标题：Dependability of results in conference abstracts of randomized controlled trials in ophthalmology and author financial conflicts of interest as a factor associated with full publication
摘要：Background Discrepancies between information in conference s and full publications describing the same randomized controlled trial have been reported. The association between author conflicts of interest and the publication of randomized controlled trials is unclear. The objective of this study was to use randomized controlled trials in ophthalmology to evaluate (1) the agreement in the reported main outcome results by comparing s and corresponding publications and (2) the association between the author conflicts of interest and publication of the results presented in the s. Methods We considered s describing results of randomized controlled trials presented at the 2001–2004 Association for Research in Vision and Ophthalmology conferences as eligible for our study. Through electronic searching and by emailing authors, we identified the earliest publication (journal article) containing results of each ’s main outcome through November 2013. We categorized the discordance between the main outcome results in the and its paired publication as qualitative (a difference in the direction of the estimated effect) or as quantitative. We used the Association for Research in Vision and Ophthalmology categories for conflicts of interest: financial interest, employee of business with interest, consultant to business with interest, inventor/developer with patent, and receiving ≥ 1 gift from industry in the past year. We calculated the relative risks (RRs) of publication associated with the categories of conflicts of interest for s with results that were statistically significant, not statistically significant, or not reported. Results We included 513 s, 230 (44.8 %) of which reached publication. Among the 86 pairs with the same main outcome domain at the same time point, 47 pairs (54.7 %) had discordant results: qualitative discordance in 7 pairs and quantitative discordance in 40 pairs. Quantitative discordance was indicated as < 10, 10–20, > 20 %, and unclear in 14, 5, 14, and 7 pairs, respectively. First authors reporting of one or more conflicts of interest was associated with a greater likelihood of publication (RR = 1.31; 95 % CI = 1.04 to 1.64) and a shorter time-to-publication (log-rank p = 0.026). First author conflicts of interests that were associated with publication were financial support (RR = 1.50; 95 % CI = 1.19 to 1.90) and one or more gifts (RR = 1.42; 95 % CI = 1.05 to 1.92). The association between conflicts of interest and publication remained, irrespective of the statistical significance of the results. Conclusions More than half the /publication pairs exhibited some amount of discordance in the main outcome results, calling into question the dependability of conference s. Regardless of the main outcome results, the conflicts of interests of the ’s first author were associated with publication.
