Publication Bias in Medical Research: Issues and Communities

Publication bias is a broadly discussed phenomenon related to the reporting of the outcome of clinical studies. As a part of the UNCOVER project this work aimed at the identification of members of the research community as stakeholders for interviews and workshops how to overcome publication bias. For this objective relevant literature was analysed by the following bibliometric approaches: networks of co-authorships and affiliated institutions, co-citation analysis and bibliographic coupling over a twenty-year timespan. Research communities were mapped and examined and research issues where identified by applying bibliographic coupling and co-citation maps. The analysis showed a high dominance of publications from evidence based medicine like systematic reviews and meta-analysis performed for different medical topics. Most authors used and cited previous research, findings and methods how to proceed with publications bias. They are to be seen as experienced "users and applicants" stakeholder group. A second dominant group of publications is related to research on publication bias from different aspects: publications and data for systematic reviews, adequacy of databases, publication of negative results, registration of clinical trials, outcome reporting, protocols of clinical trials, sponsorship bias, role of editors, ethic committees, guidelines for systematic reviews, regulation of clinical trials and methods for meta-analysis. Stakeholders were selected on the basis of research issues and affiliated organizations.