John Carlisle, a consultant anaesthetist at Torbay Hospital, used statistical jiu-jitsu to detect anomalies in the demographic data of a number of high profile articles.
John Carlisle
This is a fast moving story, but it has the potential to bring down multiple careers. The AASK trial is on the list of suspect articles.
See blog coverage here in For Better Science
Coverage in the Guardian here
Coverage in Retraction Watch is on the case (of course) here
An editorial that discusses Carlisle's methodology of detecting fraud is here
And the bombshell itself is here
UPDATE [June 8th 2017]
Most statisticians seem unimpressed.
Nick Brown @Steamtraen with a nice explanatory walkthrough
Andrew Gelman (@statmodeling) philosophical take
Tom's tweet that started a loong thread, with input from Eugene Lin, Stephen Senn and Perry Wilson:
UPDATE 2 [June 12th 2017]
Perry Wilson @Methodsmanmd distils the bewildering stats into something we can understand. Full text here (pay attention to the footnotes!), video embedded below: