Conflicts of interest and the truth of scientific discovery: an editor's perspective

When Christopher Columbus was struck by his idée fixe to find a direct route to India, he needed a sponsor for his journey. João II, King of Portugal, was unsure if this was possible. Columbus failed to convince the ruler, too dangerous and uncertain seemed the plan to sail into the Atlantic into nowhere land. But Columbus did not give up, and subsequently went to Spain, where Isabella I of Castile and Ferdinand II of Aragon, the catholic monarchs, were ruling the country with iron fists. They had just united several kingdoms save for Granada (which was the one city that successfully resisted the Reconquista) in their campaign to expulse all Jews. In January 1492, Isabel finally did conquer the last stronghold of Islamic rule on the Iberian Peninsula. Columbus visited her in Cordoba and tried to gain her support for his journey.1,2 Isabella told him that the Junta of Salamanca had concluded that the Atlantic was not crossable. Columbus boldly replied: ‘What did they say about Granada before today?’ whereupon the queen had to acknowledge: ‘That it was impregnable.’3 Columbus had won. Columbus was far from modest in his triumph: he not only asked for the title of ‘Don’, but also was given the rank of Admiral of the Spanish Fleet. In addition, he was appointed Viceroy and Governor of all new lands discovered, and for one-tenth of the revenues from his discoveries. From the very beginning, there were huge conflicts of interest marring his journey. The rest is history: Columbus discovered the new world, landed in what he called Hispaniola (now Haiti), and eventually established a route to North and South America. Was his seminal discovery devalued by his financial conflicts? Without a doubt he had asked for compensation in gold and honours, and many of his actions were incredibly selfish, but he discovered the Americas. When Andreas R. Grüntzig dared to enter a coronary artery with a self-made balloon catheter in a patient with angina pectoris and a stenosis of the left anterior descending coronary artery on 16 September 1977, most of his colleagues at the University Hospital in Zurich felt that he was out of his mind.4 As he constructed the first balloons on his famous kitchen table at home, he needed a commercial partner and filed a patent. Originally a small garage business, Schneider Medintag AG, would become an impressive international business in Bülach (near Zurich), and eventually a worldwide enterprise. At the time, Grüntzig was promoted to senior consultant, but did not receive the professorship he was looking for. Later, he too landed in the USA, was made faculty member at Emory University, and became a rich man thanks to his invention. Did the money he made lessen his discovery? Certainly not—to the contrary: he changed cardiology from a diagnostic subspecialty of internal medicine to an interventional discipline on its own. His vision of catheter-based interventions has revolutionized cardiovascular medicine to this day. Lastly, an example from the pharmaceutical industry: Hartmann Stähelin and Jean Borel earned their living at Sandoz in Basel, where their employer expected blockbusters from their scientists.5 Stähelin never got rich from his discoveries and Borel could not enjoy his glory, but Sandoz (later Novartis) made a fortune from Sandimmun®, the first effective immunosuppressant that revolutionized transplantation medicine. Did the fact that Sandoz made billions with this seminal drug devalue their discovery? Certainly not; to the contrary: cyclosporin saved hundreds of thousands of lives, and the production of the drug provided jobs for numerous individuals in Basle and beyond. What do we learn from these stories? Conflicts are found everywhere in the discovery process, but they do not necessarily hinder innovation; to the contrary they may be important motivators on the bumpy road to evidence—Bernard Mandervilles'6 and Adam Smith’7invisible hand also works in this context. Obviously, as outlined by others,8 the primary interest of any scientific endeavour, i.e. to benefit patients and society, must be aligned with such secondary interests in a way that it minimizes the risk of bias. Therefore, not all conflicts are a valid concern; they should be rated along these lines accordingly. Above all, conflicts with a variety of industry partners can result in less bias than the conflict arising from having only one industry partner. What primarily matters, however, is the test of time; it is the only reliable judge of the validity of scientific findings and their value for society at large. The philosophy of science has stressed this fact ever since the publication of Karl Popper's seminal work:9 the falsification process considers only the reproducibility of findings, not the motivation of the discoverer or the financing of its project—it is the result of the scientific process, what is left from conjectures and refutations that really matters, not the circumstances of its establishment. It is undisputable that the scientific process is influenced by our favourite hypothesis (on which we often built our careers), by hopes (we are all prone to confirmation bias) and interests, be they financial, professional or personal in nature. Although the códex of science10 will guide most scientists to good publishing practice,11 there are those who deviate from this path. Of note, retractions of published papers have increased markedly in recent years, particularly in high impact journals.12 However, lies have short legs. Unfounded findings most likely will not survive the test of time and will be falsified by others eventually. What is not confirmed by subsequent research will eventually be refuted. What do we gain then with conflict of interest statements at the beginning of each lecture, each talk, and as footnotes in every publication? We do learn who financed the research, and if the lecturer or author has received honoraria from any kind of company—beyond that however we no little more. Specifically, we do not learn anything about professional and intellectual conflicts that may have influenced the authors thinking and his research as much as financial ones do. No doubt, financial conflicts may foster our enthusiasm of the findings made, we may exaggerate the implications of the research.13 But the findings must be true; otherwise, we would not talk about conflicts but about fraud. So, what do we really learn from conflict statements? Are we in a better position to judge the true value of findings? To answer this question, a randomized