The Use of Time-To-Event Methods in Dental Research: A Comparison Based on Five Dental Journals Over a 11-year Period

Community Dent Oral Epidemiol. 2012 Feb;40 Suppl 1:36-42. doi: 10.1111/j.1600-0528.2011.00664.x.

Abstract

Objectives: Time-to-event methods are used in multivariate data analysis to describe the relationship between patient variables and the timing of an outcome event. The aims of this study were to evaluate the reporting of statistical techniques and results in dental research papers with special reference to time-to-event (TTE) methods and to create guidelines for the appropriate reporting of these methods.

Methods: All the original research reports published in five dental journals in 1996, 2001, 2005, 2006, and 2007 were reviewed. The evaluation covered 1985 articles that were based on the systematic collection and statistical analysis of research data. Differences between TTE approaches and others were assessed in terms of the justification for the number of cases, description of procedures, statistical references, software used, and statistical figures and tables provided.

Results: Fifty-six papers (2.8% of the total) used time-to-event methods, the frequency of which increased slightly from 1996 to 2007 (P = 0.061). Statistical procedures were described more extensively in the papers, which used TTE methods. Reporting of the statistical methodology in papers using other methods was in general inadequate.

Conclusions: TTE methods are underused in dental research. Authors could well take heed of these results when designing their research, so as to make more use of such methods and to present the results in a manner that is in line with the policy and presentation of the leading dental journals. Authors could also improve their statistical reporting with the help of the guidelines presented here.

Publication types

  • Comparative Study

MeSH terms

  • Animals
  • Bibliometrics*
  • Chi-Square Distribution
  • Dental Research*
  • Humans
  • Periodicals as Topic / statistics & numerical data*
  • Proportional Hazards Models
  • Research Design*
  • Sample Size
  • Software
  • Statistics as Topic