Intended for healthcare professionals

Editorials

Local research ethics committees

BMJ 1995; 311 doi: https://doi.org/10.1136/bmj.311.7006.639 (Published 09 September 1995) Cite this as: BMJ 1995;311:639

This article has a correction. Please see:

  1. K G M M Alberti
  1. Professor of medicine Human Diabetes and Metabolism Research Centre, University of Newcastle, The Medical School, Newcastle upon Tyne NE2 4HH

    Time to grab several bulls by the horns

    Nowadays, it should be obvious to all participating in human research that ethical principles should be followed and that approval of a study by an ethics committee should always be sought. Despite the appalling lessons of the second world war this has not always been the case, and even now examples of unethical research are reported regularly, with researchers dazzled by the potential scientific rewards forgetting the moral and humane principles to which they should adhere. Ethics committees thus continue to be a vital safeguard, although their birth has been long and tortured and they continue to be criticised for their often idiosyncratic ways of operating.

    The main drive to the formation of ethics committees in Britain came from the World Medical Association's Declaration of Helsinki1 and the publication of Pappworth's The Human Guinea Pig.2 The British Medical Research Council reported in 1963 on the responsibility of researchers for the wellbeing of their subjects, and in 1967 the Royal College of Physicians recommended that all research should be subjected to ethical review.3 In 1968 the Department of Health recommended that hospitals should establish ethics review boards but on an informal basis. Guidelines on how they should function appeared from the royal college in 1984 and in 1989 from the Department of Health.4 But it was not until 1991 that the Department of Health formally required every health district to have a local research ethics committee and gave more detailed guidance on the composition of these committees and how they should function.5

    The 1991 document was welcome, although discerning why it took more than 20 years from the first loose guidance to reach that point is difficult. In the past four years it has become obvious that many problems remain.6 Alternative, or at the least complementary, advice has come from the European Union regarding drug trials7 and research involving patients' records.8 In addition, some multinational drug manufacturers adhere to guidance on good clinical practice for trials on medicinal products,9 and the Association of the British Pharmaceutical Industry has specific requirements, which are now largely consistent with national thinking.

    More important are the large variations in practice among different ethics committees, which partly stem from the lack of specific guidance before 1991, when committees developed their own, local methods of working. This has led to a progressively larger groundswell of complaint from research workers, frustrated at the delays incurred, the costs involved, and the unnecessary duplication of effort. This has been a particular problem for multicentre trials, as reviewed by Meade from the standpoint of an organiser of large scale trials.10 The problems are described graphically in three papers in the current issue11 12 13 as well as two oft quoted earlier papers by Gilbert et al14 and Harries et al.15

    Interestingly, all three of this latest clutch of papers involve studies of children or adolescents and their families. All involve questionnaires with or without interviews rather than physically invasive procedures. It is worth noting that these types of “social” protocols seem to create the biggest uncertainty for ethics committees, for reasons outlined below. Each study approached a number of ethics committees, ranging from 1313 to 162.12 The problems were sadly common. Thus Middle et al had to fill in 118 different application forms12; only in the old Northern and South Western Regional Health Authorities were common forms in use. In the two other studies all forms were different. These were not just minor differences but vast variations ranging from two to 10 pages, some with full protocols, some with virtually none. Not all committees gave guidance and not all required consent forms. In two of the three studies 10-20% of committees indicated that formal approval was not required.

    The time to obtain a response also varied greatly, from six to 161 days (median 56 days) in White et al's study.11 One quarter of committees had not responded within three months in Middle et al's study.12 A large variation in changes requested was also found. In a few cases no change was requested; most commonly changes in the consent form were requested, but a large range of changes to the protocol or other aspects of the study were also requested. Often two suggestions would be diametrically opposed. The current situation is obviously wasteful of time and resources—Middle et al calculated that seven to eight weeks of staff time was needed, that more than 1000 protocols and application forms had to be sent out, and that direct costs were nearly £5000.12

    Lack of clear guidance from the top

    So what are the problems and how may they be resolved? Firstly, does the inconsistent behaviour of committees partly relate to the lack of clear guidance already alluded to? This is now well recognised and is changing as membership of committees changes. The Department of Health has organised training days for ethics committees, which have proved valuable and popular—but many more are needed. Training days have also been organised in the Northern and Yorkshire Regional Health Authority under the auspices of the Research and Development Directorate, and these will continue regularly for new members of our 16 local research ethics committees. It would also help if there was consistent central guidance on the handling of difficult areas—for example, when patients cannot give informed consent, or sociological studies.

