous. When running a clinical trial, one can encounter a vast array of issues that can lead to a situation where a particular outcome is not recorded. For instance, the instruments may fail, logistics might run late to deliver a particular test, patients might face difficulties to reach their appointments or drop out completely, etc. These unplanned events are impossible to account for ahead of time, and they will be reflected in the data that are collected, causing potential bias and loss of power in subsequent analyses.
Topics & Concepts
Pairwise comparisonComputer scienceMathematicsArtificial intelligenceStatistical Methods in Clinical TrialsMeta-analysis and systematic reviewsAdvanced Causal Inference Techniques