Reporting, interpretation, and extrapolation issues (SPIN) in abstracts of orthodontic meta-analyses published from 2000 to 2020.

Abstract

To assess the prevalence of and identify factors associated with SPIN in abstracts of orthodontic meta-analyses. Electronic search was performed within the contents of five orthodontic journals and the Cochrane Database of Systematic Reviews (CDSR) to identify meta-analyses of studies involving humans, from 1 January 2000 until 31 August 2020. Inclusion of SPIN in the abstract of meta-analyses, defined as misleading reporting, misleading interpretation, and inappropriate extrapolation of the findings, was documented. Extent of SPIN and associations with journal and year of publication, type of study, number of authors, continent of authorship, methodologist involvement, funding, and significance of the primary outcome were investigated. One hundred and nine meta-analyses were identified, with the highest proportion being published in the European Journal of Orthodontics (EJO: 31/109; 28.4%). Inclusion of SPIN, in at least one domain, was recorded in nearly half (53/109; 48.6%) of the studies, of which 30 (56.6%) included 2 or more domains of SPIN. Meta-analyses of observational studies presented 1.66 times higher risk for including SPIN in their abstracts compared with interventional ones [95% confidence intervals (CIs): 1.14, 2.40; P = 0.007], after adjusting for a number of predictors. Studies with a large number of authors (≥6) presented 1.76 times higher risk of SPIN (≥6 versus 1-3: 95% CIs: 1.04, 2.97; Wald test, P = 0.021), conditional on the pre-defined predictors. Flaws in the reporting and interpretation of the findings of abstracts of meta-analyses, as framed by inclusion of SPIN are persistent in orthodontic research, being more prevalent in meta-analyses of observational studies. Consistent, multidirectional efforts should be endorsed to improve the quality of the disseminated research findings.