Introduction: Previous studies have highlighted the need to improve the quality and readability of abstracts of Cochrane reviews. As a result, more detailed guidelines for writing Cochrane abstracts were developed during October and November 1998. . Three hundred abstracts of the 522 existing Cochrane reviews were then assessed against these guidelines by a small team from the Australasian and Nordic Cochrane Centres. In general, the results section required the most editing, followed by the conclusions and search strategy sections. The difficulties of communicating uncertainty in an abstract and presenting results succinctly were most noticeable when reviewers encountered methodological and power limitations in their reviews.

Objectives:

Methods: Following this editing process, we established a prospective study to compare the abstracts from Cochrane Library 1,1999 (before central editing) and Cochrane Library 2, 1999. The study involved an assessment of:

-- How many centralised edits were accepted in part or full by review groups. Any patterns in modifications made by review groups.
-- Any changes in quantitative or qualitative measures of abstract quality (whether centrally edited or review group edited).

Results:

Discussion: Based on this study recommendations will be made about ways to write and edit abstracts of Cochrane reviews in the future, considering the resources available centrally and within review groups.