Registering information about randomised controlled trials

Meeting organised by the BMJ Publishing Group, The Lancet, and The Association of the British Pharmaceutical Industry, and held at the British Medical Association, London
Monday, October 4, 1999

This conference took place to highlight issues and challenges related to the registration of randomised controlled trials. It aimed to promote the prospective registration of a minimum data set of trial information, tackle barriers to trial registration, pool experience, and encourage communication between stakeholders. Speakers and participants came from across the world and from diverse backgrounds.

CONTENTS OF REPORT

Keynote Address: Professor Liam Donaldson

Why register trials?: Dr Iain Chalmers

Research Ethics Committees and the interests of patients: Mrs Jennifer Blunt

The NHS National Research Register: Professor Richard Lilford

The pharmaceutical industry and randomised controlled trial registers: Dr Trevor Gibbs

The Medical Research Council and randomised controlled trial registers: Professor George Radda

Medical research charities and randomised controlled trial registers: Professor Sir Leslie Turnberg

Discussion: How can further progress be made within the UK?

Trial registration in the United States: Drs William Harlan and Alexa McCray

Trial registration in Australia: Dr Davina Ghersi

Trial registration in Europe: Dr Gerard Urrutia

International collaboration in trial registration: Dr Kay Dickersin

Practical and other barriers to trial registration: Dr Hamish Cameron/Professor Desmond Julian

Discussion: How can trial registration be fostered?

Session 1: Chaired by Dr Kay Dickersin, co-director, New England Cochrane Center, USA

Keynote Address
Professor Liam Donaldson
Chief Medical Officer, Department of Health, London

Professor Donaldson outlined the relevance of randomised controlled trials to the NHS as the bridge between research and clinical practice. He saw the NHS not only as a sponsor of trials but as the prime user of the information they provide. He described a triangle of discovery, learning and change and saw the NHS Research & Development Programme as facilitating improvements in clinical quality through the practice of evidence-based medicine. He reported that there was still a wide disparity between research and practice but that the NHS Research & Development Programme had been taking steps to address this by initiatives such as the funding of Cochrane work and the setting of evidence based quality standards for health care.

Professor Donaldson acknowledged the valuable efforts of the National Research Register in making details of all studies funded or co-funded by the NHS accessible. He applauded the lead taken by Glaxo Wellcome (now GlaxoSmithKline) and Schering Health Care in disclosing information on their trials and hoped other pharmaceutical companies would follow this lead. He fully supported the international endeavour to make information on ongoing trials more accessible.

Why register trials?
Dr Iain Chalmers
Director, UK Cochrane Centre, Oxford, UK

Dr Chalmers explained the importance of registration of trial information from the point of view of different decision-makers:

Patients, clinicians and other decision makers in the health services - to allow them to contribute to or participate in trials

Researchers and research funding agencies - to allow them to make informed decisions about their own research, to avoid duplication or promote appropriate replication, and to promote collaboration

Users of evidence - to guide policy and practice decisions and decisions about further research

Dr Chalmers showed how The Cochrane Controlled Trials Register had resulted in an increase in the accessibility of information about reports of controlled trials, but highlighted the problem of under-reporting, often because studies had not been submitted for publication because the results had been considered negative or disappointing. He illustrated the impact of such under-reporting with examples demonstrating the possible misleading implications for clinical practice and new research. He suggested that the public should have access to the results of all randomised controlled trials and that not registering and reporting randomised controlled trials is scientific and ethical misconduct.

Dr Chalmers then outlined progress in registering planned, ongoing and unpublished controlled trials over the past 15 years. A key decision was taken in 1996 when Schering Health Care decided to provide information on all their ongoing controlled trials for publication on The Cochrane Controlled Trials Register. Since then there had been encouraging progress, which was reflected in the present conference.

He concluded by emphasising the excellent progress made in establishing a framework for registering information about controlled trials through the metaRegister established by the Current Science Group, and suggested that there is now wide acceptance that all randomised controlled trials should be registered prospectively.

