Major medical journals do not follow their own rules to report clinical trial results | Science

This is a well-known problem in clinical trials: researchers start by saying that they will look for a particular result – heart attacks, for example – and then report something else when they publish their results. This practice may give the impression that a drug or treatment is safer or more effective than it actually is. Today, a systematic effort to find out if leading journals are honoring their own commitment to ensuring that results are reported correctly has shown that many people fail – and journals and writers are overflowing with 39; apology.

When journals and researchers were asked to correct studies, the answers "were fascinating and alarming. Writers and researchers often misunderstand what a proper test report looks like, "says Ben Goldacre, project manager, author and physician at the University of Oxford in the United Kingdom and advocate for transparency in drug research.

4 years ago, his team's evidence-based medicine results monitoring project (The COMPARE project) examined all the published essays for six weeks in five journals: Annals of Internal Medicine, The bmj, JAMA, The lancet, and The New England Medical Journal (NEJM). Study subjects ranged from health effects of alcohol use for diabetics to a comparison of two kidney cancer drugs. All five journals have adopted the CONSORT guidelines. A CONSORT rule is that authors should describe the results they plan to study before the start of a trial and stick to that list when they publish the essay.

But only nine of the 67 trials published in the five journals reported the results correctly, reported the COMPARE team on February 14 in the journal testing. One-quarter did not correctly report the primary outcome that was measured and 45% did not correctly report all secondary outcomes; others have added new results. (This varied among the journals: only 44% of trials Annals correctly rendered the main result, compared to 96% of NEJM trials.)

When the COMPARE team wrote the papers on the problematic articles, only 23 of the 58 letters were published. Annals and The bmj published all, The lancet accepted at 80%, and NEJM and JAMA rejected them all. NEJM The publishers explained that their publishers and their peers decided on the results that would be reported. Some of the CONSORT rules are "useful," they wrote, but the authors are not required to comply. Other publishers did not seem to understand that the trial researchers could change the results if they revealed the change. JAMA and NEJM stated that they did not always have the opportunity to publish all results.

When the authors of the essay responded to the letters that were published, their comments were full of "inaccurate or problematic statements and misunderstandings," discovered the COMPARE team in an article in the article. accompaniment in testing. Like publishers, many authors have misunderstood the rules of CONSORT, as well as the role of public registries to share the plan of a trial. Some attacked the COMPARE project as "outside the research community". Others have rejected the critics, complaining about the difficulty of their work. Still others have denied any results, according to the authors.

The COMPARE team said it hoped the journals would be inspired to better enforce CONSORT and review their standards for publishing letters. "We hope the editors will respond positively, constructively and thoughtfully to our findings," said Goldacre.