PARTICIPANTS
Click here for a full list of those attending.

INTRODUCTION
The Current Controlled Trials meeting had been arranged in order to capitalise on the enthusiasm generated by the previous day's conference: 'Registering Information about Randomised Controlled Trials', organised by the Association of the British Pharmaceutical Industry, BMJ Publishing Group and The Lancet. (Click here for a full report of the meeting, and to BMJ and Lancet editorials on the subject.)

CURRENT SCIENCE AND DEVELOPMENTS IN PUBLISHING
Vitek Tracz (Chairman, Current Science Group) explained that as publishers in medicine and life sciences, the Group relies on the advice and expertise of specialists in the field - such as those present - to inform the development of all its publishing initiatives, such as the Current Controlled Trials website and metaRegister of Controlled Trials. (Click here for information on the Group and its products.)

He outlined exciting changes occurring in the publishing field which will facilitate greater openness in the dissemination of medical information. For example, he applauded the recent creation by the National Institutes of Health of 'PubMed Central' - an NIH-operated repository for the electronic distribution of original life sciences research reports, the major goal of which is to provide unrestricted access to reports of primary research.

He confirmed that reports of primary research published by any company in the Current Science Group will be available free to individuals via the web, and will be published in full and without delay in PubMed Central, so providing wide visibility and creating a permanent record of the article. (Click here for press release.)

CURRENT CONTROLLED TRIALS
a) Purpose
Vitek Tracz observed that for some time he had felt that "controlled clinical trials are the single most important area of medical publishing, and the worst-served". Following a meeting of the Ad Hoc Group for Prospective Registration of Controlled Trials held in July 1998, the Current Science Group was asked to create a website through which a register could be made available. Current Controlled Trials Ltd was formed as a separate company, and the Current Controlled Trials website was launched in Autumn 1998.

The priorities for Current Controlled Trials have been:

to develop the metaRegister of Controlled Trials in an attempt to provide free access to a comprehensive database of ongoing and completed trials in all areas of health care. While UK-based organisations had been the first to contribute, the metaRegister was now expanding to become a truly international resource.

to help groups who would like to build their own registers of trials.

b) Journals
In time the Current Controlled Trials site will also feature:
a series of peer-reviewed controlled trial journals featuring protocols, reports on trial design, results, and trial data (with no restriction on space). The first of these will be Current Controlled Trials in Cardiovascular Medicine (joint editors in chief: Drs Curt Furberg and Bertram Pitt). Selection criteria will be based purely on the methodological quality of the trial and not on the 'newsworthiness' of its results.

Apart from reports of primary research (including refereed and non-refereed 'deposited' reports) which would be freely available, the journals will also include literature and web reports, short reviews, thematic reviews, editorials and clinical recommendations.

In discussion, Anne Greenwood (Managing Director, Current Controlled Trials and Current Science Group) confirmed that while reports of primary research would be freely available, the 'secondary' material covering reviews, commentaries and so on, would be available only on a subscription basis. She also hoped that rapid reports (for example when a trial had been stopped) would be included.

There were some suggestions from the floor about the publication of protocols on the site. Liz Wager (GlaxoWellcome, now GlaxoSmithKline) asked whether there were plans to conduct protocol reviews, followed by conditional acceptance of results papers based on the quality of the protocol. Prof. Desmond Julian (Cardiovascular Clinical Trials Forum) felt that complete protocols contain much confidential information, but that reports on trial design were very informative, and important. He added that it would be a good idea if journals would accept results articles only of those trials on which an article on the trial's design had been published. Anne Greenwood undertook to discuss this with the Current Controlled Trials journal editors.

c) Site demonstration
Anne Greenwood then demonstrated the Current Controlled Trials site, with particular emphasis on:

free registration

the Memorandum of Understanding that summarises the responsibilities of those organisations contributing data to the metaRegister and those of Current Controlled Trials in hosting the data

format and search facilities of the metaRegister

the Controlled Trials Links Register - a list of more than 50 links [now 200+] to online registers of controlled trials

the membership of the Current Controlled Trials Advisory Group.