Research Ethics Committees and the interests of patients
Mrs Jennifer Blunt
Chair, NW Multi-Centre Research Ethics Committee, NHS Executive, North West, UK

Mrs Blunt explained the role of research ethics committees as independent regulators, protecting the interests of patients taking part in research. They exist in an advisory capacity and should have no interest in the outcomes of research. Acting as gatekeepers, they protect patients enrolled in research studies, including randomised controlled trials, from unnecessary research, misleading information and fraudulent investigators and practices.

It has been estimated that, at any one time, there are currently half a million patients and healthy volunteers taking part in all forms of research trials in the UK. Of all protocols reviewed by the North West Multi-Centre Research Ethics Committee to date, about one third have been for randomised controlled trials.

She described how research ethics committees function in the UK, outlining the respective roles of the multi-centre research ethics committees (MRECs) and local research ethics committees (LRECs) in assuring patients about the validity of the research, their own welfare and dignity, and the suitability of the investigator.

Mrs Blunt supported registration of randomised controlled trials, feeling it was beneficial to both researchers and patients, who would welcome this spirit of openness. Research ethics committees could also benefit from a register supplying information on the funding status of a trial and information on whether the protocol had undergone peer review. She highlighted the difficulty in choosing the most suitable time for registration in the sequence of peer review, confirmation of funding and ethical approval.

Session 2: Chaired by Dr Richard Horton, Editor, The Lancet, London

The NHS National Research Register
Professor Richard Lilford
NHS Trials Advisor, NHS Executive, West Midlands, UK

Professor Lilford described the NHS National Research Register (NRR), a register of ongoing and recently completed research projects funded or co-funded by the NHS. This is freely available on an internet site, updated quarterly and currently contains about 50,000 records. Approximately 20,000 of these are ongoing studies. Of these records only about 1,500 are thought to be randomised trials.

As well as the benefits of registration detailed by Dr Chalmers, he explained how registration could increase funding probity by giving sponsors confidence that what they were funding was making a real contribution. In many cases a single country study might be of inadequate statistical power, but reference to a register may build confidence that overall statistical power from similar studies done in several countries was likely to be adequate.

Professor Lilford drew attention to the fact that not all trials are funded by external agencies and many small studies are undertaken alongside a clinician's routine work. Although this research may sometimes not be of such high quality, and it is thought only a very small proportion would be randomised controlled trials, it is still important to identify these studies.

Professor Lilford reported that the quality of the records in the NRR needed to be improved and that he was working with the UK Cochrane Centre to enhance this, to ensure that records referring to controlled trials could be distinguished clearly from other research, and to remove duplicate records. He continues to work with funders and investigators in order to identify all randomised controlled trials undertaken within the NHS. He envisaged the use of unique identifiers for trials to ensure accurate registration and avoid double counting.

The pharmaceutical industry and randomised controlled trial registers
Dr Trevor Gibbs
Director, International Medical Operations, Glaxo Wellcome (now GlaxoSmithKline), Middlesex

Dr Gibbs explained the reasoning behind Glaxo Wellcome's (now GlaxoSmithKline) decision to disclose information on their trials. While there were concerns about commercial sensitivity and the resource implications of disclosure, ethical responsibility and the changes in expectations of customers had been prime reasons for their decision. Glaxo Wellcome (now GlaxoSmithKline) felt that disclosure would make their products more credible and increase public confidence that clinical decisions were evidence based.

He felt that in order to make evidence-based decisions on the clinical use of products, information must be widely and quickly available. A time delay is inbuilt in the present publication system hence he looked forward to a time when electronic publishing was more commonplace. Currently, use of the Glaxo Wellcome (now GlaxoSmithKline) register requires a password, released only to research/health-care professionals. This is to prevent accusations of direct advertising to patients, which is illegal in Europe.

Dr Gibbs concluded by summarising Glaxo Wellcome's (now GlaxoSmithKline) commitment to clarify evidence for the best use of their products by publishing information about all their studies, giving each study a unique identification number.

The Medical Research Council and randomised controlled trial registers
Professor George Radda
Chief Executive, Medical Research Council, London

Professor Radda explained the role of the Medical Research Council (MRC) in funding clinical trials since the 1930s. He described the sorts of trials funded and highlighted the MRC's aim to encourage investigators to undertake trials in areas which are not traditionally well funded. There are 144 ongoing MRC trials, 69 of these are in cancer. The annual budget is £13 million and 25% of the trials are co-funded.

