Wednesday, January 21, 2009

Conflict of Interest Not Disclosed Initially in Smokeless Tobacco and Cancer Review; Correction Issued, But Issue is Still Largely Misunderstood

In a review of the literature regarding the relationship between smokeless tobacco use and cancer, a conflict of interest of one of the article's authors was initially not disclosed (see: Boffetta P, Hecht S, Gray N, Gupta P, Straif K. Smokeless tobacco and cancer. Lancet Oncology 2008; 9:667-675). The original article was published in July, 2008. That article stated: "The authors declare no conflicts of interest." After the editor of the journal was notified by a researcher with knowledge of a potential conflict of interest of Dr. Stephen Hecht (one of the study authors), an erratum was published in September 2008, which noted: "During the immediate months preceding submission of the review SH was acting in the capacity of an expert witness for the plaintiff in a future court case against a smokeless tobacco company. SH declares his participation in this case in no way influenced his writing or involvement in the review."

More recently, the researcher with knowledge of this conflict of interest made an inquiry to the journal Nicotine and Tobacco Research, in which appeared a December 2008 article co-authored by Dr. Hecht on toxin and carcinogen levels in various brands of smokeless tobacco without any declared conflict of interest. The article stated: "The authors do not have any competing interest pertaining to this work." (see: Stepanov I, Jensen J, Hatsukami D, Hecht SS. New and traditional smokeless tobacco: Comparison of toxicant and carcinogen levels. Nicotine and Tobacco Research 2008; 10:1773-1782).

Initially, the journal responded by stating: "We agree that it might have been more open to disclose Dr Hecht's involvement in this case. However, he argued that the data presented in this paper represented a scientific report which was in no way influenced by his acting as an expert witness in this case. Indeed, the request to him to act in this way reflected his expertise in the field. My senior editorial board unanimously agreed that there was no intention to deliberately mislead the readers in this instance and, indeed, the need to disclose that an author is serving as an expert witness is a moot point."

A later response from the journal stated: "Before responding to you, I contacted Dr Hecht again and took the views of my team of Deputy Editors. Our list of potential conflicts of interest includes specifically payments for providing expert opinion in court - one of my Deputy Editors reminded me of that. However, in this case it appears that Dr Hecht did not receive any payment - he waived the fee. The law firm involved, as I understand it, made an equivalent donation to a US cancer charity. Thus, we accept that there is no conflict of interest to declare."

The Rest of the Story

There are several interesting issues here that are worthy of discussion.

Q: Can serving as an expert witness represent a conflict of interest?

A: Most certainly. If an individual is being paid to be an expert witness in litigation and the research in question is directly related to the testimony to be given in the case, then there is a potential conflict of interest. Think about it the opposite way and it becomes clear. Suppose that a researcher who was testifying on behalf of the tobacco industry in a lawsuit regarding whether smoking causes cancer authored a paper on the causal relationship between smoking and cancer. We would expect that relationship to be disclosed in the paper. It is not less of a conflict just because the expert is testifying for the plaintiff.

Q: In what situations does being paid as an expert witness represent a conflict of interest?

A: Being paid to serve as an expert witness represents a conflict of interest when the research in question is directly related to the testimony to be given. The key question is whether the presence of the financial relationship between the researcher and the attorney/company could be perceived as having influenced the conduct or reporting of the research. If the research is directly related to the expert's testimony, then a reasonable person might question whether the fact of being paid to testify in the trial could have influenced the conduct or reporting of the research.

It is important to note that the research must be directly related to the testimony in order for their to exist a conflict of interest. If I am testifying that an individual's smoking caused his or her lung cancer, then there is no reason why all of my research related to smoking must include a conflict disclosure. However, if my research relates specifically to the issue of lung cancer causation by smoking, then a conflict disclosure would be in order.

In the above case, it is clear that a conflict of interest exists with respect to the Lancet Oncology article. That article reviews the relationship between smokeless tobacco and cancer. Since Dr. Hecht is presumably testifying about the causation of cancer by smokeless tobacco in the lawsuit, the research is directly related to the subject of his testimony, and the conflict should be disclosed. Here, I believe the journal acted correctly in publishing the erratum to note the conflict of interest.

Q: Does the argument that an investigator was not influenced by a significant financial relationship negate the existence of a conflict of interest?

