How to publish raw clinical data: guidelines from Trials and the BMJ

An increasing number of peer-reviewed journals and research funding
agencies require authors to make available the raw, unprocessed data
supporting the findings reported in their research articles (click here for
information on why this is important). Just this month, for example, The
American Naturalist announced
that its authors must make their data publicly available as a condition of
publication and the UK Government has also recently launched an open data website.

But there is little
practical guidance available on how data should be shared,
particularly in clinical research where sharing information about
individuals without their consent presents risks to privacy – both from a
legal and ethical perspective.

Recognizing this problem, in March
2009 the editors of the journal Trials
made a committment
to produce guidance on
preparing raw clinical data for publication.

"Preparing raw
clinical data for publication: guidance for journal editors, authors,
and peer reviewers
" – co-published today in Trials and in the BMJ – represents that guidance. It proposes a minimum
standard for anonymizing (or "de-identifying") datasets to protect
patient privacy whilst allowing clinical data to be shared.

Research article

Preparing raw clinical data for publication: guidance
for journal editors, authors, and peer reviewers

Iain Hrynaszkiewicz, Melissa L Norton, Andrew J
Vickers, Douglas G Altman

Trials 2010, 11:9
(29 January 2010)

[Abstract]
[PDF]

The guidance lists 28 items
of personal and clinical information that can make patients identifiable
and recommends that any direct identifiers, such as patients’ names and
addresses, should be removed from datasets before publication. Unless patients have consented to the sharing of their data, datasets
containing three or more indirect identifiers, such as age or sex, should be
reviewed by an independent researcher or ethics committee to determine any risks to privacy, before data are submitted for publication. If
the independent review finds privacy could be at risk, alternatives to fully
open access sharing of data must be considered.

Making raw
clinical data available will benefit future research – and to that end,
human health – and all researchers should obtain consent for sharing of
supporting data when recruiting human subjects. Until this becomes a
routine practice, however, concerns about patient privacy remain a common barrier
to the sharing of information. This practical guidance aims to help
remove this obstacle and enable other scientists and patients to
benefit from full and transparent reporting of research data.

View the latest posts on the Research in progress blog homepage

Comments