- Julian Savulescu, Sir Robert Menzies medical scholara,
- Iain Chalmers, director, UK Cochrane Centreb,
- Jennifer Blunt, former chair, Salford research ethics committeec
- a Clinical Ethics Project, Oxford Radcliffe Hospitals, Oxford OX3 9RP
- b NHS Research and Development Programme, Oxford OX2 7LG
- c Salford and Trafford Health Authority, Eccles M30 0NJ
- Accepted 19 September 1996
The results of recent empirical investigations in research synthesis imply that research ethics committees are behaving unethically by endorsing new research which is unnecessary and by acquiescing in biased under-reporting of research which they have approved. The performance and accountability of research ethics committees would be improved if they required those proposing research to present systematic reviews of relevant previous research in support of their applications; to summarise the results of these reviews in the information prepared for potential participants; to register new controlled trials at inception; and to ensure that the results of these trials are made publicly available within a reasonable period of time after completion of data collection.
Properly designed, conducted, and reported research is essential to distinguish useful from useless or harmful forms of health care.1 Research ethics committees exist to ensure, firstly, that proposed research will not expose participants to unacceptable risks and practices; and, secondly, that the potential participants can evaluate the expected consequences of their involvement and decide for themselves whether to participate.2 3 Research ethics committees have a wider responsibility to promote the public interest by helping to ensure that relevant research is done.3 In this article we suggest that the results of recent research in a new and rapidly expanding sphere of empirical investigation—research synthesis4—has important implications for the work of research ethics committees.
Research synthesis is the aggregation and integration of the results of related primary studies with the purpose of drawing conclusions from the totality of the relevant evidence. As the body of primary research evidence expands, research synthesis has become essential; but there is an emerging recognition that the quality of most research syntheses leaves much to be desired.5 6 Research syntheses of poor quality can lead to beneficial effects of treatments …