Conflict of interest bias

In discussions about controversial topics such as climate change, vaccine safety, environmental sustainability, pesticide toxicity, chronic Lyme disease, male privilege or homeopathic therapies, we often hear the argument that scientific studies are biased due to financial conflicts of interest of the researchers. The accusation of conflicts of interest is used to discredit studies, but we have to be careful to avoid a conflict of interest bias: an unjustifiable asymmetry where we see the conflicts of interest of the opponent but not the conflicts of interest of those people holding our own views. This conflict of interest bias is a version of the disconfirmation bias, where we are more critical and distrustful towards those people or studies that disconfirm our prior beliefs.

The most extreme example of a conflict of interest bias is probably the case about chronic Lyme disease. There is no evidence that chronic Lyme disease is caused by a persistent bacterial infection that can be treated with long-term antibiotic therapy. Medical associations such as the Infectious Disease Society of America (IDSA) advise against long-term antibiotic treatment, because antibiotics are ineffective in this case and a long-term therapy is expensive and can be harmful. However, Connecticut Attorney General Richard Blumenthal accused the IDSA of having undisclosed financial conflicts of interests held by several IDSA Lyme disease panelists (however, he did not name those panelists, nor did he clarify the kind of conflicts of interest). There is no evidence for such undisclosed conflicts of interest. Blumenthal and chronic Lyme disease pressure groups reject the guidelines of medical associations such as the IDSA, using the accusation of conflicts of interest as a weapon to discredit their opponents. However, they are not so critical about the possible conflicts of interest of people holding their own views. In the case of antibiotic therapy against chronic Lyme disease, one could equally say that patients involved in those pressure groups have a financial conflict of interest when they argue for insurance coverage of long-term antibiotic therapy. And of course, Big Pharma (the pharmaceutical industry) could generate a conflict of interest, trying to sell their antibiotics to chronic Lyme disease patients. How can the IDSA panelists be profiting financially by recommending to not treat patients with antibiotics?

The case of antibiotics brings us to a second example of conflict of interest bias: homeopathy. Take for example many organic livestock farmers: they often refuse to give their sick animals antibiotics, claiming that those antibiotics are harmful and merely serve the profits of the pharmaceutical industry. Instead, those farmers use homeopathic therapies for their animals. However, not only is there a scientific consensus that homeopathy is less effective than antibiotics in treating bacterial infections (at most, homeopathy has a placebo effect). A lot of studies that demonstrate the effectiveness of homeopathy have conflicts of interest. For example, researchers were paid by companies that sell homeopathic products.

The case of organic farming brings us to a third example of conflict of interest bias. Proponents of organic food claim that a lot of scientific studies that indicate that organic food is not better for our health and the environment, were performed by scientists who had conflicts of interest with the non-organic agricultural industry (e.g. with companies like Monsanto). Those proponents overestimate the conflicts of interest of the counterparty and they underestimate the conflicts of interest of their own party. A lot of scientific studies that claim that organic food is better for our health and the environment, or that non-organic genetically modified crops are unsafe, were performed by scientists who had conflicts of interest with the organic agricultural sector. Some infamous names include: Charles Benbrook (had undisclosed conflicts of interest: worked at the Organic Center and research was funded by Whole Foods, Organic Valley, United Natural Foods, Organic Trade Association and others), Gilles-Eric Séralini (consultant of Sevene Pharma that sells homeopathic antidotes against pesticides), Judy Carman (her anti-GMO research was funded by Verity Farms and published in a journal sponsored by the Organic Federation of Australia) and the Rodale Institute (a research institute that has a commercial interest in organic farming by selling organic products). These (often undisclosed) conflicts of interest are at least as bad as the conflicts of interest of e.g. Monsanto selling GMOs and pesticides. Imagine how environmentalists would react if proponents of GMOs came up with studies that had similar conflicts of interest with Monsanto.

The case of Monsanto brings us to a fourth example of conflict of interest bias. Monsanto sells the herbicide glyphosate, so of course they want to deny that glyphosate is toxic or carcinogenic. Opponents of glyphosate warn that studies showing the safety of glyphosate are biased due to the close ties between scientists and the pesticide industry. However, there are potential financial conflicts of interest among the opponents of glyphosate. For example farmers who developed non-Hodgkin lymphoma and who support the environmentalist cause against glyphosate, aim for compensation fees from Monsanto. This could generate conflicts of interest: American scientist Chris Portier proposed that the International Agency for Research on Cancer (IARC) reviewed glyphosate and participated as an outside expert on the IARC glyphosate panel. In March 2015, the IARC judged that glyphosate is probably carcinogenic (although most other regulatory agencies such as the U.S. EPA, the European Food Safety Authority, the WHO-FAO, the German Federal Institute for Risk Assessment, Health Canada and Australia’s Pesticides and Veterinary Medicines Authority declared glyphosate to be safe). Right after that IARC judgement, Portier became litigation consultant for law firms that were bringing lawsuits against Monsanto on behalf of plaintiffs who claimed that their cancer was caused by glyphosate. There are other cases of conflicts of interest between the IARC and scientists who are hired by law firms in litigations against industries. IARC reports serve as evidence in lawsuits where plaintiffs hope to win cases worth hundreds of millions of dollars in fees.

