Most Journal Articles Still Missing COI Statement on PubMed

While use of PubMed’s voluntary field for conflict of interest (COI) statements in abstracts has increased, it continues to be underutilized by most scientific journals, according to a repeated cross-sectional study.

An analysis of more than 7,000 journals on PubMed found that the proportion of journals with at least one article that included a posted COI statement rose from 25.9% in 2016 to 33.2% in 2021, according to Peter Lurie, MD, MPH, of the Center for Science in the Public Interest in Washington, D.C., and co-authors.

Among the approximately 400,000 articles published each year, the proportion with a COI statement on PubMed increased from 9% in 2016 to 43% in 2021, they reported in PLoS One.

Notably, in a separate analysis of the 40 highest-impact journals, 30.2% of approximately 4,000 articles published in 2021-2022 had published COIs; 63.3% of those had also posted that statement in the voluntary PubMed COI field.

“The most fundamental problem is that the majority of journals, even today, are not using a field that PubMed provided 7 years ago,” Lurie told MedPage Today, adding that this is still the case even though “it’s a relatively straightforward matter to link your journal to the PubMed database, because about a third of journals have managed to.”

Lurie explained that the National Library of Medicine added this voluntary field for COI statements on the abstract page of PubMed in 2017, but the extent to which it was being used was unknown. He explained that these analyses show that the field for posting COI statements has continued to be underused, even by some of the highest impact journals.

He noted that the ideal outcome would be nearly universal inclusion of all published COI statements in the PubMed field, but the reasons why that does not happen can vary.

“There are a number of possible explanations,” Lurie said. “One is [they] don’t use the field at all.”

“The second explanation is they do sometimes transfer the data, but for some reason it’s not done consistently,” he added. “The third possibility is that the conflict exists in a field in the published version, but that field is not being transferred over to PubMed.”

He noted that mismatching fields between the published article and PubMed is likely the most difficult problem for journals to solve. This could mean that journals have funding or acknowledgement sections in their articles where conflicts are mentioned, but without an official COI statement section that information does not make it into the PubMed field.

“What that basically says is that there’s a lack of consistency in the ways in which the journals collect and report information in this entire area,” Lurie said. “Anybody who submitted an article to a journal knows that, although some of them follow standardized approaches, others do not and are just idiosyncratic.”

Lurie noted there is already existing code to help journals that want to improve their reporting of COI statements on PubMed.

The authors noted that the National Library of Medicine should consider further outreach to journals that are not using the COI statement field in order increase transparency on PubMed.

For these analyses, the authors identified journals indexed on PubMed from January 2016 to December 2021, that published original research, reviews, and comments.

The articles consisted of clinical trials, observational studies, validation studies, systematic reviews, meta-analyses, narrative reviews, letters, editorials, consensus statements, case reports, and guidelines. They excluded books, legislation and government publications, article preprints, news articles, and datasets.

For each year, the authors recorded the total number of journals indexed and the number of journals with at least one article that included a COI statement on PubMed.

For the analysis of high impact journals, they identified the top 40 — based on the Journal Citation Reports — that had published at least 50 articles from June 2021 through May 2022. The authors randomly selected 100 articles from each and manually searched articles for published COI statements and compared those to the PubMed field for each article.

The authors explained that they started their search in 2016 to show change from before and after PubMed added the new field.

The study had several limitations. For example, the authors were only able to assess the utilization of PubMed’s COI statement field among the top 40 journals in one year due to the large number of articles published. They also may have missed some posted COI statements that were added after their data collection ended in 2022.