Spin in the abstracts of randomized controlled trials in periodontology and oral implantology: A cross-sectional analysis.

Abstract

To investigate the presence and characteristics of spin (a distorted interpretation to make research findings seem favorable) in abstracts of randomized controlled trials (RCTs) in periodontology and oral implantology, and to explore its associated factors and influence on the subsequent literature. PubMed was searched to identify recent RCTs in periodontology and oral implantology, whose primary outcome was non-significant. Spin in abstracts was assessed and categorized according to pre-determined spin strategies. The associations between study characteristics and the presence / severity of spin were analyzed using multivariable logistic regressions. 196 abstracts were included, 137 (69.9%) of which had spin. 57 (29.1%) abstracts had spin in the Results Section, 126 (64.3%) had spin in the Conclusion Section. The main spin strategies in the Results and Conclusion Sections were focusing on secondary outcomes (16.3%) and focusing on within-group comparisons (28.6%), respectively. The presence of spin was associated with number of centers (OR=0.28, 95% CI: 0.11-0.73; p=0.009) while its severity was associated with topic (OR=0.24, 95% CI: 0.08-0.70; p=0.009). The frequency of spin is relatively high among published RCT abstracts in periodontology and oral implantology. Findings reported in these abstracts need to be interpreted with caution.