ATLANTA - Relationships between physicians and the drug industry are a fact of life, but nevertheless one that gives many people pause. In an editorial published in the Nov. 1, 2007, New England Journal Of Medicine, Eric Campbell of the Institute for Health Policy at Massachusetts General Hospital and Harvard Medical School wrote that "the nature, extent and consequences of physician's relationships with industry have become one of the most fiercely debated issues in health care today."

But to hear the panelists at a special session on "Practical and Ethical Issues Arising During The Conduct of Pharmaceutical Industry-Sponsored Research" tell it at the American Society of Hematology Annual Meeting on Saturday, drug-company-sponsored clinical trials are like democracy: Definitely a most imperfect system, but, warts and all, still better than the alternatives.

The basic attitude of the panelists on such relationships was positive. "The kind of research we are all interested in doing . . . depends enormously on relationships with the industry. That's where most of the new drugs come from, and that's where most of the new ideas come from," panel chairman Charles Schiffer of the Barbara Ann Karmanos Cancer Center told the audience, a point that was echoed by panelist Frank Giles of University of Texas. "There's sure not a lot of drugs coming from any other arrangement" than drug company and venture capital funding, he told the audience.

Still, conflicts of interest are an inherent part of such relationships, and the panelists shared their perspectives on the pitfalls, as well as how to keep it honest.

Money may be the least of the conflicts. "Much of the conflict-of-interest furor has focused on financial conflicts," Schiffer told the audience. "But frankly, many of us think other conflicts are much more important."

Likewise, the drug industry is not the only industry with conflicts of interest surrounding the publication of scientific data - the panelists discussed scientific journals, contract research organizations and scientific societies - and for the drug industry, the watchful eye of the FDA provides plenty of incentive to be good. "This has always been a highly regulated industry," said panelist David Parkinson, CEO of Nodality Inc. "We are tracked by regulators, and subject to felony charges if we do it wrong."

Other conflicts, also known as intrinsic conflicts, are specifically the fame and career advancement that comes with running and publishing positive clinical trials. Recent studies show that many patients are more concerned about such intrinsic conflicts than financial ones. "In essence, patients are not concerned whether we're getting rich, but they are concerned whether we're getting famous," panelist Richard Schilsky of the University of Chicago told the audience.

But whether researchers are in it for the money, the fame, the betterment of mankind or all of the above, the panelists had a variety of pointers for making sure the reward is earned honestly.

First and foremost, the panelists stressed that the rewards system for bringing patients to a clinical trial needs to change. Currently, doctors enrolling a few patients in a large multicenter trial but often having no intellectual input into trial design are nevertheless rewarded by co-authorships on the publications detailing the trial.

Parkinson pointed to experimental physicists as a group that has figured out how to distribute both credit and rewards among "large, long, geographically distributed projects," and stressed that honorary authorship is not the way to go: to be authors, "investigators need to contribute ideas, not just patients."

Those who are co-authors need to make sure the paper that ultimately is published accounts for all the data that originally were collected, that planned and after-the-fact analyses are distinguished, and that the results that are reported are the predetermined study endpoints, rather than whatever data are most favorable to the drug being studied. Schilsky said that there have been "egregious" examples of whole arms of a study disappearing from final publications, as well as results that bear only the most cursory resemblance to the study as it originally was planned.

Parkinson pointed out that rooting out such shoddy analyses can and should be done by journals as well as authors. "I continue to be amazed that Phase III trials are reviewed by journals in the absence of a protocol document," he told the audience.

Assuming the data to be published are indeed a legitimate representation of the trial, the next question is whether they are new. Schiffer criticized the repeat publication of data with unclear novelty value. The panel discussed legitimate reasons for that, including the fact that different scientific meetings have different audiences and the fact that data do indeed change over the course of a study. But presenters need to make clear what is new about data being presented, and make sure that they are worth presenting a given update.

Schilsky took a particularly strong stand on the publication of Phase II trials. "Phase II trials are just sort of out there," he said, and are not designed to change clinical practice no matter how promising their results. "This is actually very dangerous" because phase II trials are designed to decide whether a Phase III trial will be worth the investment.

Conversely, investigators have a responsibility to make sure that negative trials get their day in the spotlight, too. Several factors work against such publications, though the root cause of most of them is that people - from investigators to journal editors - like to see good news, making it hard to find first the motivation and then the journal space for studies that didn't work out.

There was general acknowledgement that there is a need for the public to know when a trial has failed. Several panelists pointed to mandatory clinical trials reporting on sites like clinical trials.gov as a big step in making sure that doctors get to know about good, bad and ugly clinical trials alike.

But though the public can suspect nefarious aims when negative studies are not published, Parkinson said that on the part of sponsors, it is often more a case of not throwing good resources after bad: "It's not suppression," he said. But when a trial fails, "there is no agent, there is no market. It's people moving on."