Benchmarking open access publication rates for the pharmaceutical industry and research-intensive academic institutions
Rees, T. J.; Philippon, V.; Liew, A.; Baronikova, S.; Gattrell, W. T.; Gordon, J.; Koskenkorva, T. S.; Mysore, S.; Osorio, J.; Koder, T. J.
Show abstract
BackgroundIn biomedical and health sciences, many articles are published open access (OA). OA publication rates continue to grow, especially for pharmaceutical research. Previous analyses of pharmaceutical OA publication rates were labor-intensive and not easily automated. We aimed to create a free, live, online dashboard comparing OA publication rates between pharmaceutical companies and academic institutions. MethodsUsing the Lens - an online platform aggregating full-text content and metadata for over 280 million scholarly works - we built a dashboard showing OA publication rates for medical research articles by authors from the top 40 pharmaceutical companies (pharma) and 40 research-intensive academic institutions (academia). The dashboard details OA rates by model, license and therapy area across three time frames: 12-24 months; 0-12 months; 0-10 years. We downloaded data from the dashboard between 24 July and 4 August 2023 and performed further analysis. ResultsOf the articles published 12-24 months before data extraction, 76.6% by pharma and 69.5% by academia were OA. The most common OA models were gold (pharma, 37%; academia, 41%) and hybrid (pharma, 22%; academia, 11%). Oncology had lower rates of OA articles than other therapy areas. In the 10 years before data extraction, growth in the OA publication rate was generally more rapid for pharmaceutical companies than for academic institutions, regardless of field (change in overall % OA 2013-2023: pharma, 3.1; academia, 1.6). ConclusionsThis dashboard provides novel and regularly updated evidence on the comparative OA publication practices of pharmaceutical companies and academic institutions. In our snapshot analysis, the OA publication rate was higher for pharmaceutical companies than for academic institutions and continues to increase. Notable differences were observed in approaches, in terms of licenses and OA models, which may reflect different institutional practices and circumstances. We encourage others to use this open resource for new research and to report their results.
Matching journals
The top 7 journals account for 50% of the predicted probability mass.