SDG minutes, 1 December 1998
Minutes of the Software Development Group meeting held at the UK Cochrane Centre on 1 December 1998
Minutes approved on ?
Present: Clive Adams (minutes), Chris Cates, Mike Clarke, Monica Fischer (chair), Sonja Henderson and Mark Starr.
Apologies: Jon Deeks.
In attendance: Rasmus Moustgaard (Nordic Cochrane Centre).
Welcomes: Sonja and Chris were welcomed to the SDG.
Contents:
Potential conflict of interest
Approval of minutes of last meeting
Action points from the previous meeting not covered elsewhere
Report from RAG
"Last date searched" field in RevMan
Windows 3.X versus 9x/NT
Report from Documentation Sub-group meeting
Report from ModMan Advisory Group
Establishment of the CritMan Software Advisory Group
Change of Advisory group names
Format of references
Any other Business
Next meeting
Potential conflict of interest
This standing item was revisited and no new conflicts were identified. Sonia recognised that her perspective would be from that of the review group co-ordinator of the largest review group within the collaboration, and Chris as a Comments editor. Chris also receives NHS R&D funding and thought that could potentially be a conflict of interest in certain circumstances.
Approval of minutes of last meeting
Amendments to previous minutes
3(d) Whole paragraph should be deleted if this is to be included in the mandate for the SDG.
4(c)(iv) "these checks are described in the handbook" rather than "will be "implemented".
4(c)(vi)(1) Jon will investigate where and how to include tests of significance between subgroups.
5(a) "ModMan 2.0" to be changed to "ModMan 4.0" throughout.
5(a)(i) July 1999 instead of April 1997.
7 "citation" to change to "references".Once these amendments were noted, the minutes of the meeting of 13 May 1998 were approved.
Action points from the previous meeting not covered elsewhere
2(b)(i) Action to improve communication between the members of the SDG and between the SDG and other entities; no action as yet. The SDG felt this should be integrated into Mandate of the SDG, which is currently being drafted.
Action: Monica.2(b)(ii) Reporting on the consultation process for RevMan 4.0 in Cochrane News has not been done and the time has passed for this idea to be useful. A summary for the rationale of RevMan 4.0 should be made available as soon as possible.
Action: Mike.3 Mandate of SDG and Advisory Groups
3(b) Monica has asked both the ModMan Advisory Group and the RevMan Group to comment on their mandate.
3(f) Redrafting and circulating of the document describing the mandates of the RevMan and ModMan advisory groups undertaken.4(b)(ii) Jon Deeks will meet with Gordon Dooley in January 1999.
Action: Jon.
4(b)(v) Rasmus has prepared an internal document with the argument regarding the the 16bit version - see Item 7 in today's agenda.
4(c)(ii)(2) Mike will be following up on the specific statistical sections of the Handbook with Jon.
Action: Jon and Mike.
4(c)(ii)(4) Claire Glenton has joined the documentation group.
4(c)(ii)(6) The BMJ full-text agreement has not yet gone through. It is not entirely clear how useful this will be, but it was thought that if some "stock" letter could be obtained from the BMJ this would be an advantage.
Action: Mike (to ask Andy)
4(c)(iii)(5) Re: MetaView Summary Screen and circulation of the proposal. The Steering Group has approved the new summary screen but suggested that it could be called "Entry Screen" or similar and that some explanatory text could be available to encourage "clicking" to get the full data. The Steering Group has also suggested that the CRD could be asked to look at the new summary screen to get ideas on how to refine it. This is now too late for MetaView 4.0, but Update Software is already in touch with CRD with respect to a later version. Mark will continue to consult with CRD vis-à-vis the development of MetaView.
Action: Mark.
There was discussion as to whether the Summary Screen / Entry Screen would be able to use the preferred statistics of the reviewer. Monica will check with Gordon Dooley in order to determine this whether this was indeed possible. If it is not possible, this matter has to be revisited for further discussion.
Action: Monica (Post hoc note: Gordon confirms that it is possible)
4(c)(iv)(2) Checks on the standard deviation/standard error were to be described in the manual.
Action: Mike.
4(c)(ii) It was agreed that a discussion of NNT needs to be in the Handbook. This documentation has been subsumed under the new documentation that Mike is developing.
Action: Mike.
4(c)(vi)(1) Jon has investigated with Simon Thompson with regard to tests of significance between subgroups and agreed that it is not advisable to implement it within MetaView.
4(c)(vi)(3) Length of p-values is still up for discussion. Monica was therefore to discuss with Gordon what would be the final decision on this.
Action: Monica.5(a)(iii) As regards the update of ModMan, the changes to the CRG module headings were hoped to have been incorporated earlier, but this had not been possible.
5(a)(iv) The functionally of ModMan was to be described and outlined by the end of June 1998 - to be discussed later in today's agenda.
5(a)(v) ModMan is a 32bit application (will only run on Windows 9x/NT). Monica is liasing with the CRGs that are still using 16bit operating systems (Windows 3.x) to find out what their plans are to upgrade.
