How some research is never published.

One issue that you will hear us gripe about here (and try to find a solution for) is the lack of mediums to discuss or publish “non-data”. What i mean by non-data is anything that can’t be published but might be useful. In that list I include hypotheses that were not correct, or data that is unpublishable or unreproducible.

You ask why non-data would ever be published eh ? Well while positive results are great and lead to new avenues, all the negative results help others not waste time doing useless time consuming experiments. Another reason why non-data is important is that your non-data may actually support somebody else’s research (some examples I will get to in another post). The problem is how to organize it, and how to constitute exactly what a piece of non-data entails.

Anyway. on to the meat of this “depressing” story. Let’s talk about Antidepressants

This NYT article describes how some makers of antidepressants didn’t disclose about 1/3 of their drug related studies. This is nothing new for most of us in the scientific field. Many times when we receive funds from biotech firms or pharmaceutical firms to do research there are strings attached. That’s not to say the strings are all bad. The funding is necessary to bring some research into industry but one of the most prevalent string is that the corporation funding a research project usually has the final say on what data can be published, and when. They pay the research, they own the rights, and are accountable to shareholders.

Most of the time that is alright, but for drug trials that is another story. When patients are involved it is important to weigh the good aspects of a drug to the bad aspects. Well what if the FDA never sees the really bad data? Can they still make an informed decision on the safety of a drug? I’ll let the public think about that one. To be fair it’s not the FDA’s problem if information doesn’t come their way. There needs to be less ambiguous congressional guidelines set to monitor drug studies either earlier on in the process, or to require full disclosure.

David Vitrant