Speaking about compensation fees, a fifth serious example of conflict of interest bias can be seen in the antivaccination movement. Opponents of vaccines often claim that studies demonstrating the safety and effectiveness of vaccines are invalid because they are supposedly influenced by the pharmaceutical industry that wants to make profits from selling vaccines. However, a lot of members of the antivaccination movement also have conflicts of interest: they want compensation fees from the pharmaceutical companies for the alleged damages they incurred from vaccines. One influential person in the antivaccination movement is Andrew Wakefield, the author of an infamous study about an alleged link between measles vaccines and autism. Wakefields research about the MMR-autism connection contained undisclosed conflicts of interest, because he was paid by lawyers who were suing for vaccine injuries. Of course, Wakefields studies could help those lawyers in the lawsuits against the vaccine producers. Another example in the antivaccination movement is researcher Romain Gherardi, who published a study with undisclosed conflicts of interest, as he was paid by E3M, a French association of patients with MMF (Macrophagic Myofasciitis). They believe their disease is due to the aluminum in vaccines. As with the other examples of conflict of interest bias, antivaccination activists focus on the conflicts of interest by their opponent (the pharmaceutical industry) and deny or minimize the conflict of interest of their proponents (Andrew Wakefield, Romain Gherardi).

A sixth example is the problem of male privilege. Feminists often accused men who are critical about feminist issues for having a conflict of interest, in particular a male privilege that they want to protect. However, if male privilege leads to a bias amongst men because they are privileged, it also leads to a bias amongst women. If men want to protect their privilege and are therefore less reliable or credible in some matters, we can as well say that women want to achieve privilege and are therefore also less reliable in those matters. Everyone can be said to have a conflict of interest: those who have power want to keep it, those who do not have power want to achieve it. It is not obvious why the latter would have a weaker conflict of interest and would be more credible.

What about climate change? It is well known that many deniers of anthropogenic climate change have financial conflicts of interest with the fossil fuel industry. But deniers sometimes claim that believers can have two kinds of conflicts of interest. First, climate scientists could have been paid by the low carbon, clean energy industries (e.g. nuclear power and renewable energy sectors), and second, climate scientists could be spreading doomsday scenarios of global warming as a means to ask for more government funding for more research, to secure their jobs. The former conflict of interest with the clean energy industry is expected to be very weak, because that industry is much smaller than the fossil fuel industry, and it becomes less and less likely because fossil fuel companies are investing more and more in nuclear power and renewable energies. The latter conflict of interest is unlikely, because it results in a huge conspiracy theory where all climate scientists have to mislead the governments. That requires an impossible coordination among scientists and a huge effort to keep the truth secret. A few decades ago, climate scientists warned about global cooling and a new ice age. That would have been a better story for securing more future government funds for research, because in that story, humans are not responsible for the climate catastrophes, which means that this story would not have met opposition from huge industries like the fossil fuel industry.

Suppose both believers and disbelievers, proponents and opponents, have conflicts of interest. What should we do then? We are no longer able to use the easy strategy of looking for conflicts of interest and discrediting all studies with such conflicts of interest. Luckily, another easy strategy is to see if there is a scientific consensus. And we have to rely on the more difficult strategy of looking at the content of the scientific studies instead of the backgrounds of the authors.

The most important lesson that we can learn from the conflict of interest bias, is that we have to be fair in our judgments and acknowledge that people (scientists) who hold our own views can also have conflicts of interest. That means we should tolerate some level of conflict of interest. For example, it is important that environmental organizations have the most reliable scientific knowledge, and therefore those organizations should invite scientists to give advice or to speak at environmentalist conferences. In order to attract enough top scientists, it might be an effective, necessary and therefore good idea to pay those scientists, to give them consultancy and speaking fees. Does that mean that those top scientists are no longer allowed to sit in governmental scientific panels or advisory boards, due to their financial conflicts of interest with the environmental organizations? Of course, those organizations would not complain against the panel memberships of those scientists. If those top scientists are not allowed in the panel, the government risks ending up with a small panel with only a few scientists with lower levels of expertise. Now suppose those scientists had similar conflicts of interest with the industry (e.g. giving paid presentations at conferences sponsored by the industry). Now the environmental organizations object. This is unfair. Furthermore, it is also important that the industry can rely on good advice from scientists: the scientists have the knowledge, the companies have the capital, and both are necessary to produce good products. So we should be more tolerant or nuanced towards some conflicts of interest.

Scientists are not only susceptible to financial conflicts of interest, but also to all kinds of cognitive biases. How reliable is a scientist who warns against a synthetic chemical product, if that scientist has a naturalness bias and is a member of an environmentalist organization? How reliable is a scientist who favors a new therapy against an untreatable disease if that scientist has a family member with that disease?

Luckily, the scientific method (with peer reviewed research, statistical methods to detect biases, other scientists testing and retesting hypotheses,…) is the best strategy we have to avoid those biases of individual scientists. Instead of focusing on the biases of an individual scientist, we should look at the broader scientific picture, the validity of studies, the statistics, the meta-analyses, the scientific consensus views, the positions of scientific academies…