Example

 

"Within 3 months of the end of the final year of funding a description of the study dataset, including a code book, a SAS file of the code used for creating the final study sample, the final study variables and plan for conducting the outcomes analyses outlined in the study protocol will be made available. The investigators will create a complete, cleaned, de-identified copy of the final data set … . A section in the MGH Health Decision Sciences Center website will be created to hold study materials and it will include information for investigators interested in accessing these materials and replicating the findings. The PI will share a de-identified data set with outside investigators according to the policies in the approved IRB [institutional review board] protocol. Investigators may be required to provide evidence of IRB approval (or exemption) and/or complete a data sharing agreement" [84].

​

“The principal investigator will oversee the intra-study data sharing process. All investigators at individual study sites will be given access to the cleaned data sets. Data sets will be housed on the file transfer protocol site created for the study, and all data sets will be password protected. Investigators will have direct access to their own site’s data sets and will have access to other sites data by request. To ensure confidentiality, any data dispersed to investigators will be blinded of any identifying participant information" [85].

​

Explanation

​

Data sharing typically involves making accessible the de-identified individual participant data generated from the trial and the data dictionary detailing the data elements, along with the analytical code used to carry out the planned data analyses.

 

A trial’s data can be stored in a data archive and shared in a variety of ways, such as via an institutional repository (e.g., belonging to the university associated with the trial’s coordinating centre), or a public-facing repository, such as Vivli, GitHub or the YODA Project. Data may be shared at the time of the main trial publication, or only after a specified embargo period (e.g., after one year), enabling planned secondary projects to be completed by the trial team.

 

Data sharing increases transparency, facilitates reproducibility, reduces research waste, and provides data for additional exploration. Some trial groups have worked collaboratively to conduct individual patient data meta-analysis [86]. Data sharing is also associated with an increased number of citations [87, 88]. 

​

Concerns about data sharing involve a perceived risk of re-identification of individuals, especially in trials enrolling participants with rare diseases. Also, inappropriate use of the data by other researchers or groups with a particular agenda can be a barrier to sharing [89].

​

Consent for data sharing can be obtained from trial participants upon enrolment, while approval for data sharing is typically obtained from research ethics committees and data access committees [90]. In a growing number of jurisdictions (e.g., Canada, USA, United Kingdom), some funders now require a commitment on the part of researchers to share their data (e.g., US National Institutes of Health, UK National Institute for Health and Care Research, Gates Foundation). Similarly, some journals require data sharing statements as part of the article submission process (e.g., BMJ, PLOS Medicine) [91, 92].

 

However, data sharing does not happen for most randomised trials [93-97] For example, among 224 randomised trials of interventions for COVID-19, 68 trials (30%) reported in a trial registry the intention to share data, but individual patient data were actually shared upon request or in a repository for only 50 (22%) [96]. 

 

Access to the trial data is also sometimes restricted for trial investigators, and not only for persons outside the trial. The World Medical Association supports the principle that trial investigators retain the right to access the data [98]. Results from randomised trials may be more reproducible and trustworthy when the complete final dataset can be accessed by multiple investigators who contributed the data.

 

For some multicentre trials, only the trial steering committee has access to the full trial dataset. The rationale is to ensure that the overall results are not revealed by an individual study site prior to the main publication. Many of these trials will allow site investigators to access the full dataset if a formal request describing their plans is approved by the steering committee.

 

For some trials, especially those sponsored by industry, access to the final trial dataset is restricted to the sponsor, which can introduce potential concerns about conflicts of interest (Item 3) [99] Reviews of two samples of 108 and 292 trial protocols from 2016 found that 23% to 32% reported plans for who would have access to full dataset [9, 10]. 

​

The protocol should identify the individuals involved in the trial who will have access to the final dataset, and explicitly describe any restrictions in access as well as provide the rationale for any such restrictions. The protocol’s data sharing plans should state whether (and if so, how) the de-identified data, data dictionary, and statistical code will be shared with persons not otherwise involved in the trial. Details on developing a data sharing plan can be found elsewhere [100]. 

​

Summary of key elements to address

• What data and materials will be shared; for example:

  • De-identified individual participant data and data dictionary, analytical code used to process the data

  • Materials associated with the intervention (e.g., handbook or video for nonpharmacological interventions)

• How the data and materials will be shared with trial investigators and external parties, including:

  • Application process to access the data (if applicable)

  • Data transfer process (e.g., via repository or direct transfer to user)

  • Any plans to obtain consent from participants

• If no sharing is planned, this should be clearly stated