There was some discussion about whether material in the metaRegister would be freely accessible to all users of the site. While Current Controlled Trials had made a commitment to make all information available free of charge, some contributing organisations (for example GlaxoWellcome, now GlaxoSmithKline) chose to restrict access to the information to certain categories of user, i.e. healthcare professionals and researchers. Dr. Robert Skinner (GlaxoWellcome, now GlaxoSmithKline) explained that this was in order to comply with the legal obligation preventing pharmaceutical companies from advertising directly to patients. He acknowledged, however, that it was difficult in practice to prevent patients having access to the information.

John Creasey (retired Civil Servant and information consultant) suggested that, in addition to the Current Controlled Trials Advisory Group, a small Technical Working Group consisting of computer specialists, service intermediaries (e.g. clinical librarians who regularly search databases) and representative users, could be of advantage, provided the Group worked closely with the Advisory Group. Anne Greenwood said that this helpful suggestion would be considered.

d) Unique identifier for randomised controlled trials
Anne Greenwood touched on the need for a unique identifier for every randomised controlled trial, to avoid confusion between trials of similar titles, and to enable researchers and editors to detect multiple publications from the same trial. Current Controlled Trials is considering assuming responsibility for the allocation of such an identifier (an International Standard Randomised Controlled Trial Number).

Current Controlled Trials had consulted organisations involved in the allocation of unique numbers, and had been strongly urged to adopt a numbering system that was not descriptive in any way (that did not, for example, include a number categorising the medical condition, specialty and so on). There was a need to avoid undue complications.

Later discussion revealed support for such a unique identifier as it was generally felt the issue of duplicate records needed to be tackled. Funding organisations, including pharmaceutical companies, tend to have their own identifier for their trials. There was support for the idea of adding to the code already assigned by the funder, as it was felt this would reduce confusion. Anne Greenwood said that Current Science would look into the possibility of using this system.

PROSPECTIVE REGISTRATION OF CONTROLLED TRIALS
a) UK pharmaceutical perspective
Dr Richard Tiner (Medical Director, Association of the British Pharmaceutical Industry - ABPI) reported on trial registration issues from the point of view of industry. The ABPI currently has no policy on trial registration but he felt it likely one would need to be formulated in the near future. He emphasised that as a commercial business, the pharmaceutical industry has a different perspective from other agencies, and needs to look very carefully at the repercussions of making information available.

He reported that Schering Health Care and GlaxoWellcome (now GlaxoSmithKline) have already released information about their trials into the public domain, and that records of other trials co-sponsored by the pharmaceutical industry have been included in the National Research Register (NRR). Those included on the NRR had been registered by investigators, often without the permission of the sponsor, although Kevin Hall (on behalf of the National Research Register) pointed out that guidance had been provided on this issue. There was concern in the industry that permission had not been sought and that some trial records contained inaccuracies. A meeting was planned between the ABPI and the NRR to see how this could be resolved.

The pharmaceutical industry is the largest source of funding for trials in the UK. Its global reach means the scope for trial registration is vast. He outlined some of the concerns expressed by UK pharmaceutical companies, which include:

the amount of information to be provided. A minimum dataset had been agreed by the Ad Hoc Group for Prospective Registration of Controlled Trials at its meeting (July 1998). Some companies might be hesitant about some items, others might wish to provide more information to reduce the number of subsequent enquiries;

intellectual property issues in relation to the phase of the trial (the earlier the phase, the greater the commercial sensitivity of the information);

timing of the release of information: although some companies might consider releasing information prospectively at the appropriate phase of the trial, others might wish to do so retrospectively following licensing/launch of the product;

potential difference in attitude between (UK) subsidiary and headquarters elsewhere. For example, Dr Tiner had had the impression that internationally (particularly in the US) there was greater resistance to disclosure of trial information than in the UK. However, US delegates at the meeting had found US companies co-operative. He would look into this;

resource implications of ensuring accuracy of the information supplied to a register. The possibility of building a database of trials specifically for the UK pharmaceutical industry was being considered.

Dr Tiner felt that it would be necessary for the ABPI to produce guidelines for industry to cover the basic minimum requirements. He concluded that, although only two companies (Schering Health Care and GlaxoWellcome [now GlaxoSmithKline]) had committed themselves so far, many others were considering becoming involved, and progress is being made. The ABPI Medical Committee is actively discussing the principles of registration.

