Assessing reproducibility in military medical research: A cross-sectional study of open access practices in unclassified publications (2013–2023)

Background: Accessibility and transparency in scientific research are essential for fostering reproducibility, collaboration and the advancement of knowledge. Military medical research, which is critical for innovation, defense strategy, and medical advancements, is typically disseminated through specialized journals. Recent initiatives by the federal government have pushed to make federally funded research accessible to the general public. However, concerns persist about the degree to which military research articles are openly accessible, limiting the ability of external researchers, policymakers, and the public to engage with these findings.

Methods: This cross-sectional study evaluates reproducibility of articles published in military-focused journals from 2013-2023. Pertinent journals were identified using Google Scholar and a random sample of 250 articles was selected to be included in our study. Two investigators (ML & KM) extracted data on key reproducibility metrics in a masked, duplicate fashion, including availability of data and materials, adherence to open-access policies, and the disclosure of conflicts of interest. The focus of this study was strictly accessing military research articles published in peer-reviewed academic and professional journals, which are not classified or subject to national security restrictions.

Results: Preliminary analysis indicates that a significant proportion of military research articles remain behind paywalls, restricting access for those outside institutional or governmental networks. This was evidenced by 81 (32.8%) of articles inaccessible due to not being open-access, while the investigators were able to access 166 (67.2%) of the articles on a control network without subscriptions to any journals. Of the articles from which data was extracted, 145 were of a study design containing empirical data while 57 had no empirical data. Additionally, three articles were removed for being duplicated in the retrieval. Articles with open-access availability were also evaluated for the transparency of their methods, data sharing practices, and potential for reproducibility. Of accessible articles, 5 (3.0%) stated that their materials were available, 23 (13.9%) stated that their data was available, 2 (1.3%) linked to an accessible protocol, 81 (48.8%) lacked a conflict of interest statement, and 42 (25.3%) lacked a funding statement.

Conclusion: Ensuring the open accessibility of military research articles is crucial for promoting innovation, collaboration, and evidence-based policy decisions. This study highlights existing gaps in accessibility and provides recommendations for improving transparency in military research publications. By advancing open-access practices, the broader scientific and policy communities can better utilize military research to address global challenges and technological advancements.