The introduction of digital object identifiers (DOIs) to all ISRCTN records in the Current Controlled Trials database, announced today, is an essential part of achieving the aims of the Threaded Publications initiative – making the medical literature more transparent and better connected.
Prospective trial registration should be the first step in transparently disseminating all scientifically relevant information about a clinical trial. Trials can be very expensive to carry out, involve large numbers of subjects and investigators, and generate large amounts of data potentially constituting several scientific articles. The articles from the third International Stroke Trial are a case in point. But unfortunately many clinical trial-related articles are not published, leading to bias in the medical evidence, and even when the outcomes of trials are published in journals and their abstract supplements they can be difficult to find and difficult for systematic reviewers to distinguish from one another.
Part of our vision for Threaded Publications is for all publications reporting the methods, outcomes (including case reports) or analyses of a particular trial to be linked in a thread, regardless of where they are published. DOIs are an industry standard in publishing and are recognized and are interoperable between all publishers which use them, which means that innovations in linking online content can be built on the DOI infrastructure. We hope to achieve Threaded Publications through CrossRef’s DOI infrastructure and CrossMark tool. Registration of a trial with Current Controlled Trials with the assignment of a DOI now provides the opportunity to achieve this vision, with forward and back linking to each trial-related article regardless of where a reader first discovers an article in a thread (see an earlier Threaded Publications post for how this might look to a reader).
DOIs are widely used by academic publishers for helping to ensure the permanence and discoverability of journal articles published on the web. An initial motivation for using DOIs was to avoid broken links in online content. While journal or article website addresses can change, DOIs are permanent.
DOIs are strongly associated with enabling digital scholarly content to be formally cited in journal article reference lists, and therefore for academics to gain credit for the work assigned a DOI. Not all scientists need understand DOIs, but all scientists understand citation. The growing practice of assigning DOIs to scientific datasets and other non-paper-based scholarly outputs is partly driven by the need to incentivize more scientists, through the potential for more citations, to publish their work.
As well as helping achieve Threaded Publications, DOIs for trial records should increase the ability to cite trial records and include them in reference lists. Whether lack of recognition (citations) for registering trials contributes to not all trials being registered is unknown but currently there is no standard way of linking journal articles to clinical trial records (although journals following the ICMJE’s Uniform Requirements for Manuscripts should include trial identifiers in abstracts). By facilitating citation and persistent linking of trial records, we hope that trial registration will become more universally recognized as an essential part of scholarly communication.
BioMed Central’s clinical medicine journals are supporting this development on CCT and Threaded Publications by now requiring citation of trial records, which have a persistent URL or DOI, in reference lists. To help further achieve this we encourage other journals and publishers to begin recommending citation of trial records, as is beginning to happen for datasets. Now we’ve built the citation potential – will they come?