He aquí un corto y valiente artículo que trata el tema de los sesgos en las publicaciones científicas sobre fármacos. El autor es Michael Cote, editor de International Journal of Climate Change. Sus reflexiones sobre su práctica como editor son de sentido común, pero quizás no en todas las ocasiones reina con el brillo debido. En mi experiencia como editor he seguido las mismas reglas que Michael Cote y jamás los resultados han sido un criterio.
In this TED, Ben Goldacre describes the disastrous problem of “publication bias” in drug-science journals.Publication bias is when science journals prefer to publish more papers with “positive” results over negative results or failures. I’m baffled.
For example (and this is incredible to me), Goldacre is a medical doctor who prescribed an anti-depressant based on a positive paper that showed the drug worked. He did some research, and found that 7 trials were conducted, and that 6 papers were rejected because they had negative results. Only one paper was published, the one with the positive results.
This is extremely dangerous. He goes on to describe how patients have died and put in further harm from this type of practice.
I’m interested in this because I edit the International Journal of Climate Change. Our process is triple-blind - the authors do not know who reviewed their papers, the reviewers do not know the names of the authors, and the reviewers themselves do not know nor communicate with each other. It’s a good process. We also have clear guidance and access to support team for any questions and clarification.
I do reject a lot of papers. I don’t think I can disclose the rejection %, but it is by-far a majority. I’ve never rejected a paper because the findings were negative.
Indeed, I cannot conceive of when or how I could do that. I reject based on common problems - irrelevant study for a climate journal, the study didn’t go far enough, graduate-level work masking for scholarship, formulas were misapplied, data was misconstrued (or exaggerated), etc. I search for plagiarism, as well. Often I’ll note that papers need a complete rewrite based on grammar, poor organization, weak conclusions - but if the content is solid, then a good rewrite will make the submission more palatable.
We’re given clear instructions, guidance, and criteria matrices to help us make our recommendations - but I assure you, negative or positive results are not remotely one of those review criteria.
So I’m a bit taken aback at this talk. Is “publication bias” a problem only with drug trials? How is this bias even possible? What are the instructions and review criteria of the journals Goldacre refers to??