April 12, 2004

The public has a right to know

It is often argued that patients and their physicians have a right to access biomedical information and research. Information about clinical trials is of direct concern to patients and the public - and in particular to those who participate in trials as volunteers. Yet until recently there was no easy way to access comprehensive information about clinical trials currently underway, or the results of completed trials. We report here on recent initiatives to provide the public and the research community with online resources about clinical trials.

"Clinical trials have enormous potential for improving medical practice," says Alexa McCray, Director of the Lister Hill National Center for Biomedical Communications at the US National Library of Medicine (NLM). "But reports of such trials are often difficult to find, and in some cases do not even exist." She stresses that limited access to information is one of the greatest barriers to the practice of evidence-based medicine¹.

It is estimated that of the one million or so clinical trials conducted over the last fifty years only about half of them have been reported. Many trials are terminated early and never published. Over one third of randomized controlled trials (RCTs; see Glossary) are not indexed in MEDLINE and are effectively missed by researchers or physicians who rely on MEDLINE to search the literature. This creates what researchers refer to as the 'publication bias'. "The fact that some trial results are never published would not be a problem, except that there is good evidence that the results from unpublished trials are systematically different from those of published trials," claim Kay Dickersin and Drummond Rennie².

For many years, patient advocacy groups and physicians have demanded that information about clinical trials should be readily available to the public. Initiatives such as the Cochrane Collaboration, founded in 1993 by Sir Iain Chalmers, have been trying to address this issue. The Cochrane Collaboration produces systematic reviews of healthcare interventions, in a bid to ensure that healthcare is based on the best available research evidence. In addition, the US Cochrane Center is assembling reports of published RCTs that compare two or more interventions. As part of this effort, over 2,000 journals are being hand-searched, page by page, by people around the globe. The Cochrane Central Register of Controlled Trials (CENTRAL) currently contains more than twice as many reports of trials as MEDLINE. But such mammoth efforts are time consuming, inefficient and expensive. Many have argued that patients and researchers would benefit enormously from systematic registration of all clinical trials².

The ClinicalTrials.gov website is the result of this legislation. "The law originated from a large number of patient groups pushing very hard for greater access. So it's definitely an Open Access issue," says McCray, who coordinates the ClinicalTrials.gov project.

"Patient groups wanted access to trials information without depending only on their physicians, and they wanted a centralized location to find this information." Within 18 months, McCray's team launched ClinicalTrials.gov with a set of 4,400 NIH-sponsored trials. They also worked closely with colleagues at the FDA to produce a detailed Guidance Document describing the submission criteria.

The ClinicalTrials.gov registry was designed to include Phase II, III and IV trials (see Glossary) for serious or life threatening diseases being conducted

in the USA; the law did not mandate for Phase I trials. McCray points out that drug companies are particularly reluctant to disclose full information about early-phase trials, for proprietary reasons. She feels that those fears are misplaced, given that once patients enrol in a trial they are free to talk about their participation. The database currently contains information on 9,600 trials, including NIH-sponsored multi-center studies in the USA and studies from around 250 pharmaceutical companies; both RCTs and non- RCTs are included. "The public has a right to know about the results of the trials," insists McCray.

All records from ClinicalTrials.gov can be downloaded as tagged data in XML (eXtensible Markup Language) format for meta-analysis. "The ClinicalTrials.gov team had to think a lot about standardized data elements and controlled vocabulary, in order to generate a carefully designed data-entry structure," recalls McCray, stressing that data analysis is greatly aided by high-quality data. "Open Access isn't just about throwing information out there. If it's poorly designed or badly presented then it's not so open."

McCray sees the registry as a critical resource to allow patients to find out about ongoing US trials that they wish to participate in. Each entry lists eligibility criteria, location and contact details, and whether the trial is currently recruiting patients. The online record ends with reference to the NLM Identifier, an 8-digit 'NCT number' that can be used to track the trial. McCray sees continual updating as an essential feature of the trials registry. "We have built a very rich information system with targeted links to additional information resources," says McCray. "It's a living, breathing, up-to-date registry. And, of course, that's what everybody wants."

In order to help patients, sponsors, researchers and clinicians to track clinical trials CCT has developed a registration numbering scheme similar to the ISBN (international standard book number) system used to track all books that are published. The International Standard Randomised Controlled Trial Number (ISRCTN) scheme is open to anyone who wants to register - it is not restricted to any one geographic location or to any specific disease area. Applicants have to pay a fee of US$128 (€116; £80) per trial. "Researchers feel that there is increasing pressure to declare that they are doing a particular trial and to make them more publicly accessible," says Greenwood. Some countries also have laws enforcing trial registration. The UK's Research Government Framework is a Department of Health guideline that encourages trial registration.

The idea behind the ISRCTN scheme is to make it easy to track trials that have been initiated in a given disease area. "It is in effect a sign-posting system to take people to the specialists involved in the trial," says Greenwood. It differs from NIH's ClinicalTrials.gov in that it only requires a one-off submission with basic information about the sponsor, the disease, the trial title, the participants and the interventions used. There are also contact details for the sponsor and the lead investigator. The ISRCTN Register is fully Open Access - there is no need to register to view all the information. "All people involved in meta-analysis say that this is a fantastic initiative. In the past, trials that were started but never published were simply lost from the face of the world, whereas here they will be maintained forever," notes Greenwood.

CCT launched the ISRCTN scheme in May 2003, using UK data for the pilot project. The numbering system was developed in close cooperation with the UK's Medical Research Council (MRC). The National Health Service (NHS) has committed to register all the trials that it funds in England. "We have fairly good coverage of all government-funded trials in the UK," says Greenwood. The World Health Organisation (WHO) has recently agreed to number all its trials, as have the Canadian Institutes for Health Research. The CCT website can also be searched using NIH's NCT numbering system. CCT is keen to encourage authors to include ISRCTN numbers in abstracts of any publications reporting trial results. BioMed Central already requires all its authors to do so.

Greenwood regrets the lack of UK legislation requiring registration. "It's really a good-will operation at the moment." She hopes that registration will become the norm in the future, and is particularly disappointed by recent European Parliament legislation, The Clinical Trials Directive, that comes into force this year. This directive mandates the creation of a confidential, rather than open, register of drug trials across member states. CCT did apply for European funding in the past but was unsuccessful. "This does not offer Europe anything like the trials register run by NIH. It is tragic for European citizens. Europe could have created something like the Open Access register that they have in the US." Despite the setbacks, Greenwood is confident that registers of clinical trials will continue to grow as government, patients and the research community understand more about the benefits they bring to healthcare and the fight against disease.