Featuring a study defending double-blind peer review, what’s next in the campaign for preregistration, and trial disclosure rates in the pharmaceutical industry.
Perhaps double-blinding isn’t so bad after all via Times Higher Education
Peer review is a cornerstone of scholarly integrity in theory but can be imperfect in practice due to the fallibility of reviewers. Double-blinding has arisen in response to this fact and is increasingly used by journals conducting peer reviews with the intention of mitigating subliminal bias or, in rare cases, active sabotage. A frequent criticism, however, is that experts in the field are likely to be so familiar with the literature that reviewers are able to guess the identity of authors, making double-blinding pointless. The authors on this study are not so sure – their results show that reviewers are unable to guess the identity of authors on papers that they review 74–90% of the time.
Looking back on the preregistration challenge: what comes next? via Center for Open Science
Study preregistration is commonplace in medical research: all clinical trials must be registered before commencement – their methods specified, hypotheses stated and planned analyses outlined. Beyond medicine, however, preregistration is not standard practice, which allows researcher biases to sneak into the research and analysis process. This lack of rigour led to the famous claim by Professor John Ioannidis in 2005 that most published research was incorrect. In recent years, the Open Science Framework, among others, has been pushing to make the preregistration of studies ubiquitous across all fields. December 2018 will mark the end of this first push, the Preregistration Challenge, which has led to over 2200 researchers from many disciplines starting to register their studies. The outcomes of these efforts will begin to emerge in the next few months as papers are being written up, and the awards of the Challenge are being announced and presented.
Contrary to common perceptions, academia falls behind the pharmaceutical industry in terms of clinical trial disclosure, as shown by a study based on Ben Goldacre’s trials tracker algorithm and presented at the Evidence Live conference this week. The aim was to quantify reporting levels in medical research and to compare industry-sponsored with academic- and government-sponsored research. Its findings suggest that industry reporting rates averaged 74%, whereas for non-commercial research organizations, the average was 46% The authors concluded that there was room for improvement in both groups, but that industry was taking the lead in the push towards full disclosure. The trials tracker algorithm has been accused of limited precision, but new and more accurate tools in development seem set to confirm these findings.
New resources on open access, preprints and peer review from Network Pharma via NetworkPharma.tv
Following the latest brunch club session, Network Pharma has this week released a new suite of resources on Open Pharma topics, focusing on open access, peer review, preprints and the journal submission process. Featuring representatives from the Taylor & Francis Group, these resources aim to provide an introduction to these important topics.