Professor Radda explained that, in making its funding decisions, the MRC has financial and other responsibilities to the NHS, clinicians and patients, researchers and other funders and must therefore get it right, particularly as many of their projects involve long-term investment. He saw randomised controlled trial registration as assisting the processes of peer review to indicate the need for a trial, monitoring of progress of ongoing trials and also for disseminating the outcomes of studies. All MRC funded controlled trials are recorded on the MRC Trials Register.

Professor Radda closed by commenting on the need to enhance the quality of randomised controlled trial registers, to ensure that all completed trials were registered, and to initiate registration of planned trials. He emphasised the importance of co-ordinating these registers to ensure their reliability and accuracy.

Medical research charities and randomised controlled trial registers
Professor Sir Leslie Turnberg
Scientific Advisor, Association of Medical Research Charities (AMRC), London

Professor Turnberg explained that the Association of Medical Research Charities (AMRC) has 107 member organisations with a combined funding budget of over £400 million. Along with the MRC and the NHS, charities are thus major sponsors of medical research in the UK. The wealthiest charity is the Wellcome Trust (with a budget of £173 million), which does not support controlled trials. Of the other charities, 15 currently support trials and 5 others intend to do so.

Professor Turnberg reported on his recent survey of attitudes of the member charities towards registration of randomised controlled trials. Although the general feeling was supportive, concerns had been raised about ownership of the data and judgements about when repeating a trial would be considered duplication or appropriate replication.

A primary concern, however, was the administrative burden registration would bring for organisations who traditionally spend very little on administration and aim to put almost all of their funds into the research itself. Professor Turnberg was left with the impression that unless the registration process was very simple charities may find it difficult to participate.

Discussion: How can further progress be made within the UK?

The first part of the discussion focused on the pharmaceutical industry and whether randomised controlled trial registration was seen as a commercial threat.

Hilary Murray, Manchester Medical Ethics Group, questioned whether it was thought registration may cause a race to publish when similar trials were running concurrently.

Trevor Gibbs, Glaxo Wellcome (now GlaxoSmithKline), felt that it was not a major issue for the pharmaceutical industry and that it would add power to the trial.

Alan Jones requested feedback on the sort of arguments that could be used for companies who are not even prepared to consider sharing their trial information, particularly those based in the US. Iain Chalmers, UK Cochrane Centre, expressed the view that resistance by the pharmaceutical companies in the USA reflected a cultural difference and that pressure from consumers seemed to be beginning to force progress there. It was important to emphasise the ethical responsibility in trial registration. Kay Dickersin, New England Cochrane Center, emphasised that in the USA the law requires drug companies to register their trials and her experience and that of Alexa McCray, National Library of Medicine, had been very positive.

Jimmy Volmink, South African Cochrane Centre, reported that a proposal for the registration of all new trials in South Africa had been prepared and they were hoping for legislation, but that it would be interesting to see if industry decided to do their studies elsewhere if this legislation was passed. Richard Smith, BMJ, suggested that UK law should move along the same lines.

It became apparent in the subsequent discussion that there is a disconnect between the experience of those in the USA requesting information on trials compared with those living outside the USA. Alexa McCray asked Trevor Gibbs why information on their web site was only available to health professionals and researchers. He answered that there were legal concerns in making the information available as it may be seen as consumer advertising, which is illegal in Europe. Moreover, one of Glaxo Wellcome's (now GlaxoSmithKline) intentions in providing the information is to assist researchers wishing to write systematic reviews.

Richard Tiner, Association of the British Pharmaceutical Industry, highlighted the issue of information about trials co-sponsored by the pharmaceutical industry being posted on the NRR by NHS researchers. This needed to be resolved, as the information was often put there without permission and sometimes contained inaccuracies.

Discussion then moved on to look at whether patients should be encouraged to enter randomised controlled trials. It was felt that there was a general lack of understanding of science amongst the public and in particular of randomised controlled trials. It was suggested that by encouraging participants to enter trials, self-selection biases might be introduced. Some thought this an issue, others disagreed.

