Delivered at the HHS Conference on Financial Conflicts of Interest 8/16/00
As we know, this conference was a response to Secretary Shalalas charge to identify new or improved means to manage financial conflicts of interest that could threaten the safety of research subjects or the objectivity of the research itself. That language is important, and Ill come back to it.
Ill begin by defining a financial conflict of interest, because theres often some confusion on that score. A financial conflict of interest, I believe, is any financial association that would cause an investigator to prefer one outcome of his research to another. Let me give you an example. If an investigator is comparing drug A with drug B and owns a large amount of stock in the company that makes drug A, he will prefer to find that drug A is better than drug B. That is a conflict of interest.
Note that its a function of the situation, not the investigators response to it. If the investigator then finds that drug B is better, he may swallow his disappointment and report the facts objectively -- or he may not. Thus, there is nothing potential about a conflict of interest. Either it exists or it doesnt. What is potential is whether a conflict leads to bad research.
Note also that financial conflicts of interest are not inherent to the research enterprise. Theyre entirely optional -- unlike the intellectual or personal conflicts of interest to which theyre often compared such as the desire for Nobel-worthy results.
At one time, financial associations with private industry were largely confined to drug companies awarding grants to academic institutions for research in areas of interest to both of them. In the best institutions, this was done at arms length. The companies had no part in designing or analyzing the studies, they did not own the data, and they certainly did not write the papers and control publication. The academic institutions were the watchdogs.
Things have changed dramatically in the past few years. Arms-length relationships are a thing of the past, and financial arrangements are hardly limited to grant support. Companies now design studies to be carried out by investigators in academic medical centers who are little more than hired hands supplying the human subjects and collecting data. The companies own the data, analyze it, and control publication. Such a study is not necessarily of any real scientific importance or even interest to the investigators; it may instead be only of marketing importance to the sponsor. That is particularly true of phase 4 studies.
For their part, academic institutions are increasingly involved in deals with the same companies whose products their faculty members are studying. Some institutions are for a price -- allowing companies to set up research outposts in their hospitals and giving them access to students and house officers, as well as to large numbers of patients. They are aligning their interests with those of industry and allowing the boundaries between them to become blurred.
Consider the story in last Thursdays Wall Street Journal. It reported that Targeted Genetics Corporation will acquire Genovo, Inc. As part of the deal, James Wilson will receive 13.5 million dollars worth of Targeted stock in exchange for his 30 percent equity interest in Genovo. Recall that the University of Pennsylvania permitted Wilson to own a piece of Genovo even while he was doing research on its products. Thats hardly surprising, given that Penn itself will receive 1.4 million dollars worth of stock for its 3.2 percent stake in Genovo. Some watchdog. This story is news because of the death of Jesse Gelsinger, but its hardly unique in its outlines.
What about the integrity of the scientific literature? As Tom Bodenheimer summarized so well at this conference, there is plenty of good evidence that investigators with financial ties to companies whose products they are studying are indeed more likely to publish studies favorable to those products. And in my two decades at the NEJM, it was my clear impression that papers submitted by authors with financial conflicts of interest were far more likely to be biased in both design and interpretation.
In my view, the pervasive and manifold financial conflicts of interest that now exist have a number of bad effects, in addition to the threat to the integrity of the scientific literature and the risks to human subjects that primarily concern us here. I dont have time to discuss the other effects today. Suffice it to say that we need to remember that the mission of investor-owned companies is quite different from the mission of academic medical centers. The primary purpose of the former is to increase the value of their shareholders stock, which they do by securing patents and marketing their products. Their purpose is not to educate, nor even to carry out research, except secondarily or as a means to their primary end. I believe academics often forget this, and allow themselves to believe that marketing is really education.
With all this as background, what I would like to do in the time remaining to me is say a few words about the remedies proposed in this conference and why I find most of them unsatisfactory.
The emphasis has been on managing conflicts of interest. Managing is the word used in Secretary Shalalas charge. In particular, the focus has been on disclosure to human subjects and on institutional oversight.
In my view, disclosure simply passes the buck to the patient/subject, who is left to wonder how the investigator will balance his competing interests. Thats neither fair nor helpful to the patient, who may be both sick and desperate, and it certainly does nothing to remove the conflict of interest. Caveat emptor is simply not appropriate in this setting. On the other hand, not disclosing a conflict of interest is worse, because its fundamentally deceptive. Patients naturally assume investigators are primarily interested in patient welfare, and they have a right to know of anything that may shake that assumption. To disclose or not to disclose, then, is a Hobsons choice, and it points to the underlying problem the very existence of financial conflicts of interest.
Institutional oversight usually boils down to defining conflicts of interest that must be disclosed and documenting them. Out-and-out prohibitions are few. Some of the guidelines are truly mind-boggling in their complexity and detail. For example, Harvard Medical Schools guidelines, widely acknowledged to be among the most stringent, fill eight closely worded and nearly impenetrable pages. They deal with such matters as the dollar amount of equity interest investigators and specified family members may own in companies sponsoring their research. In short, all of this is about management not about the wisdom of permitting conflicts of interest in the first place. Its as though financial conflicts of interest were a given of nature or a Constitutional right neither of which they are.
I believe we need to stop dancing on the margins of this issue and deal with it head-on. The important question is whether financial conflicts of interest should exist, not how to work around them. Mind you, Im not opposed to cooperation with industry. Academic/industrial collaborations have led to some important advances. But the conflicts of interest that now abound in academic medicine go way beyond cooperation and many of them have no conceivably useful social purpose.
I would like to put forth the following recommendations:
I am well aware that my proposals might seem radical. Thats because our society is now so drenched in market ideology that any resistance is considered quixotic. But medicine and clinical research are special, and I believe we have to protect their timeless values of service and disinterestedness. Patients should not have to wonder whether an investigator is motivated by financial gain, and the public should not have to wonder whether medical research can be believed. The only way to deal with the problem is to eliminate it as much as possible.
Marcia Angell, M. D.