Cancer Clinical Trial Data is Rarely Made Public
(Thursday, December 8, 2022)
A survey of publications for cancer clinical research in major journals found that in about 20% of the papers, the authors promised public availability of the data related to their research and only about 1% provided data that met accepted standards for completeness and transparency. Most journals require data sharing commitment as a condition for publishing research, but it seems that most authors do not publicly share data despite stating so. Data sharing is a critical requirement for independent verification of results, validating models, and guiding future research. However, research is extremely competitive, particularly that leading to future commercial products. While there are mechanisms to protect intellectual property, frequently authors and their organizations hesitate to fully disclose their data to others for fear of helping competition. Hence, many authors intentionally hide their data behind the process. Most authors require requesters to contact them and disclose their intentions with the data prior to releasing the data. This automatically discourages competing organizations from asking for data and hampers competition. Public disclosure policies were created to address the above issue. There is no easy answer. The authors of the study suggested that journals to confirm compliance with data and code sharing policies as a pre-condition to publishing. It is also suggested to better educate the authors accepted for publishing in FAIR data sharing practices which require that the data be posted to a recognized repository, assigned an identifier, data license outlined, organized in non-proprietary formatting. But these are already being done. The deficiencies in public disclosure are due to lack of process but because of the authors’ intentions, fair or not. This is not the first-time data sharing issues have been highlighted and it won’t be the last.
Dr. Mukesh Kumar
Founder & CEO, FDAMap
Linkedin: Mukesh Kumar, PhD, RAC