J&J Seeks Objectivity, But Does It Seek All Truths?

The Wall Street Journal reports this morning that Johnson & Johnson's Scios unit "named" Duke University and the Cleveland Clinic to run a $100 million trial testing the safety and efficacy of Natrecor, a heart-failure drug widely promoted for off-label uses that may cause kidney failure among other unwanted side effects. For background on this story, see my op-ed, posted here, from a year ago.

While I'm sure that the two institutions will do all in their power to insure the independence of this high-profile study, the key word in the story is highlighted in quotes above: J&J NAMED the institutions. The writer could have used the word "chose" or "selected." Whatever. The point is that industry-funded studies will always bear the taint of conflict of interest as long as the sponsor, who has the most at stake, plays a role in organizing the study.

This is as much true in the world of environmental science as it is in pharmaceutical science. Every study that has ever been done on the subject shows a correlation between outcomes and sponsors. In other words, a study funded by a particular source is likely to come up with an outcome that reflects the self-interest of that source.

How to avoid this taint? Simple. If J&J needs a study done, it should be able to pay a fee -- the cost of the study -- to an independent agency, which then selects the clinicians and statisticians who will conceptualize the protocols of the trial. If J&J wants to spend $100 million on a clinical trial involving Natrecor, there are probably many more questions of interest to learn besides whether this particular drug is safe.

For instance, how does it compare to every other drug out there for the same condition? How useful is it when given off-label (the greatest expense associated with use of the drug)? How does it stack up in a cost-benefit analysis compared to other drugs for the same condition.

Memo to the Duke and Cleveland Clinic investigators involved in this trial: Curious minds would want to know.