Premium
Commentary: Indefensible Methods of Handling Missing Data in Clinical Trials
Author(s) -
Arndt Stephan
Publication year - 2013
Publication title -
alcoholism: clinical and experimental research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.267
H-Index - 153
eISSN - 1530-0277
pISSN - 0145-6008
DOI - 10.1111/acer.12294
Subject(s) - missing data , imputation (statistics) , cover (algebra) , computer science , data science , econometrics , statistics , mathematics , engineering , machine learning , mechanical engineering
Background Literature continues to appear using inappropriate statistical methods when dealing with missing data in treatment trials. I regularly see last observation carried forward, assumed worst‐case scenario, or some other static imputation method still being submitted and published in journals. Methods We briefly cover a few reasons why the use of more modern and defensible methods may not have completely saturated the literature. Results While some delay in complete permeation of appropriate methods is understandable, we are currently past the point for reasonable delay. Conclusions Editors and reviewers should demand appropriate statistical methods in published literature.