In later discussion, Dr Michael Atkins (independent pharmaceutical physician) asked whether the Medicines Control Agency had been approached on the issue of registering information on ongoing trials. Dr Tiner said this would be considered.

b) UK medical research charity perspective
Professor Sir Leslie Turnberg (Scientific Advisor, Association of Medical Research Charities) explained that medical research charities provide nearly £420m of UK medical research funding every year. Only a small proportion of this research was clinical trials, but nevertheless these still needed to be registered. AMRC member charities recognised the value of trials registration.

Charities traditionally pride themselves on how little they spend on administration so that the bulk of their funds can be spent on the research. Sir Leslie felt that the registration system used, therefore, would need to be very simple and easy to maintain. There was a need for a recognised process for registration so it would be clear to researchers, funders and ethics committees when registration should take place, and who should act on it (see further discussion - c).

Sir Leslie felt that the information would be best provided by the researchers themselves. He felt that some charities would resist assuming responsibility for maintaining, monitoring and checking the data, because of their limited resources.

Discussion followed.

Professor Julian (Cardiovascular Clinical Trials Forum and former Medical Director of the British Heart Foundation) suggested that it would not be so difficult to obtain a minimum dataset from charities as the number of trials was relatively small. The task could be simplified by providing a standard 'summary sheet', specifying the minimum dataset, which should be completed by applicants for funding for a clinical trial. As trials registration is an important part of the research process he felt the activity could be classified as 'research' in the charity's budget allocation rather than 'administration'.

There remained some doubt about whether medical research charities would be prepared to take on the paper work (however limited) involved in registering trials. Dr Iain Chalmers (UK Cochrane Centre) offered to take this on himself if it continued to be a problem. Hilary Leavy (National Asthma Campaign) felt her charity would have no difficulty in complying with trial registration, provided clear guidelines were made available by the AMRC.

Dr William Harlan (National Institutes of Health) emphasised that trials registers must be accurate, and therefore some kind of checking and quality control process, ideally involving the investigator, was critical.

Dr Lesley Stewart (UK Co-ordinating Committee for Cancer Research - UKCCCR), suggested that specialist trials registers already take on the role of checking and quality control and regular updating, as does the UKCCCR, which aims to collect, check and publish information from all randomised cancer trials in the UK. Thus they can help charities and other organisations by providing the administration associated with registration - as the register deals directly with trialists - and also by providing a quality check before records are made available to the metaRegister.

Dr Peter Dukes (Medical Research Council) felt that registration of trials should be seen as part of the charity's (and trialist's) ethical responsibility. It was generally agreed that once identified the information would also be useful in monitoring charitable activity.

Vitek Tracz (Chairman, Current Science Group) suggested that an advisory group (including representatives from Cochrane) might be set up by Current Science to advise and support those registering trials. Current Controlled Trials would devise a standard form for recording the minimum dataset.

c) Comments from Europe
Netherlands
Dr Eduard Klasen (Netherlands Council for Medical and Health Research) gave a short presentation on trial registration in the Netherlands. He reported that clinical trials were popular in the Netherlands, training was provided for those conducting the trials, and patients were generally well motivated to participate. There is reasonable funding, mainly from university hospitals and the government, the government mainly concentrating on interventions involving drugs. Charities also funded trials but to a much lesser extent.

While the situation in the Netherlands is generally good, co-operation between researchers is not perfect and there is currently no central registry of trials. He felt that it was a good time for the Netherlands to act and involve as many agencies as possible to achieve comprehensive trial registration.

Spain
Dr Gerard Urrutia (Spanish Cochrane Centre) reported that the government in Spain had kept a register of all controlled trials conducted for more than 15 years. This register has always been confidential, accessible only to those from the regulatory authority responsible for producing the register. There is now legislation in Spain requiring prospective registration of randomised drug trials, and the recently created Agencia Espanola del Medicamento has assumed responsibility for the management of the Spanish Register.

