When the Department of Health finally produced guidelines in 1991⁽¹⁾ on local research ethics committees (LRECs), paragraph 2.16 of the guidance was:

The Department required DHAs to have implemented the guidelines by the end of March 1992, so that those committees not already producing annual reports should have started to do so by April, 1993. In the summer of 1993 the Bulletin decided to undertake a study of LREC annual reports, and obtained generous funding from the Nuffield Foundation to do so. The objectives of the study were:

Within the above objectives, other issues would also be addressed. The editor of the Bulletin carried out the largest survey yet of research ethics committees in 1983(2). It had shown immense variations in the membership, workloads and working practices of research ethics committees. With the results of later studies, particularly that of Julia Neuberger(3), it had provided the basis for many requests to the Department of Health and other authorities to try to make the performance of LRECs more consistent and coherent. The Royal College of Physicians of London, and the Association of the British Pharmaceutical Industry, in particular, have in the last decade issued series of guidelines designed to assist LRECs in such an aim. The present study therefore provided an opportunity to see whether LRECs still showed so much variation.

In August 1993, a letter and brief questionnaire was sent to all identifiable LRECs in the United Kingdom. Various lists of LRECs, from the pharmaceutical industry, were combined with an updated version of the list held by the Bulletin since the 1983 study. Responses to the first mailing also indicated changes in the structure or existence of committees, or in their addresses. Eventually 272 possible committees were sent questionnaires and requests for annual reports. The responses have been:

Overall, 209 committees contacted the Bulletin out of a maximum possible 255: the response rate was therefore at least 82%, the exact figure depending on how many of the non-respondents still exist. That some of them do so is shown by the fact that at least six undergraduate teaching hospitals are included in the "no response" group.

Of the 270 reports received, four were for years prior to 1991, and have not been included in the analysis. The actual periods for which annual reports are written are in some cases the calendar year, in some cases the academic year, and in a small majority of cases the financial year. For ease of analysis the reports have been assigned to the year between 1991 and 1995 in which fell the majority of the period reported on. This applied also to the handful of reports for periods longer or shorter than 12 months, ranging from 8 to 15 months. The numbers of reports available for analysis were:

The crudest measure of whether committees adhere to the Department of Health guidelines is whether or not they include in their annual reports the information listed in the quotation above. In other words, do they give the names of their committee members, do they list the number of meetings held and do they give a list of proposals considered with the outcomes of such consideration? There is obviously room for debate as to what constitutes a list of proposals, since different committees provide very different sorts of lists. Some give very abbreviated titles - sometimes just the name of a drug or a manufacturer only - while others give only the committee's reference number. Some give the name of the proposer, others do not.

For the purposes of this analysis, a list of proposals includes, as a minimum, the name of the proposer, a reasonably full title of the project, and the LREC decision - approved, approved after amendment, rejected or withdrawn. The numbers and percentages of committees each year that gave a list of proposals, the names of committee members and the number of meetings held were:

1992	19 (41%)
1992	64 (47%)
1993	14 (38%)
1994	20 (54%)
1995	4 (40%)

Overall 45.5% of reports met the DoH requirement for information in the annual report.

Another indicator of adherence with the DoH guidelines would be whether the overall size and type of membership of LRECs conforms to that recommended. Thus committees should have between 8 and 12 members overall, with two or more lay persons, one of whom "should be unconnected professionally with healthcare and be neither an employee nor advisor of any NHS body". If LRECs are taking serious note of the DoH guidelines, one would expect more committees to conform to them in the years after 1991. The first table shows how many LRECs were too small (less than 8 members) or too large (more than 12 members) each year, while the second table shows how many failed to have adequate lay membership.

Adherence to many of the other DoH guidelines is more difficult to assess because there is no requirement that the necessary information be included in the LREC's annual report. Individual reports may, however, give interesting insights. Thus the guidance states: "Despite being drawn from groups identified with particular interests or responsibilities in connection with health issues, LREC members are not in any way the representatives of those groups". One annual report, however, bemoaned the lack of consistent attendance of either a nurse or a public health specialist. The relevant directors of nursing and public health had been asked to nominate representatives, and each had sent three different representatives in the course of a year.

The guidelines go on to say that "DHAs should consult local professional advisory committees and relevant health professional associations". One committee has taken this to the extreme, and now had 13 separate official bodies nominating the 18 members of the committee.

Very few annual reports give any indication of when the committee members were appointed or of when their terms of office end. So there is no way to measure compliance with the requirement that no more than two terms of three to five years be served. Several names of long standing committee members were recognised however, one of whom has been on a committee for nearly 30 years.

Compliance with the advice on working procedures and keeping a register could only be addressed accurately by an on-the-spot audit. Yet there are plenty of examples in the annual reports of good, clear, standing orders. Moreover, there is very little evidence that any significant amount of business is conducted by post or telephone, a problem that was still substantial at the time of the 1983 survey(2), when 11% of committees conducted their business without meeting.

