Data Availability Statements and Data Sharing in Urology: A False Promise?

Abstract

Background and objectiveIt is considered standard for authors of scientific papers to provide access to their raw data. The purpose of this study was to investigate data availability statements (DAS) and the actual availability of data in urology. MethodsThe DAS policies of the top ten urology journals were retrieved. Then 190 selected papers were classified according to their DAS status. Finally, we contacted the corresponding authors of papers that stated that data were available on request to enquire about this possibility. Key findings and limitationsAll journals either required or highly recommended a DAS. Among the selected articles, 52% (99/190) included a DAS stating data availability, most often on reasonable request to the corresponding author. A formal DAS was lacking in 29.5% (56/190) of the articles, with an additional 18.3% (35/190) citing various reasons for data unavailability. On contact, 23.4% (15/64) of corresponding authors indicated a willingness to share their data. Overall, data were unavailable in 73.7% (140/190) of cases. There was no difference between papers dealing with malignant and benign diseases. Conclusions and clinical implicationsThere is a gap between the intention to share data and actual practice in major urological journals. As data sharing plays a critical role in safeguarding the reliability of published results and in the potential for reanalysis and merging of datasets, there is a clear need for improvement. Easier access to data repositories and stronger enforcement of existing journal policies are essential. Patient summaryTo ensure the reliability of data and allow further analyses, major urology journals require authors to make their data available to other researchers when possible. However, in practice we found that data were only accessible for about a quarter of published scientific papers.