Desmond Julian, Cardiovascular Clinical Trials Forum, directed the discussion to consider criteria for inclusion in registration. A further point was whether randomised controlled trials within PhD theses should be included. George Radda reminded the conference that many trials have non-drug protocols and other areas such as therapy trials have their own issues.

Discussion moved on to look at how further progress could be made in implementing randomised controlled trial registration given the unanimous agreement that it was, in principle, a worthwhile endeavour. Iain Chalmers suggested that research ethics committees could play an important role in enforcing registration. Jennifer Blunt, NW MREC, replied that it would not be a simple process as there were many stages to ethical approval and several agencies involved, so while it was worth considering, it would not be an easy option. Richard Lilford suggested that various different methods of registration should be piloted as there did not appear to be a clear way forward.

Close of morning sessions

Session 3: Chaired by Professor Eduard Klasen, Director, Council for Medical and Health Research, The Netherlands

Trial Registration in the United States
Dr William Harlan
Associate Director for Disease Prevention, US National Institutes of Health (NIH), Bethesda, Maryland, USA, and
Dr Alexa McCray
Director, Lister Hill National Center for Biomedical Communications, National Library of Medicine, Bethesda, Maryland, USA

Dr Harlan briefly described the history of trial registration in the USA. During the past 5 years there had been an increased demand for information from the public and from US health care providers, in particular, for information about randomised controlled trials in cancer. This had prompted the 1997 Food and Drug Administration Modernization Act (FDAMA), stating that it is a legal obligation to make information on trials available in areas of serious or life-threatening diseases or conditions.

In addition, an increased demand for participants in trials had led to greater provision of freely accessible information from pharmaceutical companies as the principal funders of randomised controlled trials. The National Institutes of Health is the second most important funder in the USA.

Dr McCray gave further details about the FDAMA of November 1997. This Act required the establishment of a national database of clinical trial information for experimental treatments for serious or life-threatening diseases or conditions. This was applicable to both federally and privately funded trials. At the end of 1998 the National Library of Medicine (NLM) was asked to develop and host this central database. The NLM is currently working to develop the trials database to contain a basic minimum data set for all registered trials.

This minimum data set will comprise a basic description of the study (title, design, phase, disease condition and intervention), recruitment information (status of trial and eligibility criteria), location of study and contact information, and other administrative data (study ID number, NIH grant number, principal investigator). Optional information may be supplied including key words, results, literature references and URLs.

It has been designed for patients to use so should be easily accessible and user friendly. There are plans to establish a toll free telephone number to facilitate dissemination of this database. Dr McCray reported that the first prototype of this system should be made publicly available by the end of this year, although initially it will only contain NIH studies.

Trial Registration in Australia
Dr Davina Ghersi
Research Fellow, NHMRC Clinical Trials Centre, Sydney, Australia

Registration of clinical trials in Australia started in 1989 with the formation of the NHMRC Clinical Trials Centre. In 1995 the Centre was provided with funds by the New South Wales Cancer Council to establish a prospective register of clinical trials in cancer. This register can be accessed via the internet. Collaborations with a number of consumer groups are in the process of being formed to ensure that the information provided on the web pages is relevant and easy to understand.

The positive experience in cancer has enabled advocates of prospective registration to negotiate more effectively with government agencies for the registration of all randomised trials. In 1995 the NHMRC Standing Committee on the Quality of Health Care recommended that a comprehensive prospective register of trials be established, a recommendation endorsed by the Australian Health Ethics Committee in 1996. Further support came from Consumer's Health Forum which released its policy on registration of trials in 1998.

A significant development in 1999 has been the release of the report of Health and Medical Research Strategic Review by the Commonwealth Department of Health and Aged Care. The review recommends that "Australia should implement systems to record, document and disseminate information about trials and studies to reduce duplication and increase collaboration. In particular this register should cover clinical trials to allow for rationalisation and streamlining of the ethics review process." A mechanism for funding such a register is currently being considered by the Department.