    A further difficulty concerns the members of the committees and their characters. Two dominant researchers can bias a committee in one direction, while strong minded lay members can have the opposite effect. The impact of this is thoughtfully described by Foster.16 As she states, committees must balance the different basic questions that must be answered: Does the project ask an important question? Will the study design allow that question to be answered? Are the risks to the research subjects acceptable? Will the autonomy of the subjects be respected by their consent being obtained?

    Researchers and committees face several other common problems. High on the list is the lack of a common application form (in all but the old Northern and South Western Regional Health Authorities). There is also the need for researchers to attend a meeting of every committee involved, despite a successful pilot study of using a regional lead ethics committee for applications affecting three or more districts. There is also doubt about the scientific competence of committees. It is almost impossible for them to cover all areas adequately, and they may lack a basic understanding of statistics and study design, making it impossible for them to assess properly the scientific validity of a proposal. Consent forms lack consistency, and payments to doctors for drug trials vary, particularly in primary health care, although many trials now follow the BMA's guidance.17 Finally, workload varies greatly, with some committees faced with more than 500 applications a year while others deal with fewer than 40. This problem is currently compounded by student projects from both medical schools and colleges of nursing. Many such studies lack the power to generate any useful data and should perhaps be rejected, although the pressure is to pass them, thus allowing students a “flavour” of research.

    Solutions

    And what of solutions? Some are relatively simple. Thus, it defies common sense that a common application form is not already in use. Part of the problem is the philosophy of pseudodevolution propounded by the centre, so that “decisions” must be taken locally. If research and development is taken as an example then the “purchaser levy,” which will be used to fund research and development, is just a politically correct form of topslicing. Surely, firm direction could come on this one issue after consultation. Indeed, the move towards a common application form for all NHS research and development grants for applications incorporates the principle of a common application form for ethics committees. In line with this, a series of accepted examples of consent forms could also be generated.

    The organisation of local research ethics committees also requires refinement. It has been suggested that they should be serviced by the regional offices of the Research and Development Directorate, which would remove them from the financial vagaries of commissioners. The directorate could also help in organising training, which obviously needs to be a continuing activity.

    Financial questions, such as “Can the trust afford this study?” should not be a problem for ethics committees. This point should be clarified before submission, and I would recommend some form of institutional review—perhaps by trusts' research and development offices—before applications go to local research ethics committees or potential funders. Close cooperation with industry should also clarify applications for drug trials, and a much more rigorous approach to financial inducement for investigators should also be adopted.

    The biggest problem, however, remains multicentre studies and the byzantine labyrinth that currently exists, as described by the papers in this issue11 12 13 and elsewhere.17 This has led to the establishment of a working party by the chief medical officer, which should report by the end of the year, as well as suggestions by the regional directors of research and development. A consensus has emerged that a coordinated approach is required, but the ethics committees should retain the ability to turn down trials on local grounds, although the rugged individuality of some committees requires tempering. The science and the overall ethics of a multicentre study—defined as one occurring in the areas covered by five or more ethics committees—would be looked at by a “central” committee.

    Much discussion has occurred on this latter point. The feeling is that a single committee could not handle the workload, and a possible solution is that each regional office of the Research and Development Directorate would house a regional ethics committee distinct from the local committees. These could handle applications when the lead investigator was in that region and applications from other regions if they were swamped. Meetings would need to occur no less frequently than once a month. Decisions would be faxed or emailed immediately to the applicant, who would then send the decision to the relevant local ethics committees together with a copy of the application. Details need to be agreed, but this seems a reasonable compromise, allowing proper assessment scientifically, ethically, and locally and diminishing greatly the time delays, workload, costs, and frustrations for all concerned.

    Proper reporting will also be necessary so that the processes are visible, defensible, and accountable. Hopefully, matters such as the common form, accountability, indemnity of lay members, and non-NHS research (which is not covered at all at present) will also be dealt with. It would be a great pity if the opportunity to tackle all the problems was lost and activity was restricted to multicentre research alone.

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.
    17. 17.
    View Abstract