Revisiting the Cochrane Collaboration

It is surely a great criticism of our profession that we have not organised a critical summary, by specialty or subspecialty, updated periodically, of all relevant randomised controlled trials.

Archie Cochrane, 1979¹

In 1992, Iain Chalmers and colleagues wrote an editorial in the BMJ that began with the above quotation and set out challenges foreseen at the time of the opening of what became the first Cochrane Centre, in Oxford.² Some of these challenges have been met, some remain, and new ones have arisen as the centre which is now the Cochrane Collaboration strives to prepare and keep up to date systematic reviews of the effects of healthcare interventions.

At the time, the major hurdle facing reviewers was identifying relevant randomised trials, and the editorial described efforts to make the task easier. Nine years on, these efforts have contributed to the Cochrane controlled trials register, which now contains more than 300 000 records. Much work, however, remains to be done—for example, identifying trials in conference proceedings, arranging translations where needed of trials under review, or getting additional data from the trialists. One development, which should make reviewing easier, is a system to register prospectively all randomised trials and assign unique identifiers to each.³

The efforts to get more information on randomised trials is only part of what is needed to make informed decisions about health care. Unless trials are placed in the context of other relevant research, they exist as islands of information.⁴ Critical summaries are needed. The collaboration is tackling this by preparing and maintaining systematic reviews. Though these reviews rely on randomised trials, they are not restricted to such studies. One challenge facing reviewers is how to identify and incorporate data on rare adverse events not usually available from randomised trials.

There are currently over 6000 people in over 60 countries working for the collaboration. Much of this work is unpaid but vital financial support has come from governments and other funders. Ensuring the sustainability of existing funding and getting more funding for growth are major challenges because without growth the collaboration will probably wither and die.

The Cochrane database of systematic reviews in the Cochrane Library contains over 1000 reviews and 800 protocols for reviews in progress. Several hundred reviews and protocols are added each year. There is, however, a long way to go before all areas of health care are covered, and the number of reviews needed to do this is likely to be several times the number currently available. Achieving this level of coverage while keeping all reviews up to date is perhaps the biggest challenge in the years to come.

Cochrane reviews have already provided a substantial amount of evidence relevant to health care. They are relevant to a wide range of policies, and are widely cited in guidelines.⁵ Their importance was recognised in a recent report from the American Institute of Medicine.⁶ Journals such as the BMJ now recommend that their editorialists should refer to relevant Cochrane reviews where these exist. The Campbell Collaboration now exists as a sibling for interventions relating to criminology, social policy, and education.

The past decade has also seen an increasing recognition of the importance of systematic reviews as a means of assessing the effects of health care. Systematic reviews are seen as research that should precede and conclude the design and conduct of all new studies. The recent document on research governance in England has emphasised the need for reviews in the planning of research in health and social care.⁷ Organisations such as the UK Medical Research Council, the National Health and Medical Research Council in Australia, the European Science Foundation, the National Institutes of Health in the United States, and several ethics committees require evidence from a prior systematic review when considering proposals for clinical trials or planning conferences.⁸,⁹

But the search continues for ways to meet the practical demands of maintaining reviews as the living documents they should be. There is still some way to go before Archie Cochrane's challenge will have been met in full—it might never be. The challenges now are to ensure that processes which are in place, and the goals of the collaboration become sustainable and achievable.

See also Papers p 829