Those were the days. New pain relief medications known as Cox-2 inhibitors had just received Food and Drug Administration approval and the post-approval research money was flowing. Merck (Vioxx or rofecoxib) and Pfizer (Celebrex or celecoxib) lined up dozens of physicians -- orthopedic surgeons, dentists, general practitioners -- to do small studies (usually a few hundred patients or less) showing that these new drugs, no different in analgesic effect than ibuprofen or other over-the-counter medications, worked in their areas of practice.

Such trials are called "seeding trials" because their real goal is to get articles about the drugs published in specialty medical journals, which can then be delivered by drug salespersons to specialists in their community practices. I'm intimately familiar with that literature because in early 2005, I reviewed all of it for a study showing the extent to which both Merck and Pfizer engaged in seeding trials to push their new Cox-2 inhibitors. The point of doing my study was to show that the companies had little interest in pursuing the more serious question of cardiovascular risk from the drugs, which had been signaled in the very large trial that won rofecoxib (Vioxx) FDA approval. They were much more interested in seeding trials.

Now it turns out that at least one of the "thought leaders" they turned to conduct those seeding trials was himself little interested in accuracy or honesty. This morning's Wall Street Journal reports that last month, the journal Anesthesia & Analgesia retracted 10 of 11 studies on rofecoxib and celecoxib authored by Baystate Medical Center's chief of acute pain, Scott S. Reuben. The journal Anesthesiology has retracted another three articles.

The journals accused Reuben of faking the data used in the studies, which appeared between 1996 and 2008. Speaking through his attorney, Reuben suggested "extenuating circumstances" led to the scientific fraud and said he was cooperating with the journals' investigation.

Here's a typical example of his studies. Reuben and colleagues administered rofecoxib to patients about to undergo arthroscopic knee surgery. According to the now withdrawn study, those given the drug had less need for stronger painkillers after the operation.

One has to wonder: why didn't Reuben try the same experiment with ibuprofen, a commonly available non-steroidal anti-inflammatory drug that is available over-the-counter? A quick search of PubMed finds no studies authored by Reuben that tested this common painkiller.

No, it would appear that it was only the availability of drug industry funding that fueled his intellectual curiosity. After completing the trials, he not only promoted the drugs for one of the companies but urged the FDA not to restrict their use.

Someone needs to replicate his experiments using ibuprofen and celecoxib (the one Cox-2 still on the market) to see if there actually is any benefit from using these drugs in that context. It needs to be both so physicians, patients and payers have comparative data that will allow them to choose the cheapest, most effective alternative.

And if it turns out that patients given rofecoxib before operations actually had no benefit, then thousands of people undergoing minor surgery were needlessly exposed to cardiovascular risk without hope of benefit.