Wednesday, November 02, 2011

When is a Financial Conflict of Interest No Longer a Conflict?

In the past two sets of commentaries, I have revealed examples of tobacco control researchers who failed to disclose past conflicts of interest with the pharmaceutical industry. Although failing to report these "prior" conflicts was in compliance with the journal or professional organization policy, I opined that although they occurred in the past, these conflicts were still significant and relevant and therefore should have been disclosed to readers nonetheless.

Today, I address the tricky issue of: When is a conflict no longer a conflict?

Let's start with two examples:

Example 1: Professor X is a chemist at the University of Louisville. He studies the levels of particulate matter and other chemicals in ambient air for use in exposure assessment. In 2011, he publishes a paper entitled "Levels of Secondhand Smoke in Owensboro Bars: No Evidence that Regulation is Necessary." For the past 3 years, he has been employed at the University of Louisville and hasn't had any funding or consulting income from any corporations. The journal requires disclosure of any financial conflicts of interest within the past 3 years.

Is it acceptable for Professor X to file a "no" disclosure (i.e., not to disclose any financial conflicts of interest)?

According to the reasoning being used by Dr. Fiore, Dr. Carpenter, and many others in tobacco control (and according to numerous comments I received from tobacco control colleagues in response to my commentaries), the answer is a resounding "No." Since the journal only requires the disclosure of conflicts within the past 3 years, Professor X need not disclose any financial conflicts of interest. There is nothing wrong with him claiming that he has no financial conflicts of interest, and there is no reason for journal readers to have any more information about past potential conflicts.

Example 2: Professor Y is an epidemiologist at the Health Policy Institute. He studies the effects of various exposures on disease risk, as he has done for decades. In 2011, he publishes a paper entitled "No Evidence of Increased Heart Disease Risk among Secondhand Smoke-Exposed Workers. For the past 3 years, he has been employed at the Health Policy Institute and neither he nor the Institute had any funding or consulting income from any corporations during that time period. The journal requires disclosure of any financial conflicts of interest within the past 3 years.

Is it acceptable for Professor Y to file a "no" disclosure (i.e., not to disclose any financial conflicts of interest)?

According to the reasoning being used by Dr. Fiore, Dr. Carpenter, and many others in tobacco control (and according to numerous comments I received from tobacco control colleagues in response to my commentaries), the answer is a resounding "No." Since the journal only requires the disclosure of conflicts within the past 3 years, Professor Y need not disclose any financial conflicts of interest. There is nothing wrong with him claiming that he has no financial conflicts of interest, and there is no reason for journal readers to have any more information about past potential conflicts.

The Rest of the Story

Here is the rest of the story about Professor X. As it turns out, Professor X spent most of his career as a research chemist in the Laboratory Division of Brown & Williamson. He worked there for 34 years, during which time he produced articles that helped the company deny that nicotine is addictive, that secondhand smoke is hazardous, and that active smoking is as harmful as public health authorities claim. Four years ago, he "retired" from Brown & Williamson and took a position at the University of Louisville.

How many people in tobacco control still think that Professor X has no obligation to disclose his prior conflicts of interest? I would venture to say none. Moreover, I would venture to say that if this occurred in real life, a contingent of anti-smoking groups would publicly attack Professor X for failing to disclose his prior employment at Brown & Williamson and that they would attempt to discredit the study on that basis.

Now here is the rest of the story about Professor Y. As it turns out, in 1985 Professor Y became Vice President of the Thomas Institute Policy Analysis Center (TIPAC), a consulting firm funded initially by a $400,000 grant from the Brown & Williamson Tobacco Corporation. Following its initial formation, TIPAC continued to receive hundreds of thousands of dollars in funding annually from a number of tobacco companies. Professor Y worked on Research & Development projects for B&W Tobacco, such as analysis of the sensory perception of smoke and how to reduce the amount of tobacco in cigarettes. By 1995, Professor Y was a full-time consultant on environmental tobacco smoke issues for the Tobacco Institute. In 2005, he took a position with the Health Policy Institute. Both his initial salary and the Institute's start-up costs were paid for by Philip Morris. For the last four years, however, the Institute has found private donors and has not had to rely on any tobacco industry funding.

