Sharing Clinical Trial Data

Data sharing; How to maximize scientific benefit and acknowledge original investigators

Over the last decade, maximizing scientific benefit from clinical trials has been pronounced as a moral obligation towards research participants who put themselves at risk to benefit the scientific society.  Multiple statements have been issued to ensure quality of clinical trial conduct, analysis, and reporting. Recently, the ICMJE has proposed a data sharing statement requiring investigators to share with others the deidentified individual patient data (IPD) no later than 6 months after publication. This valuable step is crucial to reduce selective reporting bias in clinical trials and to allow for reproducibility and testing other hypotheses in secondary analyses.

Currently, obtaining IPD of a clinical trial is not feasible, which deprives the scientific society from the additional knowledge if these data underwent further analyses and yielded other hypotheses. Villain and colleagues [1] identified 38 research questions in orthopedic surgery. Of the 273 RCTs related to these research questions, 56 (21%) of corresponding authors were not accessible (non-functioning emails), 149 (54%) did not reply, 21 (8%) refused, 10 (4%) stated the request is under consideration, while only 37 (13%) agreed to send IPD of their data. Therefore, even with providing a clear research question and analysis plan, guaranteeing   patient confidentiality and secure data storage, and offering co-authorship, many investigators are not encouraged to share IPD of clinical trials. This attitude of some researchers to not share IPD arouses many questions about the reasons behind keeping these data private and might threaten the trust and integrity of clinical research.

The recent editorial of the New England Journal of Medicine [2] advocated that (1) “others might not understand the choices made in defining the parameters”, and (2) “other researchers might use these data to disprove what the original investigators had posited”.

Firstly, if there are special definitions for study parameters, they should be already clear in the published article in order to avoid misinterpretations by readers and misuse of clinical trial results to generate misleading evidence by health practitioners.

Secondly, if researchers are under pressure that others might disprove there results, they will give more attention to data management and analysis; maybe, some authors will analyze their data in multiple scenarios to increase their confidence about their conclusions.

The statement of ICMJE opened the door for ideas about how to provide a credit for investigators who implement then share clinical trial data sets. It should be clear that investigators who implement a clinical trial do not meet ICMJE authorship criteria in the secondary analyses because it is not likely that they will contribute to generating hypothesis, performing analysis, or drafting the manuscript of the secondary analyses. Also, it is not likely that they take responsibility for this publication or approve every secondary analysis from their clinical trial data. However, their contribution to implementing this trial is invulnerable. We propose that investigators who implement a clinical trial should be acknowledged as non-author contributors in the secondary analyses. The list of collaborators is usually available in meta-data of published articles and is citable in electronic indexing databases which gives original investigators a prestigious and continuous credit whenever IPD data of their trial were re-used.

In addition, one single group name should represent those investigators in the authors list of the secondary analysis. We suggest the use of clinical trial registration number as a group name as follow (Group of “registration number”).

Therefore, authors of the secondary analysis will acknowledge the group name among authors and acknowledge all investigators of this group as collaborators. This model of publishing is common in large multicenter studies [3] where the contribution of collaborators is vulnerable but they do not meet authorship credit.

In conclusion, obligatory data sharing will enhance the integrity of conducting and analyzing clinical trials and increase scientific benefit from their data. We propose that original investigators should be acknowledged as collaborators in the secondary analyses of IPD.


References

[1]  Villain B, Dechartres A, Boyer P, Ravaud P. Feasibility of individual patient data meta-analyses in orthopaedic surgery. BMC Med 2015;13:131. doi:10.1186/s12916-015-0376-6.

[2] Longo DL, Drazen JM. Data Sharing. N Engl J Med 2016;374:276–7. doi:10.1056/NEJMe1516564.

[3]  Bhangu A, Kolias AG, Pinkney T, Hall NJ, Fitzgerald JE. Surgical research collaboratives in the UK. Lancet 2013;382:1091–2. doi:10.1016/S0140-6736(13)62013-9.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s