Journals use conflict of interest field in PubMed inconsistently

Medical and life science journals are making increased use of a voluntary field in the PubMed database designed to make researchers' conflicts of interest visible on studies' abstract pages, according to new research published today in the journal PLOS ONE.

The National Library of Medicine, which administers PubMed, added the new voluntary field to capture researchers' conflict-of-interest disclosures in 2017, after the Center for Science in the Public Interest and others urged it to do so in 2016. Including conflicts of interest on the abstract page of PubMed is important since many readers might not have full access to the underlying article or may rely on the abstract before deciding whether to read the full text.

Researchers at CSPI and Yale University decided to examine the prevalence of the posted conflicts of interest over time.

Among approximately 7,000 journals publishing articles that appear on PubMed each year, the proportion with at least one article with a posted conflict-of-interest statement increased from 25.9 percent in 2016 to 33.2 percent in 2021.

Among nearly 400,000 articles published each year, the proportion that included a posted conflict also increased from 9 percent in 2016 to 43 percent in 2021. The study identified 5039 journals that did not use the conflict-of-interest field at all in 2021.

The researchers also closely examined up to 100 randomly selected articles from each of the 40 highest impact journals, as defined by Journal Citation Reports. They determined how many published conflict-of-interest information anywhere in their articles, and then what percentage of those used the conflict-of-interest field in PubMed. Among 3,888 articles published in these journals in 2021-2022, 30.2 percent had published conflict of interest disclosures. Of those, 63.3 percent used the PubMed field. Even high-impact journals, such as Nature Materials and Science, make inconsistent use of the field and variously use the Acknowledgments or Funding sections in published versions, without transferring the information to the PubMed conflict-of-interest field.

CSPI is writing 46 of the journals that failed to fully utilize the PubMed field urging them to do so. Some of these journals are not using the field at all, others are doing so inconsistently, and others are not transferring conflict data from Acknowledgments or Funding sections of their published articles to the PubMed field.

"Researchers, journalists, clinicians, and other users of the PubMed database should be able to know at a glance if the studies before them are funded by the drug, food, chemical, or other industry, or the government or foundations," said CSPI president Dr. Peter G. Lurie, a co-author of the paper. "Unfortunately, while journals have made some progress using this valuable new field, others are making inconsistent use of the field or aren't using it at all-and that needs to change. Journals owe their readers no less."

Lurie's co-authors on the PLOS ONE study were former CSPI scientific integrity campaign manager Stephanie Rogus, now on faculty at Texas A&M University, and Dr. Joseph S. Ross, professor of medicine and public health at Yale University.

"We hope that notifying editors of journals that are using the field inconsistently or not at all will encourage them to take the necessary steps to ensure that all author conflicts are included in PubMed's conflicts of interest field," Rogus said.

"Utilizing this field is important because all readers may not have access to the full text of the article and many rely on the abstract alone to draw conclusions about a study or to guide clinical decisions," Lurie wrote to the journal editors. "Disclosing conflicts of interest allows scientists, professionals, and the public to better assess the credibility of scientific findings and consider the potential impact on funding on research."

# # #

Topics

• CSPI

Contact Info: Lisa Flores, 202-777-8368 or Jeff Cronin, 202-777-8370