Multicentre research ethics committees: has the cure been worse than the disease?

No, but idiosyncracies and obstructions to good research must be removed

Papers pp 1179, 1182, 1183 Personal view p 1217

I first wrote about the byzantine labyrinth that surrounded obtaining ethics committees' approval for multicentre studies in England in 1995, as well as mentioning other unsatisfactory aspects of local research ethics committees.1 At that time a working party of the chief medical officer suggested the establishment of multicentre research ethics committees on a regional basis to take care of multicentre studies. These were established in 1997.2 So is it now simpler to obtain approval for multicentre studies? Are decisions reached more speedily? Are local research ethics committees restricting their comments on multicentre studies to local problems?Or has yet another layer of bureaucracy been added, making the process even more labyrinthine?

In the past two years frustrated research workers have regularly told me that the new system is a disaster. Early feedback suggested that local research ethics committees were finding their subordinate role difficult. These committees have always jealously guarded their independence. The early problems led to further guidance from the Department of Health and NHS Executive on the precise responsibility of local research ethics committees for these multicentre applications. Most importantly, the guidelines stressed the need for speed (a response within three weeks) and that objections should be based solely on local issues.

Has this worked? Two papers in this week's BMJ are highly revealing (pp 1179, 1182).3 4 Both look at the fate of a multicentre study submitted to and approved by the appropriate multicentre research ethics committee. The study of Tully et al is the larger, involving 125 local research ethics committees. One response of these committees in general has been to establish executive subcommittees to deal in timely fashion with applications to multicentre research ethics committees. In Tully's experience this did shorten the time taken to respond, although less than a third of all local committees did so within the 21 days allowed, with a median time of 41 days.3 After six months, Tully's study was still not approved by nine of the local committees. More worrying perhaps was that about half these committees asked for amendments, and two thirds of these concerned non-local issues—expressly against the Department of Health's guidance. Lux et al had a similar experience involving 99 local research ethics committees, with only a third replying within three weeks.4 Some problems remained unsolved six months later. However, they did find, like Tully et al, that fast track subcommittees did speed up the process.3 Al-Shahi and Warlow had a similar experience with a Scottish multicentre research ethics committee.5 There the median delay to review was 28 days. The time taken for approval was 39 days, with a range of 21 to 109 days. They found only three objections, although one of these was not a local issue. The other major problem identified was the vast amount of paper involved—26.9 kg in one case5 and over 100 000 sheets of paper in another.3

None of these studies looked critically at the workings of the multicentre research ethics committees themselves. There is one in each English region, and one each in Scotland and Wales. Their decisions are, however, binding for the whole of the United Kingdom. Any application involving five or more local research ethics committees goes first to the multicentre committee. So far, most of the problems seem to surround the interface between the multicentre committee and local committees—both in terms of time and individual idiosyncrasy.

So what can be done? Holley and Foster found generally high standards of practice in 27 local research ethics committees in the South Thames region, and there is little reason to suspect that the situation is different elsewhere.6 They concluded that the problems researchers have with multicentre research are structural and logistic, and not due to substandard working of local research ethics committees.7 They also noted a steady improvement over time.

Some relatively simple measures would help solve the problem. Ah-See et al noted just two years ago that 15 out of 19 local research ethics committees approached had unique application forms.8 Surely a single form with a small number of variants should be used nationwide.8 9 Certainly a common form for multicentre research ethics committees is essential. A short form containing locally relevant information could be devised and sent electronically to local ethics committees, avoiding the need to send vast piles of papers.

There remains the problem of different modes of working and standards, which are occasionally highly idiosyncratic, between different local research ethics committees. A national advisory body is clearly needed to communicate regularly with all local committees, organise training programmes, and lay down clear guidance that is updated regularly. In return, local committees need better support, and their members need reasonable payment for what is often an onerous task. The guidance on handling multicentre proposals needs major reinforcement.

Research ethics committes have two major functions. On the one hand, they must protect patients and the public against harm from research—and against useless studies, which are unethical. They perform this function well, although at times in irritatingly nitpicking detail. On the other hand, they should encourage research that will in the long run improve health care and health. Here the system is still too obstructive. So have multicentre research ethics committees worked?The answer must be a qualified yes, but further improvement is needed if we are to continue to perform timely and valuable multicentre research in the United Kingdom.