Abstracts presented at the European Association for Osseointegration (EAO) Congresses- Publication fate and discrepancies with full-length articles

15749 ORAL COMMUNICATION CLINICAL INNOVATIONS

Background

The European Association for Osseointegration (EAO) Annual Scientific Congresses present the most up-to-date research in oral implantology, and the increase in the number of abstracts presented at EAO congresses is an indicator of the rapid development of oral implantology research. However, any failure or delay in publishing an abstract in full is unscientific and unethical, which can lead to reporting and publication biases and amount to a waste of time and resources.

Aim/Hypothesis

To investigate the full publication proportion (FPP) of abstracts presented at the 2010 and 2011 EAO Congresses, to analyse the discrepancies between abstracts and their full publications, and to explore potential predictors of FPP and discrepancies.

Material and Methods

Abstracts presented at the 2010 and 2011 EAO Congresses were retrieved from the online archive of Clinical Oral Implants Research. Associated full publications were identified by searching PubMed, Embase and Google Scholar. Discrepancies between abstracts and full publications were identified, categorized and then quantified using a discrepancy score (DS, range- 0 to 20). The Kaplan-Meier survival analysis was used to describe cumulative FPP over time. Association between abstract characteristics and cumulative FPP were analysed using cox regression modelling. Association between abstract characteristics and the DS were analysed with a linear regression model.

Results

A total of 850 abstracts were included. The overall FPP was 36.4% with a median time lapse of 12 months. Higher FPP were significantly associated with oral presentation (RR = 2.13; 95% CI- 1.59 to 2.86; P < 0.001), multiple affiliations (RR = 1.41; 95% CI- 1.09 to 1.81; P = 0.008) and presence of statistical tests (RR = 1.78; 95% CI- 1.39 to 2.26; P < 0.001). Among the identified full articles, 91.3% had at least one minor change from the abstract and 70.9% had at least one major change. Higher DS was significantly associated with longer time lapse (B = 0.06; 95% CI- 0.04 to 0.08; P < 0.001) and being clinical research (B = 1.30; 95% CI- 0.52 to 2.08; P = 0.001).

Conclusion and clinical implications

Only about one-third of abstracts presented at the EAO Congresses were published in full. Among these, more than two-thirds had at least one major change in their full publications. Conference abstracts in oral implantology should be viewed with caution.