Action: Monica.
5(a)(iv) As regards inclusion of the Master List of journals being searched within the Cochrane Library: this has been discussed by the CCSG and CCAG and it has been decided not to implement it for the time being.
5(a)(vii) The Cochrane Library Group has not yet been "activated", because the Steering Group has new plans for a series of advisory groups and their structures. Monica was asked to circulate the document describing these plans once it has been revised. The Cochrane Library Advisory Group would be subsumed within the new structure. Quite independent from this, Update Software is convening a series of advisory groups to advise on the next generation of software.
Action: Mark to report on this at the next meeting. Monica to circulate document.6(a) Re: The Contact Directory and publishing co-reviewers contact details. Apparently guidelines are being developed by Georg Koch of the German Cochrane Centre. He may work with the Nordic Cochrane Centre on this and the guidelines are expected soon.
Action: feedback from Monica.7 Announcement of a new standard format for citations and ongoing trials is yet to be done. Agreement on what exactly the standard format should be is on the agenda for the meeting. Sonja asked if it would possible if CRGs that use the citation field in RevMan 3.1 to store additional information could get help to move this to the notes field in RevMan 4.0. Rasmus was unable to commit to this but agreed to investigate the matter.
Action: Rasmus8(i) Re: Telephone conference on Criticism Management System. This had been undertaken with Andy as chair.
-
-
Unique identifiers for studies and other references
Sonja pointed out that the Handbook and the software are in conflict because of the limited field length in RevMan, and that one or the other should be changed. As a result of the ongoing copy editing process that Update Software is co-ordinating through the work of the Technical Editing Group with Rachel Stancliffe as convenor, a set of guidelines is being prepared to help reviewers produce reviews with a consistent style. This will have something on the Unique Identifiers within it.
Action: Sonja to forward the guidelines to Mike for him to consider modification of the handbook. -
Postcard announcements
It had been suggested that Update Software might want to send a postcard to reviewers with information about the date of the new release and the importance of converting. However, as it is the CRG editorial bases that are the first point of contact for reviewers, it was agreed that it should be the editorial bases that convey this message. Timing of the official announcement needs to be tied in carefully with the availability of RevMan 4.0 and of support and training. Sonja and Monica will work on an announcement to go to the review group co-ordinators.
Action: Sonja to work with Monica on announcement to Adminors. -
Conversion Hotline
It was envisaged that this would be available mainly for review group co-ordinators and not reviewers. Monica will contact the CRGs to find out when they are likely to convert to the new software (RevMan and ModMan) and therefore likely to be contacting the Hotline.
Action: Rasmus and Monica. -
Distribution of RevMan
After some discussion it was agreed that it would be up to the CRGs to distribute RevMan. The Nordic Cochrane Centre therefore plans to send bulks of CDs to the CRGs on the day RevMan is scheduled to be released (19 July 1999).
Action: Monica to ask review group co-ordinators the numbers of people that will be needing RevMan 4.0. URGENT. -
Contingency plan
It was decided that, if all else fails, and individual conversion by reviewers is not on schedule, then it could be done centrally by Update Software. This was seen as a last resort. -
Simultaneous support of old and new versions
A discussion took place as to whether two versions of CRG modules could be supported at the same time. However, it was agreed that all CRGs would need to convert to ModMan 4.0. If a CRG wishes to continue to allow reviewers to prepare their reviews in RevMan 3.1, it will be the CRG's responsibility to convert such reviews into the new format in order to incorporate them into its module. No central support will be available for such conversions. -
Dropping of short titles
It was recommended to drop short titles from RevMan. On reports where it will be difficult to print the full long title (e.g. in MetaView), it was agreed that a running head would be acceptable.
Action: Rasmus. As this is a structural change to the review itself the Steering Group need to comment on this.
Action: Monica. -
Membership
It was felt that the RAG was short of reviewers, and recruitment should take place at workshops and through co-ordinators. -
Double-citing of an individual report from within a multiple-published trial.
There was discussion as to whether, when special reference was made to a single report of a multiple-published trial, this single report should be doubly cited - not only within the "included trials" listing of references, underneath its relevant study tag, but also within "additional references". The decision was to allow its double entry with "additional references". Information about this should be included in the user guide.
Action: Mike
-
Unique identifiers for studies and other references
"Last date searched" field in RevMan
-
It was decided that the "dates" relating to the first publication of the protocol and the review should be 20 characters long and should have the form: Issue x, 199x.
Action: Rasmus. -
After a discussion on what exactly a "substantive update" meant, it was clear that guidance was needed within the RevMan 4.0 manual. Discussion suggested that this may actually be more appropriate for inclusion within the Handbook but, whatever, it is clear that close dovetailing of the RevMan manual and the Cochrane Handbook need to take place.
Action: Mike. -
An additional date is needed for last search.
Action: Rasmus. -
Rather than having four dates combined with 'reasons for update' it was decided to give individual dates for each of the options listed under 'reasons for update'. The 'Discussion section amended' should be removed from the list.
