Personal tools
You are here: Home → IMSG → Minutes → SDG minutes, 5-6 April 2000
 

SDG minutes, 5-6 April 2000

Minutes of the Software Development Group meeting held at the UK Cochrane Centre on 5-6 April 2000

Minutes approved on 25 May 2000

Present: Chris Cates (had to leave before item 12), Mike Clarke (minutes), Monica Fischer (chair), Sonja Henderson, Julian Higgins, Peter Langhorne (6 April - item 9 onwards), Mark Starr.

Apologies: Peter Langhorne (5 April), Phil Wiffen.

In attendance: Rasmus Moustgaard (Nordic Cochrane Centre).

Contents:
Potential conflicts of interest
Correction and approval of the minutes of the special meeting
Correction and approval of the minutes of the previous ordinary meeting
Matters arising from the minutes of the 3 December 1999 meeting
Feedback from RevMan Advisory Group meeting on 5 April 2000
Official document defining the structure and technical contents of a Cochrane review
Content of a Cochrane review
Standards/procedures for testing of software
Alternative models to the next generation of the Information Management System
Contingency plan if problems arise after the release of MetaView 4.1
New comments and criticisms system
ModMan/ModMan Advisory Group issues
Any other business
Next meeting
Appendix 1 (relating to item 7)
Appendix 2 (relating to item 9)

  1. Potential conflicts of interest

    2. These are as before with the addition of the following:

    Julian Higgins - co-editor of the statistical section of the Cochrane Reviewers' Handbook. Sonja Henderson - employed at the editorial base of a very busy CRG.
    Chris Cates - convenor of the newly formed Criticism Management Group, which will also look at non-software aspects of the comments and criticisms system.

  2. Correction and approval of the minutes of the special meeting

    3. The minutes of the 3 February 2000 meeting were approved, subject to the correction of the spelling of Julian's surname in the conflict of interest section. The corrected minutes will now be published on the SDG Internet site.

  3. Correction and approval of the minutes of the previous ordinary meeting

    4. The revised minutes of the 3 December 1999 meeting were approved. They will now be published on the SDG Internet site.

  4. Matters arising from the minutes of the 3 December 1999 meeting

    1. Item 4f, Guidelines for old criticisms: these were on hold pending the discussions of the Criticisms Management Group.

      Action: Chris

    2. Item 4.7.b, CRG glossaries: a discussion paper prepared by Mike and Mark was discussed. The following additional options were suggested: the ability to toggle between a document and a listing of glossary terms; the use of a "right-click" to reveal the glossary term for the word on which the cursor is placed; the revealing of the glossary term if the cursor is paused over a word; the dynamic generation of a glossary of terms for any selected document; and the addition of a medical dictionary (supplemented with Cochrane terms as necessary) to The Cochrane Library. Mike will add these options to the discussion paper and circulate it, for comment, to Review Group Coordinators, Coordinating editors, the Cochrane Library Advisory Group and the Consumer Network. It was recognised that the proposal for a single glossary would require a central glossary editor. Mike and Mark will look at the feasibility of combining the current CRG glossaries.

      Action: Mike, Mark

    3. Item 10, web archive of minutes: this will be created. It will contain information additional to the minutes (eg the remit of the Groups).

    4. Item 12, proposal for additional graphs in Cochrane reviews: the "work in progress" discussion paper prepared by Mike and Jon Deeks and circulated at the RevMan Advisory Group meeting on 5 April was discussed briefly. Comments on this should be sent to Mike or Jon by 30 April. It will then be revised and circulated more widely for comment.

      Action: Mike

  5. Feedback from RevMan Advisory Group meeting on 5 April 2000

    1. Assessment of software needs for Cochrane reviews: It was agreed that an Action Group should be formed to design a survey assessing the software needs relevant to preparing a review. This should cover aspects of review production from the start of work by a reviewer to submission of the review to Update Software. The suggested membership of the group is Mike (convenor), Julian, Philippa Middleton, Paul Whatling, a review group coordinator (to be suggested by Sonja via the ModMan Advisory Group), a reviewer who uses a wordprocessor to prepare their review, a reviewer working in a developing country, and someone with experience of survey design. The piloting of the survey should involve reviewers whose first language is not English. A pilot should be completed in the next few months and the full survey should have started before the Colloquium (at which a workshop should be held). The Action Group should also advise on how to analyze the survey and to whom it should be distributed.

      2. Action: Mike, Sonja

    2. Release of RevMan 4.1: It was agreed that beta-testing of RevMan/MetaView 4.1 should be extended by three weeks. Update Software should be given a report of the findings of the beta testing by 1 May. Depending on the extent of the revisions needed, Mark hoped that it will be possible to complete the revisions in May. The statistical testing by the Nordic Cochrane Centre will then need to be repeated and there will be dedicated testing on a variety of computers by Rasmus and Mike. Mike will use the Update Software test bank to do this. It is hoped that it will be possible to release the software sometime during June-August.

