Introduction: The pace of development of the Cochrane Collaboration has led to a considerable increase in reviews and protocols published on the Cochrane Library. Little attention has been paid to how topics for reviews are decided. Given the absence of intramural funding for the Collaboration and the large proportion of reviewers with other commitments, tha topic of reviews should receive more attention given the potentially high opportunity costs of such decisions. We report the results of a study aimed at describing current mechanisms of assigning topics within CRGs.

Methods: Electronic survey with one reminder of 46 editorial bases of Cochrane Collaborative Groups registered as at 1 February 1999.
Results: Thirty CRGs responded (65% response rate). Two CRGs had carried out no reviews at the time of writing and could not answer the question. The results are summarised in the table.

What percentage of the topics of your last ten reviews were decided on the basis of:

Median 25thpercentile 75th percentile
Editorial base decision? 5% 0% 30%
Availability of evidence? 0% 0% 33%
Reviewers' suggestions? 70% 40% 100%
Funders' direction? 0% 0% 20%
A mixture of all of these? 0% 0% 20%
Other criteria? 0% 0% 20%

9 CRGs (30% of respondents) declared being in favour and 19 (70 or respondents) against prioritisation. Of the 9 respondents favourable to prioritisation, 2 (22%) indicated the capacity of a particular topic to yield one or more reviewers' ideas and 4 (44%) importance of topic as criteria for prioritisation. Of the 19 respondents not favourable to prioritisation, 9 (69%) indicated a threat to the current bottom-up approach and the rest diverse reasons for their opposition.

Discussion: Several CRGs commented on the possible acceptability of Collaboration-wide criteria for prioritisation, provided their application within each area of remit (e.g. peripheral vascular disease) are left to relevant CRGs. The Cochrane Collaboration relies on individual's enthusiasm to carry out its mission.