Seeding trials—no relation to Seed Media Group, LLC

ResearchBlogging.orgI’ve been having an internal debate about whether to write on this issue, not because it isn’t interesting, not because it isn’t important, but because it’s getting so much coverage and I’m not sure how much I can add to the conversation.

But it so infuriated me that I must blog. Science-based medicine relies on medical evidence. It relies on being able to grade medical evidence by its quality and strength, and to do this, there must be a certain level of transparency.

I’m only a little bit idealistic. I know that drug companies must fund clinical trials if we ever want to see new drugs. That is why we have “conflict of interest” statements included in most medical studies—so that we may view the data with a somewhat jaundiced eye, if necessary.

Until two days ago, I’d never heard of a “seeding trial”. Now I have, and I’m not happy.

Seeding trials are basically direct-to-doctor advertising. Instead of confining the trial of a new drug to one or two institutions, a study can be spread out among many investigators in the community. If a promising new medication is ready for widespread testing, doctors like me (but not me) out in the community can sign up to participate in the clinical trial. The pharmaceutical company will design the trial and a third party company will come out and make sure we are following the protocol correctly. This allows data to be collected from a diverse group of patients. It also allows a doc out in the community to participate in research, and perhaps make a little extra money.

It turns out that sometimes the research protocol comes not from the research department but directly from the marketing department.

You see, Merck isn’t run by fools. They (and apparently many other companies) figured out that these types of trials make really good advertising. They get a whole lot of doctors into the habit of prescribing a new medication not as a side-benefit of the trial but as its primary purpose.

In the latest Annals of Internal Medicine, and in-depth analysis of a seeding trial involving Vioxx (ADVANTAGE) revealed some very disturbing information. This document, obtained by the study’s authors, is one of many that shows how this study was not done in the name of testing the medication (later withdrawn for safety concerns), but as a giant marketing ploy.

The authors of the article specifically mention three problems with seeding trials that makes them “harmful to science and society”:

    Full informed consent is impossible, as the participants are not aware of the full purpose of the trial (as a marketing tool).
    Good research design is compromised for marketing goals. Spreading the study over as many practitioners as possible is good for business, but maybe not so good for research quality control.
    The study may be scientifically irrelevant. A more definitive study of Vioxx was being done simultaneously with the ADVANTAGE trial, rendering ADVANTAGE moot.

There are further problems. The existence of trials run by marketing departments creates distrust of scientific knowledge. We base our practice on data. Some data will always be problematic, but if the whole system of collecting data is called into question, clinical science-based practice can be set back decades.

While I’m not surprised that seeding studies exist, I am disgusted. Now that the practice has been brought into the light, we can be armed with another tool with which to interpret data, which is ultimately a good thing.

Kevin P. Hill, MD, MHS; Joseph S. Ross, MD, MHS; David S. Egilman, MD, MPH; and Harlan M. Krumholz, MD, SM (2008). The ADVANTAGE Seeding Trial: A Review of Internal Documents Annals of Internal Medicine, 149 (4), 251-258 (Link)