Protocol of convenience

December 19, 2005|By MERRILL GOOZNER

The specter of researchers hiding damaging data when drug companies financed their clinical trials is once again haunting the medical publishing establishment.

Last week, the editors of The New England Journal of Medicine accused Merck-funded researchers of not reporting three deaths in the trial that led to the approval of Vioxx, the pain reliever subsequently pulled from the market because it caused heart attacks in some patients.

Medical editors are once again scrambling for better ways to manage these conflicts of interest. The editors of The Journal of the American Medical Association have taken the extraordinary step of requiring every industry-funded researcher who submits a study based on a clinical trial to hire an independent statistician to analyze the data.

They also demanded that the principal investigator vouch for the integrity of the data and the accuracy of the data analysis - an approach that's comparable to what corporate accountants now face in the wake of the Enron Corp. scandal.

Unfortunately, piling new disclosure rules on medical researchers will not address the core problem. When industry has exclusive control over studies of its new drugs and medical devices, the most important questions usually do not get asked.

Most trials aimed at gaining Food and Drug Administration approval for a new product compare it to a placebo. Instead of asking whether it is better than what is already out there, they ask if it is better than nothing. As often as not, it is.

In some cases the clinical-trial protocol designed by industry-funded researchers compares the new drug to a rival drug but at doses designed to put it in the best light. That was the case with Vioxx, whose original trial compared it to high doses of naproxen to show that it caused less stomach distress. When the data showed Vioxx patients with four (now five) times more heart attacks, the Merck-funded researchers waved off that red flag by claiming naproxen probably was cardio-protective.

Merck defended its study last week by saying all the data had been turned over to the FDA, which still approved the drug. And in that sense, the company was right. The original study never tested enough people to definitively prove the drug caused heart problems - or, conversely, to show that it was safe.

Less than a year after the original trial was published, cardiologists at the Cleveland Clinic published an analysis questioning the safety of not only Vioxx, but also the entire class of so-called Cox-2 inhibitors (Celebrex and Bextra are the other two approved drugs in the class).

They called for a major new trial large enough to test that hypothesis. That trial was never done. Instead, the industry funded hundreds of studies testing the Cox-2s for every imaginable ache and pain, which appeared in every conceivable medical specialty journal as part of the drug companies' overall marketing campaign.

The way to get better science in front of regulators and physicians (who rely on the medical literature) is to have an independent body take control of FDA registration trials. The drug and medical device industries already pay user fees to the FDA to hasten reviews of their new product applications. The user-fee law is up for renewal next year. Why not expand those fees to include paying the government to conduct the final product registration trials?

It could be done through a new institute in the National Institutes of Health, which could contract with investigators at universities, independent research institutes or the private firms that have sprung up in recent years to conduct clinical trials. The most important thing is that the trials be designed to answer the right questions - is the new product truly safe and how does its effectiveness compare to what's already out there. Regulators are facing tougher and tougher calls on drug safety.

The government next year begins paying a large share of the nation's drug tab through insurance companies, which will influence which drugs seniors buy through their co-pays. Americans will be better off if those decisions are not based on information provided by medical literature written by researchers in industry's back pocket.

Merrill Goozner, author of "The $800 Million Pill", directs the Integrity in Science project at the Center for Science in the Public Interest. His e-mail is This article first appeared in the Chicago Tribune.

Baltimore Sun Articles
Please note the green-lined linked article text has been applied commercially without any involvement from our newsroom editors, reporters or any other editorial staff.