Evaluating Reproducibility and Transparency in Emergency Medicine Publications (Preprint)
BACKGROUND Reproducibility is the backbone to ensuring reliable and transparent research. However, evidence reveals reproducibility may be lacking within emergency medicine. Reproducibility is defined by the replication of a study design and obtaining similar findings. Studies that share research components such as raw data, protocols, and analysis scripts increase the overall reproducibility by aiding study replication. Due to the evidence reproducibility in emergency medicine research may be lacking, investigation of the sharing of these research components is warranted. OBJECTIVE We aimed to assess the reproducibility of empirical research by determining the availability of components required for replication of a study, including materials, raw data, analysis scripts, protocols, and preregistration. METHODS The National Library of Medicine catalog was used to identify English language and MEDLINE-indexed emergency medicine journals. From 30 journals that met the inclusion criteria, 300 publications from January 1, 2014, to December 31, 2018, were randomly sampled using a refined PubMed search. Two investigators were trained for data extraction and analysis and were blinded for independent extraction of data. Extracted data included statements regarding the availability of materials, data, analysis scripts, protocols, and registration. RESULTS After search criteria were set, 25,473 articles were found, from which we randomly selected a sample of 300. Of the 300 articles, only 290 articles were accessible, and of these, 178 articles contained empirical data to analyze. Of the empirical articles eligible for each type of statement, 3.94% (95% confidence interval [95% CI] 1.74% to 6.14%) provided a materials statement, 9.42% (95% CI 6.11% to 12.73%) provided a data statement, 0.72% (95% CI 0% to 1.68%) provided access to the analysis scripts, 2.17% (95% CI 0.52% to 3.82%) linked the protocol, and 7.25% (95% CI 4.31% to 10.18%) were preregistered. CONCLUSIONS Emergency medicine studies are lacking indicators required for reproducibility. The vast majority of studies fail to report factors needed to reproduce research to ensure credibility, which requires an intervention that can be achieved through the collaboration of researchers, peer reviewers, funding agencies, and journals.