      Action: Mike, Rasmus

    3. Content of a Cochrane review: discussed under item 7 below.

  6. Official document defining the structure and technical contents of a Cochrane review

    7. It was agreed that an official document should be produced that could be given to Update Software so as to define a Cochrane review (including its graphs). This should contain guidance on the presentation of reviews. The Nordic Cochrane Centre will prepare the document. It will be based on Appendix 2a of the Cochrane Reviewers' Handbook but will contain details on additional fields that are not described in the Handbook. The Publishing Policy Group (with advice from the Cochrane Library Advisory Group if they wish) will be asked to consider defining the style of a Cochrane review. Such a definition would lead to the file format that will be required by publishers of Cochrane reviews. The document should describe the requirements for links in Cochrane reviews.

    Action: Monica

  7. Content of a Cochrane review

    1. Content of a Cochrane review: The recommendations from the RevMan Advisory Group meeting on 5 April 2000 were discussed and agreed (appendix 1). These recommendations will be passed to the Cochrane Collaboration Steering Group for consideration. If approved the list will be made available to Review Group Coordinators and Coordinating editors and passed by Mike to Update Software for implementation as soon as possible - ideally from issue 3, 2000 of The Cochrane Library. Mike will also tell Update Software which elements should be published for Cochrane protocols and which should be regarded as mandatory. Where possible, items should be added to the RevMan and ModMan wishlists to check that something has been entered into those elements of Cochrane protocols and reviews that must be completed to allow publication.

      Action: Mike, Rasmus

    2. Date of search in abstract: Mike will check that the Cochrane Reviewers' Handbook advises reviewers to include the date of last search in the Search Strategy section of their Abstract.

      Action: Mike

  8. Standards/procedures for testing of software

    9. Phil had not been able to find any suitable guidance but will keep searching.

    Action: Phil

  9. Alternative models to the next generation of the Information Management System

    10. Mark's document was discussed, paragraph by paragraph. Paragraph 1: it was agreed that a document should be produced and circulated for comments. Paragraph 2: it was agreed that the new system should be document (ie review) based. Sonja will attend a workshop on document tracking at the UK Contributors' meeting and report back to the SDG. Paragraph 3: these principles were agreed and a generic question "Will this help to get Cochrane reviews done?" should be applied to all aspects of the Information Management System. Paragraph 4: this was agreed and it was noted that the Information Management System should facilitate checking of the quality of Cochrane reviews as well as allowing some checking to be done automatically. Paragraph 5: it was agreed that the preparation of Cochrane reviews should not be dependant on a single piece of software but it was recognised that the Collaboration will probably need to continue to produce a software tool that reviewers can use if they wish. Paragraph 6: Mark will give the Nordic Cochrane Centre copies of Cochrane reviews in XML for them to experiment with. Piloting will be done of moving XML reviews to and from The Cochrane Library. Update Software and Rasmus will agree on the specification for this. It was agreed that Mark's paper did not need to go onto the SDG Internet site but that the principles (paragraph 3) should do so (see appendix 2).

    Action: Sonja, Rasmus, Mark

  10. Contingency plan if problems arise after the release of MetaView 4.1

    1. Although Update Software will try to correct any problems with MetaView 4 after its live release, they will no longer develop it. If it is not possible to correct any serious problems, alternatives might be: an analysis/graphics package prepared by the Nordic Cochrane Centre and/or a new analysis package from Update Software. The package that Rasmus has been working on still needs testing but it will allow, for example, the printing of several graphs on the same page and the re-scaling of graphs. However, it was agreed not to package additional analysis/graphics packages with RevMan 4.1.

    2. RevMan 4.1 will be modified to make it possible to save the data file so that this could then be manipulated in other graphics/analysis packages.

      Action: Rasmus

    3. The availability of the Nordic Cochrane Centre software (which should not be seen as an official Cochrane Collaboration product) will be discussed at the next SDG meeting.

      Action: Monica

    4. During this discussion it was also noted that the Statistics Methods Group should be asked to define the data and analysis needed for the inverse-variance-weighted average method that was agreed at the SDG in December 1999.

      Action: Julian

  11. New comments and criticisms system

    12. A pilot system (similar to that used by the BMJ) will be tried from July until the Colloquium.

    Action: Chris

  12. ModMan/ModMan Advisory Group issues

    1. It was noted that this discussion was taking place in advance of the ModMan Advisory Group meeting on 7 April 2000.

    2. Tracking software: 75% of the Review Group Coordinators (RGCs) who had replied to a survey do not use the tracking function in ModMan. The feeling among RGCs seems to be that ModMan is not suitable for tracking reviews because, although it is good for chasing people, the report generation aspects are not sufficiently flexible. Many RGCs are trying to develop their own tracking systems. The ModMan Advisory Group will discuss how to move forward with this.

      Action: Sonja

    3. Relationship between RevMan and ModMan: the inability to use RevMan and ModMan at the same time was causing serious problems. It is slowing down the work at editorial bases where there is a need for multi user access to the software. It would be easier to manage the process at the editorial base as well as be 'safer' if they were separate. For example, it is possible to make unintentional changes to the Module submission by editing the review within RevMan without opening ModMan, including changing the version to be published. No warning is given in RevMan to tell the user that they are about to change details of a publication already approved in ModMan for publication. It would be preferable if reviews could be moved into ModMan for viewing and submission (not editing).

