[Skip to Navigation]
[Skip to Navigation Landing]
Views 858
Citations 0
Comment & Response
April 23/30, 2019

Post Hoc Bayesian Analyses

Author Affiliations
  • 1University of Utah School of Medicine, Salt Lake City
JAMA. 2019;321(16):1631-1632. doi:10.1001/jama.2019.1198

To the Editor Dr Goligher and colleagues1 used a Bayesian approach to reanalyze the negative ECMO [extracorporeal membrane oxygenation] to Rescue Lung Injury in Severe ARDS [acute respiratory distress syndrome] (EOLIA) trial. I believe this reanalysis is an example of confirmation bias2 and illustrates the common criticisms of Bayesian methodology.