Personal tools
You are here: Home → IMSG → Minutes → IMSG minutes, 12 Sept. 2001
 

IMSG minutes, 12 Sept. 2001

Minutes approved: 14 December 2001

Present:Gerd Antes, Chris Cates, Monica Fischer (convenor), Sonja Henderson, Julian Higgins, Rasmus Moustgaard, Mark Starr, Phil Wiffen.

In attendance: Jacob Riis (minutes).

Apologies: Mike Clarke.

Contents:
Welcome and apologies
Approval of agenda
Potential conflict of interest
Approval of minutes and matters arising from the minutes of the previous meeting (28 March 2001)
Update on needs assessment survey
Recruitment of new members
Graphs in Cochrane Reviews
Bi-annual report to the CCSG
RevMan 4.1 multi-user
Release of ModMan 4.1
Update on Contact Information Pilot
The Cochrane Collaboration's web site
Any other business
Date of next meeting

  1. Welcome and apologies

    2. Mike was unable to attend the meeting in person, but it had been arranged that he would join via telephone at 15.00.

  2. Approval of agenda

    3. The agenda was approved.

  3. Potential conflict of interest

    4. Changes since last meeting:
    Chris: the Criticism Software Advisory Group no longer exists. The CMAG also considers issues relevant to software development

    • Gerd Antes: Representing the Steering Group. No potential conflicts known, but possibly in relation to the German Cochrane Centre's responsibility for managing the Collaboration web site.
    • Chris Cates: Convenor of the Criticism Management Advisory Group.
    • Mike Clarke: Convenor of the RevMan Advisory Group. Representing the Steering Group. Editor of the Reviewers' Handbook.
    • Monica Fischer: Convenor of the IMSG. Co-ordinator of software development at the Nordic Cochrane Centre.
    • Sonja Henderson: Convenor of the ModMan Advisory Group. Review Group Co-ordinator of the largest review group which desperately requires a tracking system.
    • Julian Higgins: Representing the Statistical Methods Group. Co-editor of the statistical section of the Reviewers' Handbook.
    • Rasmus Moustgaard: Technical Advisor. Ex-officio. Working on software development at the Nordic Cochrane Centre.
    • Mark Starr: Representing Update Software. Involvement in the development of commercial software, e.g. Metaxis, and software for the Cochrane Collaboration as publisher of The Cochrane Library.
    • Phil Wiffen: Representing co-ordinating editors. No potential conflicts known.

  4. Approval of minutes and matters arising from the minutes of the previous meeting (28 March 2001)

    5. The minutes of the previous meeting were approved.

    1. Matters arisen from minutes of meeting on 7 December 2000

      1. XML coding (item 4.a)

        2. Rasmus and Lois have communicated about this, but have agreed to wait for the needs assessment results.

      2. Importing reviews from RevMan 4.1 into CLIB (item 4.b)

        3. The problem is that the RTF codes used by the text control in RevMan may change when new Microsoft software is installed. Rasmus and Lois will continue to monitor this.

      3. Review Group glossaries (item 4.c)

        4. Sonja reported that the module guidelines have been updated with new guidelines for including glossaries in CRG modules. The need for a central set of glossaries was discussed, and Mark reported that Update Software are looking at a single commercial solution from X-refer.

        It was discussed whether the issue of glossaries fall within remit of IMSG. Since they may be relevant for other language versions of reviews and potential comments and criticisms in other languages, development of future systems may need to take them into account.

      4. MetaView bugs on web site (item 4.e)

        5. Rasmus reported that Update Software's web site for MetaView did not list all the bugs that he had submitted to Gordon. Matthew Gilbert and Phillip Basset are responsible for tracking bugs and updating web sites, so Rasmus should resend the reports to Phillip Basset.
        Action: Rasmus.

      5. Expenses claim forms (item 4.f)

        6. Monica will send out the forms again after the meeting.
        Action: Monica.

    2. Feedback from Criticism Management Advisory Group (item 5)

      1. ComCrits FAQ

        2. The document 'The Comments and Criticisms system - some questions and answers' is being updated. It was agreed that new version should be published on the Collaboration's web site and followed by an e-mail to relevant mailing lists, including: ccinfo, comcrits, adminors, centres and allcoeds.
        Action: Chris

      2. Problems with notification

        3. There had been problems with notification of comments not being sent to the criticisms editors, but the comments do appear on the web site. Chris will add to the FAQ that the web site should be checked regularly.
        Action: Chris

        Mark reminded the group that request for changes to the Comments and Criticisms system maintained by Update Software should go through the official channels: Mike Clarke, in his capacity as Liaison Officer between the Cochrane Collaboration and Update Software.