Trial Registration in Europe
Dr Gerard Urrutia
Centro Cochrane Español, Barcelona, Spain

Dr Urrutia gave a detailed description of how the Spanish Registry of Clinical Trials had developed and then a brief comparison with the level of trial registration in other European countries. Dr Urrutia explained that clinical trials in Spain are regulated by law. Since 1978, no clinical trial involving medications may be carried out without having previously received authorisation from the national regulatory authority. Clinical trials comparing techniques or procedures are not covered by the law, however.

Since 1982 it has been possible to create and maintain a national clinical trials database (Spanish Clinical Trials Database). Between 1982 and 1988 the information on each trial contained in this database was restricted to basic identifying and administrative details. In 1988 the data set was expanded to include methodological and ethical data. In May 1999, the Spanish Drug Agency assumed the role of regulatory authority for clinical trials and developed a new database. The information contained in this database is collected on a "Clinical Trials Summary Sheet" which has eliminated the fields for qualitative evaluation which had previously been included.

Dr Urrutia outlined the problems encountered with the Spanish registry to date, including the lack of information on whether trials were completed or interrupted, or whether the trials were finally published. The information on the database has always been totally confidential. Dr Urrutia reported that the Spanish Drug Agency had recently met with the Spanish Cochrane Centre. Although there are still issues to be resolved and consultation needed with the pharmaceutical industry, it has been agreed in principle to contribute a limited data set to the metaRegister of Controlled Trials set up by the Current Science Group.

Dr Urrutia concluded by outlining future perspectives for Europe, in particular, regarding a proposal for a European Directive on Clinical Trials to attempt to reconcile the local legislation of the various member states of the European Union. This proposal mentions the need to create a European database of trials. Dr Urrutia's survey of other European countries indicated that national clinical trials registries required by law are either non-existent or only in very early stages of development.

International collaboration in trial registration
Dr Kay Dickersin
New England Cochrane Center, Providence, Rhode Island, USA

Dr Dickersin described the changes in international collaboration in trial registration in the last decade. Over the years, there have been numerous efforts to develop trial registers, none of them designed to be comprehensive, by individual scientists and government initiatives around the world. Calls for more comprehensive mechanisms to register trials prospectively have been publicised since the 1980s, and since its start in 1993, the Cochrane Collaboration has promoted this idea.

Dr Dickersin outlined the efforts that have been made to launch comprehensive trials registers. These include government initiatives such as the NIH trials register in the US (1974-79), ClinProt/Physician Data Query for cancer trials (1975 to present), NHS National Research Register in the UK (1998), and those from publishers such as CenterWatch in the US (1994) and the Current Science Group's metaRegister (1998).

Since 1987, Dr Dickersin and her colleagues have been compiling an "International Register of Registers" and have observed a recent dramatic increase in the appearance of new registers. Until 1998, fewer than 100 registers had been identified and these were mostly available on paper only. However, since 1998 she and her colleagues have identified over 500 online registers of randomised controlled trials (over 30% being funded by government), and the number is increasing rapidly.

She attributes much of this expansion to a recognition of the power of the internet and an increase in its use over the last 5 years, as well as consumer interest in the issue. Despite the large number of trials registers currently online, the subject matter they cover is not what those searching the internet for health information appear to be seeking, however.

Dr Dickersin felt that committed partnerships between investigators, the pharmaceutical and device industry, publishers, consumers, and others were required in order to bring about useful international registers. These partnerships require a consideration of all stakeholders' needs.

Dr Dickersin considered how an international register of trials might be developed and felt that a single comprehensive register may not be possible. The register proposed by her group at Brown University would comprise a core content, made up of the contents of multiple searchable registers, linked but not actually merged, and would have multiple levels of access and individualised output, depending on the needs of the user.

Session 4: Chaired by Mr Mike Wallace, Schering Health Care, West Sussex.

Practical and other barriers to trial registration
Dr Hamish A Cameron
Cardiovascular Therapy Area Director, AstraZeneca, Gothenburg, Sweden

Dr Cameron outlined a number of perceived barriers and issues related to registration, from the perspective of the pharmaceutical industry. These included problems resulting from international pharmaceutical companies having offices in countries with varying legislation, issues of competition between companies, confidentiality of information and intellectual property issues, the influence of changing customers and the impact of electronic communication.

