What the World Health Organization asks of sponsors
The World Health Organization asks sponsors to register a trial's methods before it starts, to publish its results on time both in registries and in peer-reviewed journals, and to make the results of past trials available. Its International Clinical Trials Registry Platform draws trials from a network of primary registries.
What the statement on results disclosure asks
In April 2015 the World Health Organization issued a statement on the public disclosure of clinical trial results, with a detailed rationale. The statement reiterates the organization's earlier statement on registering trial methods before a trial starts, and extends it to the timely publication of results, both in clinical trial registries and in peer-reviewed journals. The aim is to ensure that negative results are not underreported and so do not distort the efficacy conclusions drawn from publicly available data. The statement also calls on sponsors to make the results of past trials available. It suggests that later publications and data-sharing initiatives identify the trial's registry number more rigorously, presumably to make meta-analyses easier.
Publishing results in journals within 12 months
One novelty of the statement is that results should also be published in peer-reviewed journals, with open access, within 12 months, to address the well-documented issue of publication bias. To one medical writer that timeline looks very tight, given that the turnaround between manuscript submission and publication can be anywhere between 2 and 6 months. As of 2015, publication rates were lower than rates of results disclosure on trial registries, as might be expected given the relative complexity of publishing in a peer-reviewed journal and the lack of legal requirements. Ben Goldacre, a prominent campaigner for full disclosure of clinical trial results, suggests in a commentary on the statement that the battle for publication in journals may not be worth fighting, arguing that journal articles are often spun and may not report the primary outcome measure. He also points out that academic publishing decisions can be arbitrary, even though many journals may pay lip service to considering negative trials.
Making the results of past trials available
The most substantial novelty in the statement is the requirement that the results of old trials be made available, on the grounds that evidence-based decisions today are based on potentially biased data collected in the past. In 2015 attention was shifting to the disclosure of the results of past trials. The same medical writer judged this a desirable goal that would be difficult to achieve in practice. The writer also found the case for it persuasive but held that putting it into practice is fraught with problems.
Goldacre recognises that throwing resources at disclosing trials with no bearing on current clinical practice is not an efficient approach. He suggests a directed approach, in which preliminary retrospective registration of trials can serve as a guide for researchers who wish to request more details and results.
Even with this directed approach, disclosure may not always be feasible. Retrospective trial registration is likely to leave gaps in the record. Even if a trial is identified and further details are requested, the pertinent information may be hard to retrieve. Many companies will have undergone restructuring or mergers, and the employees responsible for an old project will almost certainly have changed project, company or even country. Archiving systems have also changed, and at best some of the older studies will be available only as scanned versions. In principle the health authorities should also have these trials archived, but many of the problems of data retrieval by companies would apply to them too.
A further obstacle is that the rigorous requirements of ICH only started to be universally applied from the mid-1990s onwards. Before then, which information was recorded, whether the primary endpoint was properly reported and even whether there was a predefined statistical analysis plan would have been somewhat haphazard.
Registries joining the Primary Registry Network
On 25 July 2019 the Lebanese Clinical Trials Registry became a member of the Primary Registry Network of the International Clinical Trials Registry Platform. It was also becoming a data provider, and trials registered with it were to be added to the platform's database. The registry was established by the Lebanese Ministry of Public Health with the support of the WHO Lebanon office and the IT team of the WHO Eastern Mediterranean Regional Office. It was to contribute to research transparency in the Eastern Mediterranean region and to ensure the registration of all clinical trials conducted in Lebanon.