      3. Comment for first publication of a review

        4. Mark reported that a new review that already has a comment attached will appear in issue 4, 2001 of CLIB. It was discussed whether this should be allowed. It was agreed that the IMSG should recommend to the Handbook Group that notes from the Editorial Team should should not be in the form of comments, but that situations may arise where individual peer reviewers' comments could be included.
        Action: Monica

        This issue should also go on the agenda for the CMAG.
        Action: Chris

    3. Access to data manipulation in published graphs (item 6.c)

      4. The PPG awaits a recommendation from the IMSG on this issue. See item 7: Graphs in Cochrane Reviews

    4. MAG matters: review tracking software (item 9)

      1. Next MAG meeting

        2. The MAG will meet on 27 September, where tracking will be on the agenda.

      2. Software specifications

        3. Sonja emphasized that the working paper on tracking software specifications should reflect good practice. The role of the Quality Improvement Manager was discussed, see item 13.b

      3. Identifying existing software

        4. Phil had not been able to find a person that could identify existing software suitable for tracking. Nor had it been possible to recruit a student from Loughborugh. Two people at the Norwegian branch of the NCC have expressed an interest and have become tentatively involved.

        Update Software have begun using 'Starteam' software for tracking their 'documents', but it is an expensive tool. Mark invited interested parties to visit Update Software for a demonstration.

  5. Update on needs assessment survey

    1. Status

      2. A brief update on the status of the survey was included as item 4.c of the draft of the Bi-annual report to the Steering Group, which had been circulated.

    2. Editorial bases

      3. Sonja felt that the response rates for editorial bases might be low, and it was agreed that Mike should request responses from those that have not replied yet.
      Action: Mike.

      It was agreed that Sonja and Phil should put the survey on the agendas for the CRG and Co-eds meetings in Lyon. Julian agreed to do a simple summary of results, based on Sonja and Phil's input on what kind of information they will need.
      Action: Sonja, Phil, Julian.

    3. Need for results

      4. Rasmus stressed the usefulness of getting preliminary results soon, as many projects are on hold in the meantime.

  6. Recruitment of new members

    1. Potential conflicts of interest

      2. Before discussing the applications, an extra round of declarations of potential conflict of interest with respect to the individual applicants was carried out. Only those who declared a conflict are listed below.

      Member representing developing countries:
      Julian Higgins: works with one of the applicants on several reviews.

      Convenor of Technical Implementation Advisory Group:
      Monica Fischer: Works closely with Rasmus Moustgaard.
      Rasmus Moustgaard: being an applicant, Rasmus left the room when the position was discussed.
      Gerd Antes: collaborates with one of the applicants.

    2. Applicants chosen

      3. All received applications were strong, and it took lengthy discussion to reach an agreement. The actual discussion was not minuted.

      Decisions were made by majority vote when Mike joined the meeting.

      1. Member representing developing countries

        2. It was agreed to ask Elisabeth Pienaar to join the IMSG.

      2. Convenor of Technical Implementation Advisory Group

        3. It was agreed to ask Rasmus Moustgaard to convene the Group.

        Georg Koch and Daniel Spinner were suggested as group members.

        Once approval of the new members have been obtained by the Executive Steering Group, Monica will write to the applicants about the outcome.

        Action: Monica.

    3. The Contact Advisory Group

      4. Prompted by the fact that nobody applied for the position as convenor, the need for the group was revisited, and it was decided that an ad-hoc Contact Management task group should be formed instead.

  7. Graphs in Cochrane Reviews

    1. SMG guidelines for graphs

      2. Julian circulated drafts of the guidelines for graphs in reviews he is preparing for the Statistical Methods Group (titled 'Considerations and recommendations for graphs of statistical data included in Cochrane Reviews'). The guidelines have not yet been discussed by the SMG. He advised the group that conflicts of opinion on the contents were to be expected.

    2. Technical working paper on graphs

      3. A working paper that investigates the various possibilities for storing, generating, and displaying graphs within Cochrane reviews from a technical perspective had been prepared by Rasmus and Monica.

      1. Raw data availability

        2. One of the assumptions made in the working paper was that the raw data would always be available for download somehow. Mark felt that this assumption needs to be evaluated for two reasons: one is possible problems in terms of Intellectual Property Rights (IPR); the other is that this is an additional item to publish.

