How do statistics in dental articles differ from those articles published in highly visible medical journals?

Abstract

Statistical methods play an important role in medical and dental research. Earlier studies have found that the current use of methods and statistical reporting lead to errors in interpreting results. This study aimed to compare statistical methods and reporting between dental articles and reports published in highly visible medical journals. We analyzed 200 papers published in 2010 in five dental journals and 240 papers published between 2007 and 2011 in the New England Journal of Medicine (NEJM) and the Lancet. We summarized the characteristics of the informed authors, classified the articles by study design type and reviewed the main strategy in the analysis of the primary research question. We also assessed the frequency with which the articles report various statistical methods. We then examined the differences between the dental and medical articles. The median number of authors in articles in the dental versus those in the Lancet and NEJM articles was 5 versus 12. International co-operation in the dental journals was lower than in the medical journals. The proportion of papers reporting “significant” results was 62.5 % in the dental journals and 48.3 % in the Lancet and the NEJM. The percentage frequencies of statistical procedures used in the two sets of articles indicate a broader use of statistical methods in the Lancet and the NEJM, and that both journals were significantly more likely to use advanced statistical methods. Improving the application and presentation of statistical methods in dental articles is essential to meeting the current and future goals of dental research.