Editorials

A clinical trials register for Europe

Stop talking and start making it happen

For more than a decade, clinical trialists and their sponsors have been saying that they want all controlled clinical trials tagged and listed somewhere while they are in progresspreferably on an international register that's simple to use, searchable, and free to anyone who wants to know who is studying what and where. No one doubts that registering ongoing controlled trials is a good idea. The Americans have made an excellent start with their publicly funded register (www.clinicaltrials.gov).¹ And the United Kingdom's Medical Research Council and the NHS Research and Development Programme have made important progress through a meta-register of controlled trials (www.controlled-trials.com), established by Current Controlled Trials, a publisher. Why has it not happened more widely in Europe? Last year the European Science Foundation, an umbrella organisation, advised all its member organisations to register controlled trials through www.controlled-trials.com and to assign each of them a unique identifieran international standard randomised controlled trial number (ISRCTN).² Last month, the foundation hosted a meeting in Frankfurt to review progress, and more importantly, to urge members to stop talking and do something.

The compelling arguments that started this debate 10 years ago remain the same: ^{3 4} The international research effort is chaotic. It is impossible to find out, even with inside knowledge, who is studying what. This means that patients and their doctors don't know about trials they could take part in, researchers don't know if someone has already started a trial they want to do, funding bodies and governments can't easily set research strategies (because they don't know what's going on), and research that's never published disappears without a trace.

All this amounts to duplication of effort and large biases in the information doctors use to treat patients. These biases are not just inconvenient, they cost lives. The "disappearance" of unpublished data on the dangers of prophylactic antiarrythmic drugs in people with heart attack contributed to tens of thousands of avoidable deaths.⁵ There is a growing consensus, voiced at this meeting, that failure to register and report controlled clinical trials somewhere public is a form of research misconduct.

To be fair, local initiatives exist in some member states, notably the United Kingdom, the Netherlands, and Germany. But the European effort adds up to little more than a proliferation of national registers, many of which are confined to trials of new drugs and closed to everyone but the national regulatory body for medicines. Information about ongoing clinical trials remains inaccessible to the people who need it most. Why?

The meeting talked about three major challenges: firstly, money. Setting up and maintaining a public register of trial activity is expensive. A heavyweight bid for funding from the European Commission failed unexpectedly in February this year, leaving existing initiatives demoralised, struggling, and dependent on a mixture of private and public finance. The web based "register of registers" run by Current Controlled Trials is simple, searchable, publicly accessible, and free to users. It contains 24 registers of trials from across the world and nearly 14 000 trials, many of which now have international standard randomised controlled trial numbers. This unique identifier helps eliminate double registering and makes it easy to trace the lifecycle of a trial from protocol to publication. Unfortunately, the European Union's decision not to support public registration until it is required by law means that to offer free access to users the company now has to charge trial sponsors for an international standard randomised controlled trial number. Worse, there are no registers from mainland Europe on the "register of registers."

A publicly funded European database is under development. But it will be confined to trials of drugs, and completely closed. Trialists and sponsors will have access, but only to their own entry. The meeting decided that working with controlled-trials.com was the option most likely to succeed.

The second challenge is how to convince the drug industry to be more open about their ongoing trials. Glaxo Wellcome made a commitment to trial registration in 1998, but the rest of the industry has not followed Glaxo's lead.⁶

Thirdly, how do you persuade researchers to register their controlled trials? In an ideal world people would declare their trials for the greater good, to help with recruitment, and to make their work more visible. In reality, a few extra forms on top of the mountain of paper work already generated by a trial are often too much. Some countries, including the United States, use the law. The Food and Drug Administration Modernization Act 1997 says that all trials must be registered, but only if they are testing new treatments for serious or life threatening diseases.⁷ The commercial sector interprets this as AIDS or cancer trials, which leaves big holes in clinicaltrials.gov and other US registers.⁸

Core items on www.controlled-trials.com Title (including acronym) Disease Inclusion criteria for patients Interventions Contact details for the principle investigator Source of funding Name of trial sponsor

The European Science Foundation meeting decided instead to push for registration to be linked to funding as it already is for many controlled trials in the United Kingdom, where the MRC releases funds only for trials registered on controlled-trials.com and identified by an international standard randomised controlled trial number. The idea is that funding agencies across Europe, led by members of the European Science Foundation, insist that all trials are registered somewhereprobably on a national registerthen ensure that core items are also placed on controlled-trials.com where everyone can see them (box).

The foundation will review progress in six months and challenge the European Commission to follow through its repeated calls for trials to be registered. In the meantime, what should the rest of us be doing? Lobbying politicians. It worked in the United States where a sustained campaign by patient groups led to a change in the law. Patient groups should be urging European politicians to open up the planned European database or pay for another more accessible initiative. Ethics committees should apply pressure to researchers at the ethical review stage. Finally, editors of medical journals should ask authors to register their trials and commit to publishing a trial's international standard randomised controlled trial number alongside the published paper.

Alison Tonks, associate editor. 

BMJ, (atonks{at}bmj.com)

Footnotes

Competing interests: None declared.