By Merrill Goozner

Confronted by evidence that its blockbuster heart drug harmed patients, Scios, a unit of Johnson & Johnson, almost did the right thing. It asked Eugene Braunwald of Harvard Medical School, the doyen of U.S. cardiologists, to recruit an outside panel to evaluate Natrecor, a biotech substitute for nitroglycerine.

In June, Dr. Braunwald's panel delivered what appeared to be a hard-hitting report. It recommended physicians limit use of the drug to hospital settings. That would eliminate the weekly outpatient "tune-ups" that had accounted for 90 percent of Natrecor's $700 million in sales in 2004 and generated significant fees for cardiologists. In recent weeks, the 10-member panel has pestered the company to follow through on its recommendations.

Tough medicine from an independent panel? Hardly. Their report wasn't as tough as a recent analysis in the Journal of the American Medical Association, which concluded that even in hospital settings the pricey drug should only be used as a last resort. And it was at odds with a New England Journal of Medicine article, which questioned whether the Food and Drug Administration should have approved Natrecor.

How can we account for this conflicting advice? How can practicing physicians and medical consumers sort through these varying opinions?

One way is to take a look at the financial ties of the authors. The toughest reviewer was NEJM writer Eric J. Topol of the Cleveland Clinic, who recently severed all ties with pharmaceutical companies and financial firms after The Wall Street Journal revealed his collaboration with a hedge fund betting against Merck stock. The lead JAMA author, Jonathan D. Sackner-Bernstein of the North Shore University Hospital on Long Island, performed work for GlaxoSmithKline while it held Natrecor's European rights, where it has not been approved.

The Braunwald committee, on the other hand, included the two physicians who conducted Scios' clinical trials. Even Dr. Braunwald confirmed that he has done work for Scios in the recent past. In other words, the closer the reviewers were to Scios, the less harsh they were in their evaluation. "This was not an FDA panel," Braunwald said. "They wanted the best advice, and quite frankly, they got it."

Yet, numerous studies have shown that clinical trials sponsored by companies with a stake in the outcome are much more likely to generate favorable results than trials sponsored by either the government or non-profit institutions.

These conflicts of interest have now become part of the Food and Drug Administration reform debate. When the FDA assembled advisory panels to consider the relationship between antidepressants and teenage suicides, the cardiovascular risk from COX-2 pain relievers and the safety of silicon gel breast implants -- all highly controversial with reams of conflicting medical opinion -- the agency allowed scientists with direct financial ties to the manufacturers or their competitors to sit on the panels as voting members. In the COX-2 case, those conflicted scientists were the swing votes that allowed continued marketing, votes that were later overturned by the FDA staff.

Responding to the public outcry, the House of Representatives recently enacted an FDA appropriations rider that would prohibit the agency from placing scientists with conflicts of interest on its advisory panels. At the heart of the bill is the notion that outside advisers must be scrupulously objective when weighing the evidence. FDA advisory panels should be a science court, not an adjunct to the pharmaceutical consulting industry.

The FDA is fighting the bill. At his confirmation hearing, new commissioner Lester Crawford argued the public benefits from having industry-affiliated scientists on the agency's advisory panels because they are inevitably the best and brightest in their fields. That's why industry hires them. Moreover, the staff worries that many medical fields have few experts without relationships to the firms that make products for their specialty.

The FDA is wrong on both counts. As the Lancet recently editorialized, "It is hard to believe that in a country with 125 medical schools -- not to mention the pool of international experts -- the FDA cannot find experts who do not have financial ties with the companies whose products are under review." Indeed, if the FDA staff would just canvass the nation's top medical schools, they'd find hundreds of qualified scientists ready to leap at the opportunity to serve the public as objective arbiters of the latest advances in medicine.

The headlines suggest we live in the greatest era of medical progress known to man. But the reality at the FDA is far more mundane. Many new therapies offer only marginal improvements for most patients. The latest cancer drugs, for instance, offer additional months of life, not years. Many new drugs are no more effective than medicines already on the market and may entail trading off one risk for another. The COX-2s are a tragic example.

When the FDA sits down to calculate the tradeoffs between these risks and benefits, it must sift through limited data derived from clinical trials sponsored by the firm that manufactures the product. It's often a close call. That's why it has 30 advisory panels -- to give it unbiased, objective advice. The public will find those panels far more credible if the members do not simultaneously work for companies whose products are under review.

Merrill Goozner is author of "The $800 Million Pill" (University of California Press), and directs the Integrity in Science project at the Center for Science in the Public Interest, www.cspinet.org.