trial involving 503 internists was performed (and later published in the New England Journal of Medicine).14 They were asked to rate the quality and credibility of abstracts reporting the design and results of drug trials of different rigor. In a randomized fashion, they were told that the trials had been supported by the National Institute of Health, by a pharmaceutical company, or by no sponsor at all. Of note, independent of the methodological quality and scientific rigor of the presented trials, the participants downrated trials where the pharmaceutical industry was stated as being the sponsor. Thus, transparency of sponsorship creates a hitherto unknown novel negative bias—an unbiased judgment seems to be difficult to attain in the current environment. It reflects a profound loss of trust in modern society.15 In publishing—which is the main focus of this ViewPoint—conflicts may arise (i) from authors when writing the paper and when proposing reviewers; (ii) by editors when considering the paper for the review process; (iii) by reviewers when accepting and writing a review; and (iv) again by the editors, when deciding on acceptance or rejection. All submitting authors have conflicts of one kind or another, but only a few have financial conflicts—which in today’s money-driven world gets the most attention. If they do, they are asked by most journals—including the European Heart Journal (see www.eurheartj.oxfordjournals.org)—about real and perceived conflicts related to their work according to the recommendations of the International Collaboration of Medical Journal Editors (ICMJE). This transparency act has become routine, but in itself is far from enough for proper scientific publishing. More importantly, authors should become aware of all their conflicts, be they financial or intellectual in nature. It should be a categorical imperative—as proposed by Kant more than 200 years ago16—for scientists to try to overcome their influence on their work and writing. When planning and performing their work, they should consider the ‘Six Red Flags’ as recently proposed by Bagley.17 Clinical trials should be registered18 and their design published beforehand. Good scientific publishing finally requires balanced interpretation of findings, a paragraph on limitations of their study and conclusions that are fully supported by the presented data—and thus, the full right to publish data of projects supported by industry accordingly regardless of outcome should be agreed upon in writing beforehand. Above all, therefore, honesty and precision should lead their way. What about invited authors? Those who are supposed to write an editorial or review? Some editors have taken a strict approach:19 The British Medical Journal recently declared zero tolerance on education articles with financial links to industry. The editors believe that the risk of bias to be particularly important for clinical educational articles that are designed to guide patient care, when authors' biases may be less visible to general medical readers. If this were true, then the review process would be useless. Again, although less than perfect, peer review is the optimal tool we have at hand to avoid such bias. And we have no choice: if all authors with any ties to any industry are excluded from any educational articles, who will write such articles? Clinicians without any scientific experience to write about a particular subject? The New England Journal of Medicine had followed a similar policy with a previous editors for good reasons,20 but currently seems to pursue a more flexible policy. Reviewers are peers with particular knowledge in the field of the submitted work. Editors must assure that their profile in their database properly reflects their experience and competence. When invited they should specifically be asked—as the European Heart Journal does—about potential financial, but also about personal and intellectual conflicts that may inappropriately influence their assessment. If so, they should refrain from accepting the invitation or—in uncertain situations—openly declare their potential conflicts allowing editors to take these into account when deciding. The peer review process is more reliable, if three to four reviewers provide their assessment rather than one to two. Thus, in order to obtain a balanced assessment of submitted work, editors should invite an appropriate number of reviewers plus—if the work is seriously considered—a statistical reviewer. This process, although less than perfect, is the best way to provide a revised manuscript in which unbalanced and biased statements are corrected and proper statistical analysis provided. Editors alike have interest that may inappropriately influence their decisions and they must be aware of it. If there are conflicts, a consulting editor should handle the manuscript in question, as is common practice in the European Heart Journal and this fact should be made transparent for the readers in the publication. Beyond that editors must be careful not to overrate ‘hot’ topics to increase visibility and citations of their journals. Indeed, in highly competitive fields authors are often pushed to submit preliminary findings that are often later falsified.21 Typical topics of that kind were gene therapy some years and recently stem cell therapy. We have to remember that the road to evidence is a bumpy one22 and too much enthusiasm early on may mislead readers and the scientific process at large. The scientific process and publishing practice are surrounded by numerous conflicts of authors, reviewers, and editors. Conflicts may be financial, but also intellectual, personal, or institutional in nature. Conflicts are not only negative, but may be essential motivators on the bumpy road to evidence—Adam Smith’ invisible hand also works in science. Conflicts are no indicator of the truth of scientific findings, but rather the test of time. Nevertheless, authors, reviewers, and editors must be aware of their own conflicts and try to minimize their influence on their work. Proper peer review, although not perfect, is currently to best tool to minimize bias, to assure a balanced presentation of results and appropriate statistical analysis. The opinions expressed in this article are those of the author and do not necessarily reflect those of other editors of the European Heart Journal or the European Society of Cardiology. This work was not funded by any agency or any company. The author of this ViewPoint is proud to acknowledge that he has worked with almost all pharmaceutical and device companies over the last decades.