Table 3.

Some identified challenges (and suggestions) for data sharing and reanalyses

Identified problems	Suggested solutions for various stakeholders
Data sharing policies leaves responsibility and burden to researchers, to obtain all necessary resources (time, money, technical and organizational tools and services, ethical and legal compliance, etc)	Patients and clinicians should help to develop awareness of a common ownership of the data, intended as common responsibility also in providing all necessary resources to make data sharable for effective and ethical use Researchers should pre-emptively address and seek funding for data sharing Funders should allow investigators to use funds towards data sharing Academic institutions should reward data sharing activities through promotion and tenure
Getting prepared and preplanning for data sharing is still in progress in trials units. There is considerable heterogeneity between different procedures to share data and types of data that are shared	All stakeholders should adopt some clear and homogeneous rules (eg, guidelines) for best practices in data sharing Clinical trials groups should develop comprehensive educational outreach about data sharing Institutional review boards should systematically address data sharing issues
Data sharing on request leaves researchers exclusive discretion on decision about whether to share their own data to other research groups, for which objectives, and under which terms and conditions	Editors should adopt more binding policies than the current ICJME requirement When routine data deposition is not ethically feasible, clinical trials groups should prespecify criteria for data sharing by adopting effective and transparent systems to review requests
There could be some difficulties in contacting corresponding authors, limiting data sharing on request	Researchers and editors should favor data deposition when it is ethically possible
Some identifying information such as date of birth can be found in some databases	Researchers must ensure that databases are deidentified before sharing Institutional review board should provide guidance on the requested level of deidentification for each individual study. It included seeking consent for sharing IPD from trial participants
Shared databases need to be effectively sharable (eg, complete, homogeneous), including meta-data (eg, descriptive labels, description of pre-analysis processing tools, methods of analysis)	Clinical trials groups should develop comprehensive educational outreach about data sharing Researchers should provide details concerning the detailed labels used in the table and analytical code Editors who ask for data and code to be shared should ensure that this material is reviewed
Reproducible research practices are not limited to sharing data, materials, and code. Complete reporting of methods and statistical analyses are also relevant	Clinical trials groups should develop comprehensive educational outreach about data sharing Researchers should also share their detailed statistical analysis plan Reporting guidelines should emphasize best practices in data sharing about computational reproducibility