Individual patient data sharing – have we considered all angles?
Data transparency is a hot topic at the moment and there are many different views, not only on the best way to proceed, but also on the advantages and disadvantages of the process. Retraction Watch (http://retractionwatch.com/) asked Liz Wager, former chair of the Committee on Publication Ethics (COPE) and currently a publication consultant at Sideview, what her views are regarding the recent proposal on the publication of anonymised individual patient data (IPD) issued by the International Committee of Medical Journal Editors (ICMJE).
The problem of non-publication of trial data is well known and anything that may deter investigators from reporting their findings needs to be minimised. Liz points out that the ICMJE are only asking for the data that support the conclusions drawn in a particular article, which will not help with the identification of selective reporting or outcome switching which, could be argued, is a more significant publication challenge. Liz also queries the demand for IPD access and highlights a recent comment from Ben Goldacre who stated that the number of applications for trial summary data far outstripped requests for IPD.
Additionally, although Liz understands that researchers may be worried about the “poaching” of data, she does not see this as an issue. She is more concerned that IPD sharing may have unforeseen effects on future transparent trial reporting. Specifically, the added costs of collating the data into anonymised user-friendly formats, whilst ensuring compliance is difficult to calculate. However, these costs may have an impact on researchers whose funds are limited.
Liz is keen to stress that she is not against IPD sharing and can see benefits to doing so, but questions whether trial registration and result reporting for all studies regardless of outcome are perhaps more important issues. You can read and comment on Liz’s article here.
Patient centricity / lay summary, Reporting guidelines, Selective publication, Transparency
Leave a Reply