Whether proven, alleged or merely suspected, conflicts of interest have a corrosive effect on a science-based profession.
Since becoming your editor, I have enjoyed hearing your ideas about the content of The Journal, its presentation and how it can influence the practice of dentistry. Somewhat to my surprise, a significant fraction of the calls and letters every month seem to revolve about a matter that lies at the very foundation of what we — JADA in particular and scientific publications in general—are all about. At issue are conflicts of interest, and the extent to which they taint the integrity of scientific work. The most commonly voiced concern is that industrial sponsorship is utterly inconsistent with honest science.
Whether proven, alleged or merely suspected, conflicts of interest have a corrosive effect on a science-based profession. Was a study designed to fairly test a hypothesis, or to buttress a preconceived idea? Are the reported results truly representative? Were study patients appropriately selected and ethically treated? The scientific literature provides the rational basis for selecting products, treatments and concepts. If we lose confidence in the published record, we also lose much of our independence of thought.
I think it’s time to address this important issue head-on, rather than indulge in vague and unanswerable innuendo. Conflict of interest is an ever-present threat, true, but it is not unmanageable.
Science is not conducted by saintly practitioners in some pure and cerebral realm. Authors are mere human beings, subject to the standard-issue human foibles. We editors are ink-stained wretches forced to pass judgment on our intellectual betters. And yet, we have to make the system work to discover and convey new knowledge you can rely on.
I will not deny that science presents opportunities and temptations to take shortcuts, or worse. But there are powerful mechanisms at work to prevent errors (intentional or otherwise) from entering the canon. Let’s consider how the system works to protect you.
There are at least seven characteristics of the scientific enterprise, which act as a check on unethical behavior. Not all apply to every research project, and it is important that we understand their applicability to a particular case. Nevertheless, these standards constitute a "layered defense" against conflicts of interest in the scientific literature.
Personal integrity.
I believe that most people, given a choice, want to take pride in their work; this includes holding up their heads as honest, unbiased individuals. If you are terminally cynical, you can attribute integrity to mere self-interest. Whatever the motivation, the disincentives to publishing bad science are strong, and the sanctions are severe. Outright fraud can land you in jail and will almost certainly end your career. Poor-quality work, especially with taint of bias, will damage your reputation among peers; nobody wants to appear foolish in public.
Right to publish.
When practitioners, universities or other third parties perform research, it is common for the investigator to be contractually guaranteed the right to publish the results. This means that negative findings can’t be unilaterally suppressed. Obviously, this is a two-edged sword for the sponsor, who must balance the risk of exposing disappointing results against the added credibility that attaches to an independent study.
Regulations.
In the United States and other developed countries, almost anyone conducting research (including privately sponsored health-related research) must comply with a panoply of government regulations. Many of these involve financial controls, patient safety and animal welfare. But an important subset of regulations are designed specifically to ensure that financial conflicts do not bias study results or negatively affect patient care. Every research institution that receives federal support must have a conflict of interest review board charged with the task of identifying and managing conflicts of interest (for example, equity holdings and contingent payments).
The U.S. Food and Drug Administration.
This layer only applies in the United States (though comparable oversight exists elsewhere) and only to research on drugs and devices that come under FDA jurisdiction. But when it applies, FDA scrutiny is a strong layer of protection. Knowing that it will be subject to rigorous review, companies try to do their research as thoroughly as possible. Do it wrong and the FDA will find out, and back you go to square one, losing years of valuable time.
Peer review.
Any reputable journal sends out papers for review by peers, who are knowledgeable both in the scientific field and in the process of scientific publishing. These people are, on the whole, very careful and subject the paper to meticulous review. They ask for clarification, justification and explanation. Reviewers help the author convey worthwhile research clearly; just as importantly, they sometimes let the editor know when the research is unsuitable for publication. As authors, we may complain about reviewers’ nitpicking, but we know that they are an indispensable part of the quality assurance process.
Intelligent readers.
You don’t have to swallow papers whole. Read critically and carefully. Consider the source. Know the difference between statistical significance (for which there are clear-cut rules) and clinical significance (which is for you to decide). When listening to a presentation, feel free to ask about sources of support. There should be nothing to hide.
The scientific method.
Ultimately, science is self-correcting. Publication of methods and results invites others to repeat the experiment. What can’t be confirmed is consigned to the dustbin of history.
Overall, I think we do a good job of separating the scientific wheat from the chaff. Think back on the scientific scandals in recent years and then consider how few in number and how quickly exposed they have been. As George Will recently wrote in reference to a questionable work of history, such "malfeasance ... actually reveals something heartening—a considerable strength in America’s scholarly community. Its critical apparatus is working. Scholars and their journals are doing their duty, which is to hold works of scholarship up to the bright light of high standards."1
We at JADA do not want or expect blind faith in what we publish. We fully expect to earn your trust by seeking out high-quality work and subjecting it to rigorous review. Continue to do your part by reading critically, offering reasoned comment and dissent, and serving as peer reviewers when asked. Trust is not to be lightly bestowed, nor lightly withdrawn.