Another big issue is the practical implications of adopting a disclosure policy. Resources would need to be found for staff to set up and administer such registers and also to field enquiries.

Dr Cameron felt that it might take quite a while for many pharmaceutical companies to participate in registration of trials, but that with reassurance and support, it would happen gradually. He hoped that industry might recognise that having a publicly accessible register of all trials could be seen as a valuable resource.

Professor Desmond Julian Emeritus Professor of Cardiology (Newcastle upon Tyne), Cardiovascular Clinical Trials Forum, London

Professor Julian described the problems he had encountered in his own experience of developing a database of clinical trials in cardiovascular disease (Cardiosource, Elsevier). He experienced a lack of time/interest/feedback from organisations, and suspicion expressed by industry.

Professor Julian outlined two approaches to establishing registers - top-down and bottom-up. Top-down registers are those created by funding organisations, for example, government agencies and charities, but seldom from industry. Bottom-up registers are derived from data collection by speciality groups, patient groups or research groups, the data being obtained from top-down registers, publications, and by direct contact with individual researchers and industry.

He felt the main considerations in producing a useful trials register were to ensure good search functions, to cater for your audience, to keep the register up to date, and to link to other appropriate publications. He saw the emergence of internet publishing as providing an ideal means of solving many of these problems. In addition he felt trial registration had to be made as easy as possible for contributors.

Discussion: How can trial registration be fostered?

Anne Greenwood, Current Science Group, reported that she was working with the Cochrane Cancer Network to pilot a system for allocating a unique identifier (ISRCTN: International Standard Randomised Controlled Trial Number) to trials when they are submitted to the metaRegister.

Richard Tiner, ABPI, requested clarifications on the apparent disconnect between UK and USA opinions on the degree of co-operation from pharmaceutical companies in the USA.

Kay Dickersin suggested that a pharmaceutical company's willingness to co-operate depended primarily on what they were being asked to contribute. They certainly would not be interested in making information on their phase I trials available. Alexa McCray agreed with this, reporting that in her experience there was not a problem with obtaining information from late phase II onwards. She also suggested that the degree of co-operation might depend on how much information was being asked for and that an agreed minimum data set should not cause problems.

Iain Chalmers felt progress towards registration was being hindered by poorly organised internal information systems within large organisations. This meant that it was difficult to pass information on to third parties, even if the will existed. If internal systems were maintained more efficiently it would be much easier to provide what was requested.

Both Hamish Cameron, AstraZeneca, and Mike Wallace, Schering Health Care, stated that from an industrial point of view there was still a degree of anxiety about what registration would mean for the pharmaceutical industry. They hope and expect that with good dialogue this feeling will dissipate.

Richard Smith, BMJ, suggested that to make really fast progress there needed to be law to enforce registration. While some agreed with this, others disagreed. It was suggested that it would be better to approach the issue by teaching those in the health professions about their ethical responsibilities.

It was felt that if there were going to be a law to enforce registration of trials then the proposition for such a law would be most successful if it came from consumer groups. Hazel Thornton, Consumers Advisory Group for Clinical Trials, reported that the consumer movement in the UK was developing slowly, and was less powerful than those in some other countries, such as the USA and Australia.

Kay Dickersin reported that the culture of the USA supported a strong consumer movement, with the feeling that it was possible to demand change and get it. The UK has never had this culture. Janet Darbyshire, MRC Clinical Trials Centre, agreed but reported that things were changing and that there was strong consumer involvement in the field of HIV and AIDS, which demanded information on current research.

Thank you to the BMJ Publishing Group, The Lancet, and the Association of the British Pharmaceutical Industry for organising this meeting.

Report by Allison Hirst (Cochrane Cancer Network) and Cath Rounding (UK Cochrane Centre) with thanks to all speakers at the conference, Inneke Loke and Iain Chalmers for their comments.

Cath Rounding will be happy to help with any enquires: tel: +44 1865 516300, fax: +44 1865 516311, e-mail: crounding@cochrane.co.uk