      2. Using the same software elements in RevMan and the CLIB

        3. Although RevMan and the CLIB do not use the same graph software, and are unlikely to do so in the future, Mark recommended that it should be investigated to what extent the various versions could share core elements.
        Action: Mark (to suggest who at Update Software should work on this), Rasmus.

      3. Access to data manipulation in published graphs

        4. Mark informed the group that Update Software were working on versions of the CLIB for new platforms, e.g. small hand-held devices. It would not be possible to include the possibility of data manipulation on such platforms.

        Julian reminded the group that there did not exist a consensus in the SMG on which single method of analysis is most appropriate for any specific meta-analysis. Therefore, the SMG has major concerns with only being able to see the reviewer's default.

        Ultimately, the group agreed upon the following recommendation for the PPG:

        Versions of analysis graphs where outcome data entered into RevMan cannot be manipulated may be published, provided that the data being plotted are presented alongside the graph and that the complete set of data from which these summaries were calculated can be obtained somehow.

        This recommendation should be cleared with the SMG before being sent to the PPG.
        Action: Julian to clear with the SMG, Monica to forward recommendation to the PPG.

      4. Database of raw study data

        5. Mark suggested that a future solution might involve creating a central database with all study data.

    3. Additional figures

      1. Terminology

        2. Julian felt that current terminology was confusing. It was therefore agreed to use the term 'additional figures' instead of 'additional graphs'.

      2. Heart Group pilot

        3. The Heart Group had piloted a method of including additional figures in reviews where image files were submitted along with the module to Update.

      3. Implementation

        4. The group agreed on implementing the additional figures in the following way:
        • A new version of RevMan (RevMan 4.2) will be developed. It will be a non-mandatory release. RevMan 4.2 files will be backwards compatible with 4.0 and 4.1, but users of the older versions will not be able to use the new features or see the additional figures.
        • A method of adding additional figures to reviews as graphics files (bitmaps) will be implemented. The image files will be stored in the current RevMan 4 data structure by re-using some of the existing fields. This feature is not meant to be used for inverse variance meta-analysis, which will be re-calculated in the CLIB from the raw data. It can be used to store other approved graphs (some of which may be produced using the new statistical module), advanced tables which cannot be produced by RevMan, and perhaps other figures if the inclusion of such are approved by the CCSG.
        • The CLIB will display the additional figures coming through RevMan. This will probably be as pop-up windows activated by links.
        • The responsibility for what to include should lie with the editorial bases, but guidelines are needed. Graphs that can be generated from within RevMan/MetaView should not be replaced by figures.
        • The Technical Implementation Advisory Group should develop specifications for the technical restraints on images (e.g. format and size).
        Action: Mark (to suggest who at Update Software should be involved at their end), Rasmus.

    4. MetaView future

      1. MetaView 4.1 in CLIB

        2. Mark informed the group that the plan is to include Metaview 4.1 with next generation of CLIB (like 3.1 is included in current online CLIB). The next generation is scheduled to become available as an alternative to the current online version after the colloquium.

      2. New statistical module

        3. It was decided that in RevMan 4.2, MetaView 4.1 will be replaced by a statistical module developed by the Nordic Cochrane Centre. The module will have support for calculating inverse variance meta-analysis and may also support the generation of other graphs as proposed by the SMG in the draft document 'Considerations and recommendations for graphs of statistical data included in Cochrane Reviews' by Julian Higgins. Julian agreed to suggest names of statisticians who could be approached about testing the new statistical module. Rasmus agreed to send the data set with which MetaView 4 had been tested to Julian.
        Action: Rasmus, Julian.

    5. Inverse variance method

      6. It was agreed that:

      • The new version of RevMan will allow the entry of raw data (effect sizes and standard errors) needed for meta-analysis using the inverse variance method. The data will be squeezed into the current RevMan 4 data structure by re-using some of the existing fields (the end user will not notice this).
      • In the CLIB (next generation), graphs will be generated from the raw inverse variance data. Update Software may or may not use the same software as RevMan to produce the graphs. The CLIB may not provide support for manipulating inverse variance meta-analyses (these may be presented as fixed images only).

  8. Bi-annual report to the CCSG

    1. Task group budget

      2. It was agreed that the IMSG budget should include funding for 5 task group meetings (one such task group meeting would be to look at contact management).

    2. Travel expenses

      3. It was decided to change all travel costs in the budget from UK vs. continental to short vs. long travel distance. Long travel distance should be budgeted with £400 per return journey.

