@article{553525d8dba14b41811c435abbe35d1a,
title = "Withholding conflicts of interest: the many flaws of the new ICMJE disclosure form",
abstract = "In this article, I describe and analyse the proposed new International Committee of Medical Journal Editors form for disclosing conflicts of interest and conclude that it has many flaws. The form does not mention 'conflicts of interest' even once in either its body or its title, it introduces a conceptually confused categorisation of different potential conflicts and it ignores future conflicts and intellectual biases. Finally, many of the authors of the new form have themselves failed to declare relevant potential conflicts of interest.",
keywords = "research ethics, publication ethics, JOURNALS",
author = "D. Shaw",
note = "Funding Information: In terms of the directly related category, authors must report {\textquoteleft}all those financial and non-financial relationships and activities that supported the work and, if financial, were paid to you or your institution{\textquoteright} and {\textquoteleft}report any [such] relationships and activities, paid or unpaid, present from the time of initial conception and planning of the work to the present{\textquoteright}. This is relatively straightforward, but it does mean that authors could have received funding from a company that ended the week before they conceived a new project, and would not have to declare that funding, despite it possibly biasing their work. (It may not always be easy to identify precisely when a topic was conceived in any case.) Another issue with the phrasing on the new form is that it concerns only {\textquoteleft}relationships….that supported the work{\textquoteright}. But what does it actually mean for a non-financial relationship to support a work? If I am married to the director of a pharma company that is interested in the results of my work but is not funding it, that is a non-financial relationship, but it does not support my work. Nonetheless, one would hope that I would declare it as it is highly relevant. (This type of relationship is also not covered by topically related relationships—see below.) In contrast, topically related activities and relationships {\textquoteleft}include all other financial and non-financial relationships and activities within the broadly defined field addressed in the work{\textquoteright}. This sounds very broad, but is in fact too limited, for a number of reasons. First, let us imagine an author who has been doing research on diabetes. He discloses that his research was supported by a large grant from a pharma company, and that he has other funding on diabetes unrelated to the work (and thus topically related). However, he does not mention (and does not need to, according to the form) that he also received money for transplantation research from the same company. Receiving the extra money from that company makes the potential for conflict of interest much greater, but the form does not require disclosure as the grant was not directly related, or topically related. Furthermore, if the transplantation grant came from a different company, that might be relevant if that company also happened to have a vested interest in the diabetes research, despite the funding not being topically related. This highlights the flaw in dividing conflicts in this way: an author could be receiving money from or be employed by a company, but would only have to declare that funding if it funded the research being submitted, or is on a broadly similar topic. As already stated, the potential for bias is much greater if one has several grants from a company rather than just one, but the new form would not capture this information. The aforementioned fact that my wife is the boss of a pharma company would also not need to be declared here according to the form{\textquoteright}s guidance, because that is not a relationship within the field addressed in the work. Publisher Copyright: {\textcopyright} 2021 BMJ Publishing Group. All rights reserved.",
year = "2022",
month = jan,
day = "1",
doi = "10.1136/medethics-2020-106136",
language = "English",
volume = "48",
pages = "19--21",
journal = "Journal of Medical Ethics",
issn = "0306-6800",
publisher = "BMJ Publishing Group",
number = "1",
}