Ethics approval for a national postal survey: recent experienceBMJ 1995; 311 doi: http://dx.doi.org/10.1136/bmj.311.7006.659 (Published 09 September 1995) Cite this as: BMJ 1995;311:659
- Claire Middle, survey researchera,
- Ann Johnson, developmental paediatriciana,
- Tracey Petty, administratora,
- Lois Sims, programmera,
- Alison Macfarlane, medical statisticiana
- Correspondence to: Dr Macfarlane.
- Accepted 29 June 1995
Before any study can be done on a national scale, permission must be sought from a large number of local ethics committees. We describe the resources required and the problems and issues that arose when we sought approval from ethics committee for a national postal survey in England and Wales.
Methods and results
We wanted to mount a postal survey in all parts of England and Wales of a national sample of children who were born in 1988 and were stratified by birth weight. After identifying the children from birth registration and tracing them through the NHS central register, we planned to send self administered questionnaires to the children's parents, general practitioners, and teachers. We identified 162 local research ethics committees in England and Wales from the Medical Research Ethics Committees Directory.1 Each was sent a letter outlining the project and asking how an application should be made.
Seventeen committees told us that it was not appropriate to apply to them for ethics approval for this study; the remainder needed some type of formal application. Thirteen committees did not have a specific application form but requested between one copy and 21 copies of the protocol. Of the 132 committees that used an application form, 118 had a unique form, ranging in length from two to 18 pages. In the former Northern region a single application was considered by a “lead” committee on behalf of the seven others. In the former South Western region an application followed a standard format but each committeeconsidered it separately. Seventy six committees requested 10 or more copies of the completed application form and protocols.
In all, we sent a total of 1095 protocols and 1116 application forms, together with a number of supporting documents. We estimated that, in total, this required seven to eight weeks of staff time. The cost of staff time, photocopying, and postage was estimated to be £4606, an average of £32 for each committee. This does not include the costs and time spent on additional telephone calls to clarify issues, responses to requests for additional information, and setting up and maintaining a database to record details of correspondence.
Three months after submitting applications to the 145 committees, 113 had responded. Eighty two expressed no objection to the study. Others raised a number of issues, including concerns about the aims of the study, its cost, confidentiality, consent, and the wording of the questionnaires and information sheets. Thirty one requested resubmission.
Seeking approval for a study covering a large geographical area is time consuming and expensive for both researchers and ethics committees. Researchers have to obtain ethics approval before they can seek funding for the project. The location and composition of committees linked to commissioning authorities and of those set up by provider units are constantly changing. There may be additional costs in the future if committees levy a charge for considering applications.
These issues, together with the concerns that, despite guidelines,2 local committees differ widely in their response to applications,3 have led to a debate on the role of a national ethics committee in multicentre clinical trials and national surveys.4 5 We suggest that there may be complementary roles for a central committee and local committees. A national committee could be asked to consider the scientific and methodological aspects of the study before theapplication is submitted to the funding body. A single copy of the application and the national committee's recommendations could then be sent to local committees, whose unique understanding and knowledge of their populations and local factors could inform their decision whether to review the project protocol further using a standard application form available in electronic form. This arrangement would reduce the time and cost for both local committees and researchers, without jeopardising the interests of research subjects.
Funding AJ and AM were funded by the Department of Health. CM, TP, and LS were funded by Action Research to test the feasibility of the survey.
Conflict of interest None.
(c)Crown Copyright 1995.