Following up approved research proposals has always been an area of considerable difficulty, which is probably why the DoH guidelines are rather vague on the point, requiring only that significant deviations from the agreed protocol, or matters affecting the safety of a research project, should be reported to the committee. Of the 173 committees from whom reports were received, just 12 provided substantial evidence of knowing what progress, or lack of it, each approved research project had made. That evidence will be considered in more detail later.

One final issue of conformity to the guidelines concerns confidentiality. Paragraph 2.15 reads:

"LREC members do not sit on the committee in any representative capacity and need to be able to discuss the proposals which come before them freely. For these reasons LREC meetings will normally be private and the minutes taken will be confidential to the committee".

Obviously this guidance is open to debate when one main purpose of RECs is to protect the public - a protection that needs to be seen to be provided - and elsewhere, in the United States for instance, equivalent meetings are open to the press and the public. Given that the advice exists, however, the Bulletin was surprised to receive minutes of meetings from six committees, five of which had not provided an annual report. Only one set of minutes contained sensitive information, however, in that it recorded considerable discussion of a complaint made by a research subject.

The first conclusion is therefore that there is scant evidence of LRECs making any substantial effort to conform precisely to the DoH guidance. The 1995 annual reports are no better at providing the basic information central to a good report than those of 1991 with, overall, fewer than half of the committee reports meeting the DoH standard. In the case of committee size, there has been a slight increase from 1991 to 1995 in the proportion of committees whose total membership lies outside that recommended. Yet at the same time all committees now appear to have fallen into line with the recommendation on the minimum number of lay members. What perhaps we are seeing is a significant number of LRECs exhibiting the independence that is crucial to their work, and regarding the DoH guidance not as rules but just as guidelines.

For several years there has been a regular succession of papers in UK medical journals purporting to show how slow and idiosyncratic LRECs have been in their consideration of multi-centre research projects. These papers have several things in common. None report studies in which a researcher has deliberately set out to discover how well LRECs handle multi-centre trials (MCTs): invariably they are anecdotal reports arising because a researcher has had difficulty obtaining the approvals needed for a clinical MCT. Likewise none of the papers provide enough detail about the delayed protocols to allow readers to judge for themselves whether there may have been serious or difficult ethical problems. Indeed the results to be presented below suggest that several of these MCTs will have had inadequate patient information sheets. A final common point is that none of the authors consider

a) that anything might have been wrong with their MCT;
b) what sort of workload the LRECs were facing; or
c) whether LRECs are adequately supported for the amounts of voluntary work they do.

What has not appeared in the journals is any accurate assessment of LREC workload and how it is changing. Any committee member can tell one that the load is continually increasing but only occasionally do anecdotal figures appear in confirmation. The second element in this present study was therefore an attempt to describe accurately LREC workloads in the first half of the 1990s, and to compare them with those of earlier years.

The 1983 survey(2) was the first to examine UK LREC workloads in detail, although it concentrated on how the committees handled proposals for research with children. It did, however, ask how many research proposals each LREC had reviewed in total in the years 1981 and 1982. The answer, from 174 committees, was a mean of 41 proposals in the two-year period, with a range of 0 to 346.

In the present study, four committees provided evidence of their workload annually from 1982 to 1992: none were in districts with designated teaching hospitals. The following table of the mean workload of these four committees is likely therefore to be reasonably representative of districts with general hospitals that have not recently had large increases in the numbers of student projects to be reviewed.

The figure for 1982 may be a little low in comparison with the mean of 20.5 found in the larger (1983) survey. Nevertheless these four committees provide strong evidence of a steady and inexorable rise in the workloads of LRECs during the 1980s, such that there was at least a fourfold rise between 1982 and 1992.

The evidence of the main bulk of committees in this study for the years 1991 to 1995 confirms the above picture, and shows further workload increases to 1995. The table below gives the mean number of proposals, and the range, for each year: the distribution is far from being a normal distribution and so no figures are given for the standard deviation.

It might be argued that the figure for 1995 is unrealistically high, and based on too small a sample. The same 10 committees reviewed an average of 111 proposals in 1994, however, which is only 4 above the average for that year. So it is likely that the true figure for the mean number of proposals seen by LRECs in 1995 was quite close to 120.

The overall figure for the mean number of proposals conceals, however, the distribution of the workload between committees. The table below shows that two-thirds of committees in 1993-5 reviewed less than 100 proposals each year, i.e. less than the mean for any of those three years. The mean is heavily weighted by, predominantly, the teaching hospitals: of the 84 annual reports received for 1993, 1994 and 1995, 12 teaching hospital reports showed annual totals of proposals reviewed that were greater than 250.