Better radiology in the BMJ

BMJ 2008;336:628 (22 March)
Editorials
Better radiology in the BMJ
Despite improvements already made, we still have more to do
The first 150 words of the full text of this article appear below.
The BMJ has a history of being self critical. In 1976 we published a review of 62 BMJ research papers, which showed that 32 of them had statistical faults and five had made claims in their abstracts that were not supported by their results.1 These mistakes occurred because at that time statisticians were not commonly involved in research papers, either as authors or peer reviewers.2 To improve the reliability of published research, the BMJ was one of the first medical journals to introduce statistical review of all published research papers, and the role of statisticians in peer review continues to be of interest today.3 We now plan to do the same for radiology, using external advisors and radiologists to guide us and improve our content.
Twenty years ago radiology was mainly plain radiographs viewed on light boxes. Now digital technology means that multislice computed tomography, magnetic resonance imaging, Doppler ultrasound, . . . [Full text of this article]