There have been recent discussions between the Agencia and the Spanish Cochrane Centre about making a limited amount of data publicly accessible through the metaRegister. Agreement needs to be reached with contributors to the register, particularly sponsors. As 90% of trials are funded by industry in Spain, significant confidentiality and intellectual property issues need to be resolved. Over 500 new trials are submitted for approval each year in Spain (80% of which are phase III or IV), making the register a very valuable resource.

OPEN FORUM
Anne Greenwood (Current Science Group) opened the meeting to the floor, requesting feedback on how best to get those involved in trials to submit information. In particular she sought advice on potential pitfalls in seeking information from individual trialists rather than funding organisations.

a) Information submitted by individual trialists
Dr Chalmers commented that information submitted by individuals to the NRR was often of poor quality and contained inaccuracies. There was also concern that accepting submissions from individuals might encourage duplication of records.

Kevin Hall (Connect Consulting/National Research Register) reported that, for the reasons stated, the NRR team had felt it impractical to collect directly from individuals. However, they had now recognised that, when organisations are requested to submit information, the task is often delegated to someone who has little or no connection with the study. This allowed more errors to enter the system, which the NRR is now addressing.

It was generally agreed that the best person to submit information was the investigator(s).

b) 'Unfunded' trials
Small trials may be conducted by PhD students or by qualified staff, in tandem with clinical practice, without direct funding. The only way, at present, of finding out about these studies would be through the ethics committees, which do not currently have mechanisms to make this information available.

c) Process for trials registration and the role of ethics committees
It was reported that Professor Richard Lilford (NHS Executive) - who could not be present at the meeting - was investigating the relative merits of various different processes and sources of trial information, in an attempt to find the best way to create a more comprehensive record of ongoing trials within the NHS.

There was a feeling that ethics committees were well-placed to supply information to the metaRegister once a trial had received ethics approval. Mrs Jennifer Blunt (North-west Multi-centre Research Ethics Committee - MREC) stated that ethics committees were advisory bodies, not part of NHS management, and members worked in a voluntary capacity. Their administrators are already overwhelmed by paper work and have limited manpower and resources.

Ideally, ethics committees would need assurance that projects were peer-reviewed, funded, and in a definitive form before giving approval. This is not often achieved in practice. Ethics committees are also not in a position to distinguish between unnecessary duplication and justified replication of research studies. Nevertheless, she suggested that MRECs might look into the practical aspects of registration via ethics committees. It was possible that a registration sheet could be issued with the final letter of approval of a study, but committees would not be able to monitor compliance.

It was agreed that all involved in approving a study should agree a definite point for trial registration. This should be once the trial is ready to start recruitment, that is after the protocol has gone through any changes following peer review. It was agreed that the best person to submit registration details was the researcher and it should be considered their ethical responsibility to do so.

There was an urgent need to clarify the process, which involves the principal investigator, trial funder, the local and/or multi-centre ethics committee and the hospital trust management, in order to confirm the person/body responsible for trial registration and the appropriate timing. Iain Chalmers (UK Cochrane Centre) called on these people/bodies to clarify the process. Jennifer Blunt reported that trial registration would be on the agenda for discussion at the next meeting of all Multi-centre Research Ethics Committee chairs in December.

Dr Robert Skinner (GlaxoWellcome, now GlaxoSmithKline) felt that it was inappropriate to expect ethics committees to supply the information for trial registration. Every trial should have a 'sponsor' or 'sponsor-investigator' as defined by International Conferences on Harmonisation Good Clinical Practice Guidelines. For pharmaceutical industry funded trials, the individual company, as sponsor, should easily be able to provide this information. He mentioned a draft European Clinical Trial Directive, and wondered whether those drafting it might be approached to include emphasis on the need for a unique identifier.

It was generally agreed that there were many unresolved issues, but that the need for prospective registration of controlled trials was so great that it was important to make a start and then to learn by experience.

Anne Greenwood and Vitek Tracz concluded the meeting with the promise that Current Controlled Trials would:

produce a report of the meeting which would be available on the Current Controlled Trials website

take forward many of the suggestions made, and

keep delegates informed of developments (via the website).

Current Controlled Trials would like to thank Cath Rounding of the UK Cochrane Centre for her significant contribution to the preparation of this meeting report.