      Action: Sonja

    4. Development of ModMan: the Nordic Cochrane Centre would like to develop a new ModMan that would produce the same output as ModMan 4 and would be optional. This would require agreement with Update Software on the specification of the output file they need. Mark agreed that, in principle, Update Software would not object to receiving suitable submission files from sources other than ModMan 4. The ModMan Advisory Group will discuss functional changes to the software and suggest these to the SDG. If Update Software is unable to make the necessary changes, the SDG will suggest to the Steering Group that responsibility for the development of ModMan should be transferred to the Nordic Cochrane Centre. If accepted, Mike, in his role as Liaison between the Collaboration and Update Software, should then notify Update Software officially about the transfer and ask them to provide the file specification for ModMan to the Nordic Cochrane Centre. It was recognised that any major new changes would lead to a mandatory release of the software.

      Action: Sonja

    5. Contact details: Guidance is needed on whether RGCs should continue to enter contact details into ModMan if this information is not being used elsewhere, as would be the case if the proposal by Georg Koch is accepted. Monica will let Georg Koch know that his proposal should be sent to the Steering Group for discussion/approval.

      Action: Monica

  13. Any other business

    1. Expansion of the Software Development Group: A request has been received from someone wishing to become a co-opted member of the SDG to provide "independent" computer expertise. He has not replied to a request to supply information on what he would bring to the Group. If he does reply, he should be told that the SDG is currently trying to redefine its membership and role. This redefining should include the issue of co-opted members and full members to represent Cochrane entities.

      Action: Monica

    2. Comments/criticisms on Internet versions of The Cochrane Library: it was noted that it is possible to submit comments and criticisms through the Internet versions of The Cochrane Library but that the relevant button is not always labeled "Comments/criticisms".

    3. Minute taking: Monica will ask the Cochrane Collaboration Steering Group for secretarial support for minute taking for the Software Development Group and its Advisory Groups (RevMan, ModMan and Criticisms).

      Action: Monica

  14. Next meeting

    1. Special meeting: There will be a special meeting of the SDG on 30 May (provisionally at Heathrow airport) to discuss the Information Management System. Monica will tell Fenella Rouse to expect input from the SDG for the Cochrane Collaboration's Business Plan after this meeting.

      Action: Monica

    2. Regular meeting: Monica will be on maternity leave from August 2000 to February 2001. The next meeting of the SDG will need to take place during this time. It will be in either November or December 2000, attached to the (to be arranged) RevMan Advisory Group meeting.

      Action: Mike

  15. Appendix 1 (relating to item 7)

    16. RECOMMENDATION FROM THE REVMAN ADVISORY GROUP AND SOFTWARE DEVELOPMENT GROUP (6 April 2000)

    The following elements of a Cochrane review in RevMan should be published when the review appears in The Cochrane Library and reviewers and Review Groups should try to ensure that all these sections have been completed. If any of the sections marked with a * below are empty, the review should not be published on The Cochrane Library until something has been added to the section. The Collaboration's Advisory Group on Quality will be asked to make recommendations on how the sections marked with [Q] should be shown on The Cochrane Library.

    *Title of review
    *Name of contact reviewer
    *Contact details of contact reviewer
    *List of reviewers for citation
    Contributions
    Sources of support - intramural [Q]
    Sources of support - extramural [Q]
    What's new - text
    What's new - issue protocol first published
    What's new - issue review first published
    *What's new - date of last substantive update
    What's new - date new studies sought but none found [Q]
    What's new - date new studies found but not yet included/excluded [Q]
    What's new - date new studies found and included or excluded [Q]
    What's new - date reviewers' conclusions section amended
    Notes - published
    Synopsis [Q]
    *Abstract
    *Text of Review
    Acknowledgements
    Conflicts of interest
    References to studies - included studies
    References to studies - excluded studies
    References to studies - studies awaiting assessment
    References to studies - ongoing studies
    Other references - additional references
    Other references - other published versions of this review
    Tables - characteristics of included studies
    Tables - characteristics of excluded studies
    Tables - characteristics of ongoing studies
    Comparisons and data
    Meta-analysis graphs
    Additional tables
    Comments and criticisms - title
    Comments and criticisms - summary
    Comments and criticisms - reply
    Comments and criticisms - contributors

  16. Appendix 2 (relating to item 9)

    17. PRINCIPLES FOR THE DEVELOPMENT OF THE COCHRANE INFORMATION MANAGEMENT SYSTEM

    1. The primary objectives should be:

    a. To facilitate production of Cochrane reviews by optimising use of resources and supporting reviewers.
    b. To ensure the scientific quality of the resulting reviews.

    2. The following generic question should be applied to all aspects of the Information Management System "Will this help to get Cochrane reviews done?".

IMSG minutes | MinMan home

Document Actions