Action: Rasmus.
-
It was decided that the "dates" relating to the first publication of the protocol and the review should be 20 characters long and should have the form: Issue x, 199x.
-
It was agreed that the RevMan 4.0 should not be developed for Windows 3.x. This decisions is also supported by the fact that Microsoft is no longer supporting Windows 3.x. Mark added that it was unlikely that the Cochrane Library would continue to be supported in a 3.x version.
Report from Documentation Sub-group meeting
Membership consists of: Mike Clarke (chair), Jon Deeks, Monica Fischer, Claire Glenton, Jacob Riis and Rasmus Moustgaard.
-
MetaView documentation
The user guide for MetaView will be incorporated into the RevMan user guide. Technical documentation of the statistical formulas will also be included. Update software is responsible for the software specifications of MetaView. -
RevMan 4.0 documentation
On schedule for the beta release. The online documentation will probably have many links to the Handbook, but for the beta release these will probably be dummy links - see 8 (iii). -
Handbook
Will not be ready for the beta release of RevMan so there may be a dummy version produced so that the links within the RevMan manual can be tested. Training material within the Handbook is being revamped. Developing a tutorial is being considered.
Action: Mike.
Sonja emphasised the importance of an easy print version of the handbook - and was reassured by Rasmus - there would be some sort of way of automatically updating the contents page to fit in with pagination.
Action: Rasmus.The use of .PDF files was discussed and dismissed - as the Acrobat reader was not considered to be well known. Instead it had been decided to use Word.
-
MetaView documentation
Report from ModMan Advisory Group
Sonja reported issues that had come up at the Baltimore meeting:
-
Generation of reports
Options to generate reports seems limited. Lois will investigate if these can be expanded.
Action: Sonja to follow up. -
Link to HIREx
CRGs are waiting for instructions on how to update contact details within ModMan and link to HIREx -
User guide
It is unlikely that ModMan 4.0 user guide will be available by time of the release of the beta version. See 9 (iv). -
Training days for ModMan (and RevMan)
These will be crucial. After the training days for ModMan guideline production will be swift - the training itself providing many ideas for the guidelines. It was stressed that the editing of reviews will be in a version of RevMan and not in ModMan. It was understood that a review co-ordinator would be able to have two versions of RevMan 4.0 working at the same time, as long as they were installed in different directories.
Action: Sonja to liase re guideline production, and therefore link to HIREx.
Action: Rasmus to ask Monica to include Module Guidelines in ModMan 4.0.
Action: Rasmus to make sure that two RevMan 4.0s can work simultaneously. -
Communication of dates
Sonja stressed how communication of deadline dates in 1999 was most important. There needs to be a very formal announcement as regards the update of the modules. The draft guidelines has already been circulated, but nothing in writing about when CRGs are expected to produce a draft module.
Action: Mike to write to Andy and Monica to encourage formal notification of the need for the new module to arrive by the 29 January.Rasmus clarified that the support for ModMan would be Lois and Nick (from Update Software) to help with technical support.
-
Training dates
The dates for the UK ModMan workshop were set for one and half days stretching across 26-27 of April 1999 (Oxford). This would be followed by the RAG meeting 28 April, 11.00-16.00, and the Software Development Group 29 April, 10.00-16.00.
-
Generation of reports
Establishment of the CritMan Software Advisory Group
The establishment of this Group is under way with Chris Cates as its convenor. Nothing is to be implemented as regards criticism management within RevMan 4.0 but there is the possibility of some 'patch' to be put in at a later date. There was encouragement for this Group to work from the perspective of the needs of the criticism editors, +/- other types of users. It was recognised that the Criticisms Management Group is advising more generally on all aspects of criticisms management. Chris Cates is a member of this group to ensure that both groups are kept informed of new developments and that relevant ideas are exchanged.
Action: Chris and Monica to draw up the mandate for CritMan Advisory Group
Action: Chris to clarify the membership of this group.
Action: Chris to request the mandate of the San Francisco CC's group.Change of Advisory group names
The CritMan Software Advisory Group has been encouraged to clarify its role by adding into its title the word "software" and to distinguish it from the Criticisms Management Group. It has been suggested that the names of the other groups should also be changed. This is felt to be inappropriate as the RevMan and ModMan Advisory are not just software-specific.
-
Rasmus requested that field tags needed for different types of reference should be specified. With this information, he could organise a system by which each type of reference was compiled within RevMan 4.0. Field tags were to include Medline ID.
Action: Mike -
CRG guidelines for modules - Date has been set for the 29 January 1999 for CRGs to produce the modules. These are then to be commented on by various people for the next few weeks. In turn their comments are to be fed back to the CRGs for consideration and possible revision by April 1999. It was pointed out that the time for the comments to be compiled was greater than the compilation of the new modules themselves.
Action: Mike to to contact Monica and Andy about the CRG modules. (Urgent - within 3 weeks) -
29 April 1999 at 10.00 - 16.00 at the UKCC.

