The credibility of the New England Journal of Medicine is at stake after the former Editor of the British Medical Journal, Richard Smith, publicly criticized editorial decisions that he says delayed the publication of corrections to Merck & Co's VIGOR study of the now-withdrawn painkiller Vioxx (rofecoxib).

Writing in the Journal of the Royal Society of Medicine, Dr Smith said: "It is unfortunate that the New England Journal of Medicine waited five years before publishing a correction to its VIGOR study, which was published in November 2000. Concerns about the correctness of the study were raised with the NEJM's editor, Jeff Drazen, as early as August 2001, the drug was withdrawn in 2004 and yet it wasn't until December 2005 that the journal published its first correction."

The VIGOR (Vioxx Gastrointestinal Outcomes Research Study) trial involved some 8,000 patients who were given either the conventional non-steroidal anti-inflammatory drug naproxen or the COX-2 inhibitor Vioxx, and showed those given the newer drug experienced less gastrointestinal side effects than those on naproxen (121 versus 56).

However, VIGOR also came under scrutiny after demonstrating significantly fewer heart attacks in patients taking naproxen compared to those receiving Vioxx. Merck asserted for years that this effect was a result of naproxen's ability to block platelet aggregation and so has a cardioprotective effect, but was forced finally to withdraw the drug in 2004 in light of new evidence from the APPROVe trial showing it in fact doubled the risk of heart attacks and stroke in patients taking it for more than 18 months. Most recently, a new analysis of the APPROVe trial suggests Vioxx may even exert these adverse cardiovascular effects after as little as four months' treatment.

Then, in December last year, the NEJM alleged that the authors of the article deleted pertinent data shortly before submitting the manuscript for publication, which emerged only after they were made public by the US Food and Drug Administration. Merck insisted that the three excluded heart attack cases did not materially affect the conclusions of the study and were reported after the pre-specified cut-off date for the article, while the NEJM has since defended itself by saying the correct data were available to view on the FDA's website.

But Dr Smith says that is not good enough. “Doctors, pharmacists and health professionals do not obtain their information by searching through copious data on the FDA website; they get it from leading journals," he said. "The journal failed its readers. It has damaged its reputation.”