    3. Cost of telephone conferences

      4. The estimated budget for telephone conferences for Advisory Groups: 500£.

    Action: Monica to incorporate changes to the report.

  9. RevMan 4.1 multi-user

    10. Rasmus has developed a working multi-user enabled version of RevMan 4.1. But since RevMan was not originally designed for multi-user capability, he would not recommend basing RevMan 4.2 on the multi-user code, because of the risk of introducing bugs.

    1. RevMan 4.2 multi-user

      2. Since the need for a multi-user version is limited to (large) editorial bases, RevMan 4.2 will be based on the original code. It may be possible to produce a multi-user version of 4.2, possibly with only some of the new features, but presently the NCC cannot commit to doing so.

  10. Release of ModMan 4.1

    11. The announcement of the release awaits testing that downloading works. It is likely to be announced 13 September. Although not mandatory, the MAG will recommend an upgrade to version 4.1, and version 4.0 will not be supported after 2001.

  11. Update on Contact Information Pilot

    1. Status

      2. Almost the entire HIREx databook has been imported as have 2000 records from the UKCC/Secretariat Outlook database, which gives a total of approx. 6700 records. A lot of the data, however, seem to be of low quality. Currently, there are 12 registered users, but more need to be recruited to generate sufficient feedback.

      It was decided that the members of the IMSG should be given access.
      Action: Jacob

    2. Future plans

      3. The steps involved in taking the project forward is as follows

      • Expand the user base (but only to a limited size to ensure that support needs can be met).
      • Obtain and structure feedback.
      • Cross-check feedback with available needs assessment results.
      • Implement possible changes.
      • Evaluate.
      • Possibly propose global implementation to IMSG.

      Action: Jacob.

    3. Tracking elements

      4. Sonja suggested that the pilot should allow administrators to make simple tracking notes for persons.
      Action: Rasmus

  12. The Cochrane Collaboration's web site

    1. Current problems

      2. There was agreement that the site needs to be easier to navigate but Gerd informed the Group that the current level of funding only allows for low-level maintenance. However, funding for a full-time person should be imminent.

      Some useful navigational features, such as a search engine would require that all mirror sites have the same server platform.

    2. Mirroring

      3. The need for mirroring was discussed, and it was agreed that it should be maintained if feasible - especially for the information of general interest. Already a number of infrastructural sites are not part of the main site, and thus not being mirrored: RevMan, ModMan, Titles.

      To allow new features to be included, Gerd will circulate a proposal for standardizing the server infrastructure. One option could be buying server space at relevant locations.
      Action: Gerd.

    3. Group remit and responsibility for website

      4. It was clarified that the web site contents belong under the PPG's remit - but that decisions about implementation, delivery, and platform are relevant to this group.
      Action: Monica to include this in the report to the CCSG.

  13. Any other business

    1. New name for advisory group

      2. Since the acronym TAG is already in use, it was decide to find a new name for the 'Technical Advisory Group'. The decision was left up to the group's convenor.
      Action: Rasmus

      Post-hoc note:
      The new name has tentatively been set to Technical Implementation Advisory Group, TIAG.

    2. Role of Quality Improvement Manager

      3. The role of the Quality Improvement Manager in relation to the IMSG was discussed. It was clarified that the QIM's role will be to visit entities and identify good practice and then generate recommendations based on this. The recommendations generated should, where relevant, be reflected in any software being developed. Therefore the QIM should be expected to present reports to the IMSG.

      The roles of the IMSG, QIM and Quality Advisory Group in relation to the SG, were discussed, and it was agreed that the report to the SG should request a clarification of those roles.
      Action: Monica

    3. Intellectual Property Rights

      4. Mark informed the group that Update Software as publisher considers that when a review has been submitted for publication from RevMan, it becomes copyright of Update Software.

      Mark also advised that IPR should be a concern in IMS work. An example of possible issues are 'what constitutes a modified Cochrane Review that can be published elsewhere'. Mark agreed to write up a brief paper about IPR and the IMSG.
      Action: Mark

      Mark also drew the group's attention to Digital Object Identifiers, a system that Update Software are planning to use for reviews.

    4. New RAG member

      5. It was agreed that Valeria Marinho should be invited to join the RAG

      Action: Monica (on behalf of Mike).

  14. Date of next meeting

    15. 14 December, 2001 at the UKCC meeting room, Oxford.

Document Actions