Physicians’ perceptions of educational articles were not influenced by financial conflict of interest (COI) statements, according to a study published in BMJ Open.
This randomized, controlled trial was designed to investigate how COI statements affect perceptions of educational articles in a random sample of physicians in the United Kingdom (N=749). Four permutations of 2 clinical reviews (1 on dyspepsia and 1 on gout) were created; the articles varied only in in terms of COI statement.
Blinded volunteers were randomly assigned to receive 1 review, then complete a questionnaire after reading. Blinded analysis of covariance and factorial analysis of variance were used to assess each review’s influence and the type of COI on readers’ confidence in the conclusion (primary outcome). Level of interest, the article’s importance, and the likelihood of changing their practice after reading were secondary outcomes.
Continue Reading
After adjusting for job type, sex, age, and years since qualification, no significant difference was seen between groups in readers’ confidence in the article (gout: P =.32; dyspepsia: P =.78) or perceptions of its importance (gout: P =.09, dyspepsia: P =.79). For readers of the gout review, articles with consultancy and advisory board COIs were rated as significantly less interesting than articles with no COI (P =.028 with Bonferroni correction), but no significant difference was observed for the dyspepsia review (P =.83). Among participants who treated the condition and read an article with recommendations that differed from their own practice, no significant difference was observed between groups in their likelihood of changing their practice (gout: P =.59, n=59; dyspepsia: P =.56, n=80).
Limitations to this study included an initial low response rate, use of only National Health Service doctors, self-reported outcome measures, an inability to pool results from the two sampling approaches, the fact that participants were told they were part of a research project and were given an article that they might not usually read, and older respondents.
Study investigators conclude that overall, COIs have little impact on clinical readers’ perceptions of educational articles, but more research is needed to determine why. “It is possible that the simple presence of a COI is not sufficient to attract attention,” the investigators suggested. “Rather, in addition to reporting COIs, authors or journal editors should consider positioning the COI in relation to the topic of the article so that any context-specific risk of bias is clearer to the reader.”
All authors have a current or former employment connection with The BMJ.
Reference
Schroter S, Pakpoor J, Morris J, Chew M, Godlee F. Effect of different financial competing interest statements on readers’ perceptions of clinical educational articles: a randomised controlled trial. BMJ Open. 2019; 9(2):e025029.
