StATS: Reforming conflict of interest statements (November 12, 2007).

In a recent talk, I was asked to disclose any conflicts of interest. I did so, but I also commented that the organization I was speaking to, as well as most other organizations have a poorly thought out conflict of interest policy. The problem with most of these policies is that they make it hard to disclose conflicts and easy to pretend that no conflicts exist. I believe that rather than having an easy box to check, make the statement of no conflict a more detailed one.

Conflicts are only of potential interest when you are presenting information about a commercial product, or you are advocating a change in regulations that might influence a company's bottom line. If your talk is about how everyone should present their data using confidence intervals, it doesn't matter how much money you have gotten from any source.

If you mention a commercial product, then the only conflicts that matter are financial ties associated with a company that produces that commercial product or with a company that produces a competing product. I received a nice gift from someone I helped with a consulting assignment. It was a CD of "Opera Digest." Now this is potentially reportable if I was reviewing their product or if I was advocating that Wagner play in the background of all operating rooms. But for any reasonable article that I might write, the gift of "Opera Digest" should not be reported.

So asking for details about any financial support from any drug company is rather silly. You only need to know about companies that produce or compete with products that you are discussing in your talk.

Advocacy is a bit trickier. Just about every paper advocates something. Advocacy only becomes a concern when you are advocating changes in the policies of a regulatory body, such as the U.S. Food and Drug Administration. Some changes (the FDA should hire more Epidemiologists) do not have any impact on any commercial concern, but others do. If you are advocating that the FDA should allow drug companies to present meta-analytic data in place of randomized trials, this change would greatly lower the approval costs of a new drug and improve the financial bottom line of many drug companies. This may still be a perfectly valid recommendation, but when you make it, be sure to disclose any financial interests associated with companies that might benefit from your proposed reform.

So instead of a no conflict of interest box, include the following questions.

  1. What commercial products are mentioned in your talk? (if none, skip to 4).
  2. For each product, detail any financial ties to the company that makes this product.
  3. For each product, detail any financial ties to companies that make a competing product.
  4. Are you advocating any changes in policies or regulations in your talk? (if none, skip to 7).
  5. Detail any financial ties to companies that might benefit from your proposed change.
  6. Detail any financial ties to competitors of companies that might be harmed by your proposed change.
  7. Are there any other relevant financial ties that your audience would want to know about?

The phrase "financial ties" needs a careful definition, of course. You must mention employment by the company. You must also mention any consulting arrangements and private research grants from the company. You should mention any stock ownership, excluding ownership via a mutual fund where you have no direct control over the investment choices. Finally, mention any gifts you have received from a company of more than nominal value. Attending a training class sponsored by a company does not count as a conflict, unless the company defrays part or all of your travel or provides lavish materials or food (the rubber chicken that most hotels provide during catered lunch doesn't count). Any financial tie involving a direct family member (spouse, sibling, or child) deserves the same disclosure as a financial tie to you directly.

Some people, for example, have stated that being a recipient of a government grant produces a conflict of interest, because it encourages people to skew results towards those that the governmental agency wants to see. It's an argument that on its face looks good. I believe, however, that this is a red herring that is raised by people who want to distract attention from the fact that they have very real corporate conflicts that they want to minimize. If you expand the sphere of people who have to declare a conflict, then your conflict doesn't stand out quite as much. Furthermore, although there are many empirical examples of how corporate financial ties can produce skewed research results, I am unaware of any research that shows a similar problem with government funded research. That doesn't mean that there isn't a problem, just that the critics have not made a convincing case yet.

There has been a lot written about non-financial conflicts of interest in general and intellectual conflicts of interest in particular. I'm skeptical about these as well, but the whole topic needs more discussion.

This page was written by Steve Simon while working at Children's Mercy Hospital. Although I do not hold the copyright for this material, I am reproducing it here as a service, as it is no longer available on the Children's Mercy Hospital website. Need more information? I have a page with general help resources. You can also browse for pages similar to this one at Category: Conflict of interest.