An analysis of retractions of dental publications.

Abstract

To comprehensively report on the characteristics of retracted publications in the field of dentistry. We searched MEDLINE (via PubMed), PubMed Central, Web of Science and Google Scholar databases for dental retracted articles from database inception to 02 July 2018. In addition, we scanned the search engine, Google Scholar, and the website, Retraction Watch (www.retractionwatch.com), for retracted dental studies. Two researchers independently screened titles, abstracts and full text of search results. Descriptive data was collected on each retracted article including reason for retraction, study type, journal impact factor, and time between publication and retraction. Regression models were used to evaluate the association between journal impact factor and retraction characteristics. A total of 138 dental retracted studies were included. Reasons for retraction were misconduct (N = 100, 72.5%), with the most frequently reported misconduct being overlap/plagiarism (N = 53, 38.4%). In vitro (28.3%), case reports (21.0%) and narrative reviews (13.8%) were study design most frequently identified in retracted studies. The median time between article publication and date of retraction notice was 1 year (interquartile-range [IQR] = 0 to 2 years). More than half of the retracted articles (n = 82, 59.4%) were cited post-retraction. A retracted article reporting a randomized controlled trial was more likely to appear in journal with higher impact factor than a retracted case report (mean difference [MD] = 2.2; 95% confidence interval [CI] = 1.2,3.1. Articles retracted after 2012 were likely to appear in journals with a lower impact factor (MD=-1.3; 95%CI=-1.8, 0.8). Research misconduct is the main reason for retraction of dental articles. A substantial proportion of these articles were still being cited after their retraction. This report of dental retraction studies informs that more transparency is needed with data reporting in dentistry to improve writing practices in dentistry. A more complete report of retractions and their causes would provide more accurate information to inform researchers and editors to avoid or reduce future cases of retractions. More complete and accurate reporting would increase the overall trust in dental research.