A: This is perhaps the least understood aspect of conflict of interest in research. The answer is no. A conflict of interest is not defined on whether an investigator is influenced due to a financial relationship. A conflict of interest is defined on whether the investigator could be reasonably perceived as being influenced by that financial relationship.

So the initial response of the journal Nicotine and Tobacco Research is inappropriate because Dr. Hecht's argument that "the data presented in this paper represented a scientific report which was in no way influenced by his acting as an expert witness in this case" is not relevant. The relevant question is whether or not the conduct of the review or presentation of the results could be perceived to have been influenced by his being paid as an expert witness in the case. To determine the answer to that question, one does not interview the investigator to assess whether or not he sees himself as having been influenced or not. Instead, one must examine the nature and strength of the relationship between the research in question and the financial interest and assess whether it could reasonably be perceived that the expert testimony might influence the conduct or reporting of the research.

So if the journal is correct in declaring that there is no conflict of interest here, they are correct for the wrong reason. An appropriate reason to declare the lack of a conflict of interest would be a judgment that the subject of the research (the various levels of toxicity of different types of smokeless tobacco products) is not directly related to the testimony. I cannot make such a determination without knowing the expected nature of Dr. Hecht's testimony. If he intends to testify regarding the differing levels of toxicity of various smokeless tobacco brands, then a conflict of interest appears to exist. If that is beyond the scope of his testimony, then there may not be a conflict here - but not for the reason stated by the journal.

Even the erratum published by Lancet Oncology shows a misunderstanding of this central issue. The erratum noted that the investigator "declares his participation in this case in no way influenced his writing or involvement in the review." Again, that is irrelevant. An investigator never declares that his or her participation in a case influenced his work. An investigator never declares that a financial interest caused a bias in his research. If this were about investigators voluntarily coming forward when they believed that they had been influenced by a financial interest, we would be waiting for the cows to come home.

It is critical to recognize that in most situations, the influence of a conflict of interest on the conduct or reporting of research will not be conscious or intentional. The conflict most often works by having a subconscious effect on the research. This is why conflicts of interest should not be defined by whether the financial interest actually affected the research or not. The key issue, instead, is whether the financial interest could reasonably be perceived as having influenced the research or the reporting of its results.

When the journal writes that "there was no intention to deliberately mislead the readers in this instance and, indeed, the need to disclose that an author is serving as an expert witness is a moot point," it is demonstrating a misunderstanding of conflict of interest. It is not about whether there was an acknowledgment of a deliberate attempt to mislead readers. It is about whether or not there was a conflict of interest. If there is no conflict, there is no need to disclose it. If there is a conflict, it should be disclosed. The fact that an investigator is not trying to pull the wool over anyone's eyes does not negate the responsibility to report a conflict of interest if it exists.

Q: Does making a donation to charity in lieu of accepting payment for serving as an expert witness negate an otherwise relevant conflict of interest?

A: Simply, no. The individual is still essentially getting paid for his services. The fact that he chooses to do something with this money other than put it into his personal bank account is his own prerogative, but it does not make the financial interest go away.

Thus, the journal's essential reason for claiming that a conflict of interest does not exist in this case is fallacious. The journal is basically admitting that if Dr. Hecht were being paid for his testimony, a conflict would exist and should be disclosed. However, that is indeed the situation. Whether the money goes into his own bank account or is donated to charity, he is being paid for his testimony. Money is coming out of the law firm that would otherwise not be. And its use is being directed by the expert witness.

Now, let me close by noting that I am not arguing here that a conflict of interest exists with regards to the Nicotine and Tobacco Research article. That judgment depends upon the nature of Dr. Hecht's testimony in the lawsuit and whether or not the differential toxicity of various brands of smokeless tobacco is directly related to his testimony. But in either case, the journal's reason for declaring that there is no conflict of interest is faulty.

Researchers and scientific journals need to understand that conflict of interest is not a matter of wrongdoing. Declaring a conflict of interest does not mean that a researcher has done anything wrong. And the reverse of that is also true. Not doing anything wrong does not mean that a conflict of interest fails to exist.

Conflict of interest is simply about whether or not there is a reasonble perception that an investigator's work could be influenced by a significant financial interest. The disclosure of that conflict is important because it allows readers to take that into consideration in evaluating the research and its conclusions. The disclosure is not intended to alert readers of any wrongdoing. In fact, the only wrongdoing that is commonly present is the failure to disclose these conflicts of interest.

No comments: