Over the course of the past several years, I have provided a number of critiques of research whose conclusions I believed were flawed and in which one or more investigators had a financial conflict of interest, leading me to suggest that the situation had the appearance of the conflict possibly affecting the interpretation and presentation of the research findings and conclusions. It has come to my attention that a number of readers have misinterpreted these commentaries as having questioned the integrity or character of the investigators with the conflicts of interest.
Today, I want to clarify and state explicitly that this is not the case. I do not believe that having a financial conflict of interest represents any wrongdoing, nor do I think that allowing such an interest to affect one's interpretation of the study findings represents any wrongdoing or reflects on the integrity, character, or honesty of the investigator.
Rather, I think that these situations involve bias, which is not related to wrongdoing, dishonesty, or lack of integrity. Bias is simply a perspective through which one sees, designs, conducts, interprets, and presents research. It has no moral value; it is simply the existence of a perspective, schema, or frame that tends to lead to the systematic interpretation of research data in a particular direction, or with a particular slant or angle. There is nothing wrong with it from the perspective of ethics, character, or integrity. In short, it is simply a reflection of being human.
Let me flesh this out a bit.
First, there is nothing wrong with an investigator having a conflict of interest. In fact, much of the great progress in the area of pharmaceuticals has been made specifically because researchers with relationships with pharmaceutical companies have been involved in the drug development process. Pharmaceutical funding of clinical trials is critical and without it significant advances in medical treatment would not occur. Researchers consulting for pharmaceutical companies is a natural offspring of these collaborative relationships and can contribute positively toward the identification and development of new and more effective medications. So there is nothing wrong with an investigator possessing a conflict of interest.
Second, if an investigator has a conflict of interest and conducts research despite that conflict, she is not doing anything wrong or unethical, assuming that the research is in compliance with her university's conflict of interest policies.
Third, even if a researcher is influenced by a conflict of interest in the conduct or presentation of research results, it is not wrongdoing. It does not necessarily reflect on the character or integrity or the investigator. Except in rare cases, the influence of conflicts of interest is subconscious. The concern is not that the investigator will intentionally and knowingly skew the presentation of results to favor the drug under study. That would be blatant investigator bias and would be unethical. However, in the overwhelming number of cases (and in all of the ones I've blogged in the past few months), the concern is that a conflict of interest could have a subconscious effect on the reporting of the results and the interpretation of findings. When this occurs, it is not a reflection of wrongdoing or lack of integrity; rather, it is simply a reflection of being human.
Several investigators have responded to my comments by stating things like: "I would never allow a conflict to affect my research." These comments show a misunderstanding of what I am arguing. I am not arguing that investigators are purposely skewing their results because they have a financial relationship with a company and want to make sure that the research doesn't hurt the financial value of the company. Instead, I believe that conflicts of interest operate beneath the consciousness of the investigator. By definition, the conflict would influence the investigator without his knowledge or awareness. You cannot be human and at the same time be completely immune from having a financial relationship affect the lens through which you view research. It's only human.
Let me illustrate this by being honest and disclosing bias in my own work. As my readers well know, I have been quite critical of the American Cancer Society for its role in supporting the FDA tobacco legislation which I believe is a huge victory for Philip Morris at the expense of the public's health. However, during the years that I held an ACS research grant, I would be willing to bet that I was less likely to criticize the ACS during that time. I am not aware of any situation in which I consciously decided not to criticize the ACS because of my grant funding from that organization. However, it is very likely that subconsciously, I avoided even becoming aware of ACS actions to potentially criticize. My eyes probably focused elsewhere during that time. So yes, my receipt of an ACS grant did create a bias: I was systematically less likely to criticize the ACS or to view the ACS' actions with as much critical analysis as I normally would. This was not a conscious effort on my part. But it occurred.
I would go so far as saying that I would (subconsciously) view any organization more positively if it were providing funding to me, funding that was helping me to support my family and kids. And if that organization was a company that manufactured drugs, I would (also subconsciously) also view those drugs in a more positive light.
So my previous criticism of studies involving conflicts of interest should not be interpreted by readers as asserting that the investigators did anything wrong or lacked integrity. In contrast, these are investigators with high standards of integrity who are human, and therefore are potentially influenced subconsciously by financial conflicts of interest.
Finally, I hope readers will understand that in my view, addressing issues such as the interpretation of the data and disclosure of conflicts is not an “attack” on anyone; rather, it is the opposite: it helps ensure the integrity of the science and the scientific process. In fact, rules about conflict of interest are set up not to penalize investigators for wrongdoing, but to protect them from potential questioning of the possible influence of the conflict on their work. My hope is that these commentaries will help advance the field, if nothing else by forcing us to re-think and reassess our current policies and programs (even if we end up making no changes, it still does us a service I think). I hope my commentary will be seen in that light.
I have the utmost respect for the work and integrity of scientists who have studied the effectiveness of NRT and view them as valuable colleagues in a common battle against tobacco-related disease and death. Helping to ensure that policy decisions are made based on unbiased, accurate, and free-from-financial conflict conclusions is not an attack on individuals; it is an attempt to improve our understanding of the scientific evidence and to protect scientists from having their work undermined because of the appearance of potential conflicts.
No comments:
Post a Comment