How many people in tobacco control still think that Professor Y has no obligation to disclose his prior conflicts of interest? I would venture to say none. Moreover, I would venture to say that if this occurred in real life, a contingent of anti-smoking groups would publicly attack Professor Y for failing to disclose his prior history of involvement with the tobacco industry and that they would attempt to discredit the study on that basis.

It is clear from these examples, I hope, that past conflicts of interest are indeed critical for readers to know about. Just because Professor X and Professor Y had nothing to disclose during the past three years does not indicate that they have no financial conflicts of interest that need to be disclosed in their current publications, the journal's requirements notwithstanding.

Furthermore, it would be hypocritical and inconsistent to argue that past conflicts are only relevant when it comes to tobacco. If past conflicts are relevant, then past conflicts are relevant.

Therefore, it is my feeling that researchers do need to disclose their past conflicts (albeit briefly). A financial conflict of interest does not disappear simply because the conflict is no longer active.

Let me reinforce this point with one further example:

In 2007, a tobacco researcher who is near and dear to you (actually, despised by many may be closer to the truth) - Dr. Michael Siegel - accepted $400,000 from a start-up electronic cigarette company to advise the company on how to best market its products, deal with potential FDA and other regulatory matters, and conduct valid scientific studies to test the safety and effectiveness of the product. The consulting work concluded by the end of 2007, Dr. Siegel was paid in full in that year, and has not had occasion to do any further paid consulting for electronic cigarette companies.

Now, in 2011, Dr. Siegel is (I am indeed) about to publish a study in which he reports the results of qualitative research with electronic cigarette users to determine the effectiveness of the product. How many of you believe that there is no need to disclose my 2007 receipt of nearly a half million dollars from an electronic cigarette company because it was more than 3 years ago? (the journal to which I am submitting the paper requires only conflicts within the past 3 years to be revealed)

I would be willing to bet (up to $400,000 in fact) that if I did not disclose my prior financial relationship with an e-cigarette company, anti-smoking advocates would grill me over the coals for failing to disclose that fact, even though it occurred four years ago and I am not required to report it by the journal. I am sure I would be attacked even though I would be in full compliance with journal and institutional policies.

Clearly, past conflicts of interest are relevant and I believe that researchers have the obligation to disclose past conflicts that are related to the research in question.

Is there any circumstance in which a past conflict of interest does indeed "disappear" or become inactive to the point that it need not be disclosed?

I can think of just one. If a researcher were to renounce his past involvement with a particular industry, vow to never accept money from that industry in the future (whether in the form of grants, employment, honoraria, consulting, lecture fees, etc.), to make this vow public (by publishing it on a personal or institutional web site), and to demonstrate the commitment by going a certain period of time (e.g., 5 years) in compliance, then one could potentially argue that the investigator has truly put the financial relationship in the past. As there is no opportunity for any funds to be received from that industry ever again, this could in fact put an end to the potential influence of the previous conflict.

Short of that, there is simply no way to argue that a past conflict is no longer present. For example, in the example above, I accepted e-cigarette money in 2007. I then write a paper in 2011 about the effectiveness of e-cigarettes and I don't disclose any conflicts since none are "current." Then, the day after the paper is published, I get a call from "Heavy Metal Cigs," which wants to pay me $1 million for helping them plan a definitive clinical trial on the effectiveness of electronic cigarettes (note: my phone lines are open). So was I really unconflicted in 2011 when I published my paper, or was I just waiting for the right offer to come along? And how could we ever know the willingness of any particular researcher to take corporate money in the future, short of a public vow?

The rest of the story is that prior conflicts of interest - even if "active" before a journal's "disclosure period" - need to be disclosed if they are directly related to the topic of the research, as they are still potentially relevant to the reader. If the conflicts are small in magnitude and unlikely to affect the appearance of impartiality, then that is a judgment that the reader is free to make. But without the information, then we clearly lack the information we need to be able to make an informed judgment.

Unfortunately, the last example was only an example. Had I actually received $400,000 in 2007, I would certainly not still be driving my 120,000+-mile 1998 Saturn SL.

No comments: