Time to publication as full reports of abstracts of randomized controlled trials in cystic fibrosis

Abstract

Objectives. To determine 1) what proportion of abstracts of randomized controlled trials (RCTs) presented at international conferences on cystic fibrosis (CF) are published as full reports, 2) time to publication, and 3) factors that might delay or prevent publication.Methods. At the end of 1995, the Cochrane CF Group's register of RCTs contained 199 abstracts describing 180 RCTs. Abstracts were identified by handsearching 44 abstract books of three international CF conferences over a 30-year period. We searched the register for subsequent full reports of these RCTs and used survival analysis to investigate time to publication. Using the log-rank test, we examined 1) whether there is a difference in time to publication between reports where the investigators concluded that the test treatment was as effective as or better than the control treatment, and reports where it was not, and 2) whether there is a difference in time to publication according to sample size.Results. Thirty-two percent of the 178 abstracts analyzed were subsequently published in full. Survival analysis indicated that the proportions published before 12 months, 2 years, and 5 years were 8.1%, 29%, and 40% respectively. No difference in time to publication was identified when the abstracts were stratified according to conclusions or sample size; no significant association (P > 0.05) existed between time to publication and both sample size and conclusions together.Conclusion. Only a small proportion of abstracts of RCTs presented at international conferences on CF are followed by full publication, and usually only after several years. Therefore, many potentially valuable studies do not reach a wide audience. However, we found no consistent factors which might delay or prevent publication.

Fate of abstracts published in the proceedings of the first annual Perinatal Society of Australia and New Zealand Congress in 1997

“…At 42 months post‐congress, it was determined that 45% of abstracts of original research presented at the PSANZ Congress in 1997 were published in peer‐reviewed biomedical journals. Studies assessing the publication rate of research presented at scientific meetings are summarized in Table 2 1,4−31 . The median publication rate for the studies in Table 2 is 47% (interquartile range 39−53%), which is a similar rate to the proportion found in the present study.…”

Analysis of characteristics of randomized clinical trials in leukemia that are associated with how results are reported

“…This may be because manuscripts were submitted first to a prestigious journal with a high impact factor, and then rejected and resubmitted to a journal with a lower impact factor. We found, as did Liebeskind et al (10), Misakian and Bero (11), Cheng et al (12), Stern and Simes (13), and Ioannides (17), that sample size did not appear to be a significant predictor of time to publication.…”

“…Other studies of pipeline bias have examined different areas: acute stroke (10), passive smoking (11), cystic fibrosis (12), studies submitted to a hospital ethics committee (13), papers published in a particular journal (14), various health‐related research projects (15), manuscripts rejected by a radiology journal (16), and HIV (17). Although our findings may be specific to childhood lymphoblastic leukemia, any similarities to the other studies may suggest some generic features of bias in the publication of health research.…”

“…In retrospect the latter may have been preferable. Different definitions are used in another two studies (12, 15). The only ‘two dimensional’ results variable we used is ‘clinical significance.’ This combines the direction of the result with our opinion of the ‘strength of the impression given by the article,’ regardless of any p ‐values reported.…”

Publication rates of presentations given at annual conferences of The Wildlife Society, 1994–2006

“…Smith et al () noted a median publication time of 27.8 months from presented abstracts at an annual American Urological Association meeting and found that abstracts were published sooner if the studies originated in the United States and included statistical hypothesis testing. Cheng et al () conducted a similar survival analysis for time‐to‐publication from 3 international cystic fibrosis conferences and they determined that 29% of the abstracts were published within 24 months, 40% within 60 months, and no significant factor affected time‐to‐publication, such as multiple authors or positive results. We also observed a shorter time‐to‐publication interval for manuscripts published after presentation in subsequent years of our analysis.…”