Scope, definition, and consequences of publication bias in clinical trials

Objectives

  • Framing of hypotheses, definition of suitable and measurable bibliometric features, and statistical indicators
  • Extraction, pre-processing, and standardization of data from information sources
  • Investigation of the influence of registered vs. non-registered studies on the obtained “bibliometric profile” in the cases of systematic reviews 
  • Interpretation of characteristic features and conclusions given the current measures against publication bias

Task 2.1 - Definition of features and indicators for bibliometric analysis and data extraction

  • Review of literature, selection and definition of features for bibliometric respectively text-analytic analyses
  • Selection and definition of indicators, network indicators; and indicators of relational maps, including thematic clusters
  • Compilation of data sets from systematic reviews and additional data sets from clinical trial registries (including but not limited to Clinicaltrials.gov) and other data sources (journal impact factors, etc.)

Task 2.2 - Bibliometric analysis of characteristic features distinctive between registered vs. non-registered studies

  • Coupled bibliometric and downstream statistical analysis on datasets by using stand-alone, network, and relational map analysis 
  • Conduct “co-analyses” (including co-author, co-citation, and co-keyword analysis) to measure and investigate the bibliometric profile of registered vs. non-registered studies across features 
  • Differences between registered vs. non-registered studies, assessment of characteristic features, and comparison between data sets

Task 2.3 - Characteristic bibliometric features of publication bias and conclusions

  • Description of informative features and means for finding differences between registered vs. non-registered studies in terms of bibliometric profile
  • Interpretation of characteristic features in registered vs. non-registered studies in relation to measures to reduce or prevent publication bias and framing of lessons learned