ANALYSIS of more than 5,000 supposedly randomised trials, including in big names such as the Journal of the American Medical Association and the New England Journal of Medicine, has uncovered dozens in which the baseline data clearly is not random.
- The study, by English anaesthetist and renowned fraud hunter Dr John Carlisle, could lead to dozens of papers being retracted.
- The analysis points to a need for screening of trials not only before they are published but also going back decades.
- A major question now is what to do with intelligence about researchers who appear to be misinterpreting or fudging their data.
The new study, published in the journal Anaesthesia, applied a mathematical screening tool to 5,087 randomised, controlled trials published inn six anaesthetic and two general medical journals.
Dr Carlisle used a computer program designed to spot statistical inaccuracies to uncover the fraud of serial fabricator Dr Yoshitaka Fujii, a Tokyo-based anaesthetist, who holds the world record for most retractions by a single author after having 183 publications withdrawn.
Later, Dr Carlisle teamed up with University of Sydney anaesthetist Dr John Loadsman to point the figure at Dr Fujiis’ collaborator, Dr Yuhji Saitoh.
Now Dr Carlisle is applying the method to many papers, rather than focusing on individual researchers, investigating the distribution patterns
Read More >