Factors associated with open access publishing costs in oncology journals.

Abstract

11032 Background: The open access (OA) publishing model represents an exciting opportunity to facilitate dissemination of scientific information to global audiences. In contrast to many traditional models, which require readers to pay subscription fees or rely upon institutional subscriptions for article access, the OA model grants free access to all consumers. However, OA publication is often associated with significant article processing charges (APCs) for authors, which may thus serve as a barrier to publication. In this investigation, we aimed to identify journal-level factors associated with OA publication costs in oncology journals. Methods: We identified oncology journals using the SCImago Journal & Country Rank database. All journals under the “Oncology” category that offer an OA publishing option with APC data openly available were included. For all journals, we searched journal websites and tabulated journal characteristics, including APC amount (USD), OA model (hybrid vs full), journal 2-year impact factor (IF), H-index, number of citable documents, primary treatment modality (surgery, radiation, medical, non-specific), treatment site (e.g. breast, etc), and continent of origin. Pearson correlation was used to evaluate univariate linear relationships between variables; for variables with significant correlation, we generated a multiple regression model to identify journal characteristics independently associated with OA APC amount. Results: Of 367 oncology journals screened, 266 met final inclusion criteria. The median APC was 2810 USD (range 0 – 5200). On univariate linear correlation regression testing, journals with the full OA model (p < 0.001), higher journal IF (p < 0.001), higher H-index (p < 0.001), greater number of published articles (p < 0.001), and those from North America or Europe (p < 0.001) tended to have higher OA publishing costs. When these co-variates were analyzed in a multiple regression model, only full OA status (p < 0.001), higher IF (p < 0.001), and North American or European origin (p < 0.001) persisted as independently associated with greater OA APC. Conclusions: Large APCs may serve as a barrier to OA publication and therefore create or exacerbate disparities among scientific investigators seeking to share their research. In this investigation, we find that OA publication costs are greater in oncology journals that utilize the hybrid OA model, have higher IF, and are based in North America or Europe. These findings may inform targeted action to help the oncology community fully appreciate the benefits of open science.