To share or not to share? Expected pros and cons of data sharing in radiological research

Abstract

The aims of this paper are to illustrate the trend towards data sharing, i.e. the regulated availability of the original patientlevel data obtained during a study, and to discuss the expected advantages (pros) and disadvantages (cons) of data sharing in radiological research. Expected pros include the potential for verification of original results with alternative or supplementary analyses (including estimation of reproducibility), advancement of knowledge by providing new results by testing new hypotheses (not explored by the original authors) on preexisting databases, larger scale analyses based on individualpatient data, enhanced multidisciplinary cooperation, reduced publication of false studies, improved clinical practice, and reduced cost and time for clinical research. Expected cons are outlined as the risk that the original authors could not exploit the entire potential of the data they obtained, possible failures in patients' privacy protection, technical barriers such as the lack of standard formats, and possible data misinterpretation. Finally, open issues regarding data ownership, the role of individual patients, advocacy groups and funding institutions in decision making about sharing of data and images are discussed. Key Points• Regulated availability of patient-level data of published clinical studies (data-sharing) is expected.• Expected benefits include verification/advancement of knowledge, reduced cost/time of research, clinical improvement.• Potential drawbacks include faults in patients' identity protection and data misinterpretation.

Artificial intelligence in medical imaging: threat or opportunity? Radiologists again at the forefront of innovation in medicine

“…The attention to inter- and intra-reader variability [ 19 ] and the work committed to improve the repeatability and reproducibility of medical imaging over the past decades proves the need for reproducible radiological results. In a broader perspective, the trend toward data sharing also works in this case [ 20 ]. The key point is that AI has the potential to replace many of the routine detection, characterisation and quantification tasks currently performed by radiologists using cognitive ability, as well as to accomplish the integration of data mining of electronic medical records in the process [ 1 , 7 , 21 ].…”

Statistical significance: p value, 0.05 threshold, and applications to radiomics—reasons for a conservative approach

“…Efforts should be instead paid to the researchers' training and to policies that may further improve research quality, such as a priori registration of a trial protocol, the need for a professional statistician for data analysis, and data sharing [6]. Especially, data sharing has the potential for verification by independent authors of the results presented in a given publication [28]. When data are shared, they may be used by other researchers to perform alternative or supplementary analyses.…”

Clinical Imaging Research: Higher Evidence, Global Collaboration, Improved Reporting, and Data Sharing Are the Grand Challenges

“…This may be facilitated by requiring the checklist to be viewed before beginning a study, with responses submitted to a research the clinical impact by involving interested members of the scientific community who were not directly involved in the primary trial (48,49). However, a recent study showed that only three of 18 general imaging journals encouraged data sharing and only one journal requested it (48). Structured reporting in radiology clinical practice will certainly improve the potential for data sharing beyond clinical trials.…”