Titles must be registered with the editorial base using a standard form (available at the Editorial Office in Leiden from the Review Group Co-ordinator Anja Helmerhorst), to prevent possible wasted time and duplication of effort if more than one person unknowingly embarks on the same review. The FRG encourages reviewers to collaborate with at least one other person, so that there is a mix of content and methodological expertise. In Leiden at the Editorial Office a database with collaborators and their interests is maintained. It is also advisable to involve people from more than one centre. Reviews must have a nominated contact reviewer who is responsible for liaising with the editorial team and other reviewers. All titles will be reviewed by all members of the editorial team and considered for their potential overlap, mixture of expertise, feasibility and suitability for a systematic review. This process should take no longer than two weeks. Once a title has been approved, it will appear in the module.
If more than one person proposes to do the same review the Review Group Co-ordinator will attempt to establish a collaboration between the parties and will arbitrate in the case of disagreements.
Protocols must be delivered within three months of a title being registered. Titles which have been registered for more than twelve months, for which protocols have not been forthcoming, are 'de-registered' and become generally available once more for review by other Group members.
The Group encourages reviewers to attend a workshop on 'Developing a protocol', before commencing work. Reviewers are also sent 'Practical guidelines for writing a protocol', when their title is accepted.
At least two editors/external referees evaluate the protocol using a refereeing checklist and a form for providing comment. They are asked to return completed assessments within four weeks. Comments are forwarded by the review group co-ordinator to the contact reviewer. Following any necessary revisions, and after approval by the refereeing editor, protocols are submitted to the CDSR. Once the protocol has been accepted, it must be sent to the editorial base electronically (in Revman).
Reviewers are aware of the identity and contact details of the editors who comment on their protocols and, in the event of a difference in opinion, are free to engage in discussion with them to establish a solution.
Copy-editing of protocols is done at the editorial base and reviewers sent are asked to approve the final version prior to publication on CDSR.
Reviews should be submitted within twelve months of the protocol being accepted for publication on CDSR.
Reviewers are sent a copy of 'Practical guidelines for writing a review', once their protocol has been accepted. Advice on specific problems/queries is available from the editorial base. The complete review is forwarded to the editorial team and evaluated by the same editors/external referees (minimum 2). External referees might be requested for advice when deemed necessary. The refereeing editor summarises the comments of the referees and outlines expected revisions. At this stage editors/external reviewers will randomly check a sample of the included trials to ensure that data have been extracted correctly and quality of the included /excluded studies has been assessed adequately. Following any necessary revisions and approval by the refereeing editor, the completed review is submitted to the CDSR. The editorial process to referee a review takes four weeks; the whole process from the formal submission of a review, feedback of comments, making revisions/addressing comments, to receiving overall editorial approval will take approximately twelve months. Reviewers are aware of the identity and contact details of the editors who comment on their reviews and in the event of a different opinion, are free to engage in discussion with them to establish a solution. Copy-editing of reviews is done at the editorial base and reviewers sent are asked to approve the final version prior to publication on CDSR.
Authors are advised of appropriate trials and are asked to examine reviews on bienual basis with a view to updating them. If no new trials have been identified, and no changes are desired by the editorial base or the reviewer, then the review will be updated as it stands. If changes are required, reviewers have nine months in which to update the review. Revised reviews are edited by a minimum of two editors and a statistician prior to re-publication. Reviews that include preliminary data from on-going trials are required to take subsequent data into account within nine months of the publication of the data. Revised reviews will be edited by two editors and a statistician as above.
Reviewers are free to update reviews more often than once per two years if they wish.
It is likely that, in the future, reviews will need to be updated in the light of comments received through the 'Comments and criticisms' facility on the Cochrane Library. It is expected that reviewers will be obliged to provide a response to criticism, and, if significant changes were made to the review as a result, the revised review would be subjected to the same editorial requirements as an updated review.