RAG minutes, 28 April 2003

Minutes approved on 10 September 2003

Present: Phil Alderson, Jacqueline Birks, Mike Clarke, Sonja Henderson, Dymphna Hermans, Julian Higgins, Monica Kjeldstrøm, Valeria Marinho, Heather Maxwell, Tess Moore, Rasmus Moustgaard, Elizabeth Pienaar, and Lindsay Stead.

In attendance: Claire Allen (minutes)

Apologies: Gerd Antes, Chris Cates, Katharine Deane, Bill Macguire

Resignations since last meeting: Kate Henshaw, Phil Wiffen

Contents:
1. Welcome
2. Approval of agenda
3. Approval of minutes and matters arising, not on this agenda, from the RAG meeting held on 14 November 2002
4. RevMan 4.2
5. Search strategies
6. Citations for other Cochrane reviews within a Cochrane review
7. Needs Assessment Survey - Developing Country Survey
8. New publishing arrangement for the output of The Cochrane Collaboration
9. Wishlist for RevMan
10. Sources of support - providing additional information
11. RevMan 5
12. Future plans for the Information Management System
13. Umbrella reviews
14. Any other business
15. Date of next meeting

Welcome

Mike welcomed everyone to the meeting, and thanks were expressed to those members who had resigned from the RevMan Advisory Group since the meeting on 14 November 2002. Apologies for absence had been received from the people listed above.

Approval of agenda

The agenda was approved with the addition of 'Remit of the RevMan Advisory Group'.

Approval of minutes and matters arising, not on this agenda, from the RAG meeting held on 14 November 2002

The minutes were approved without amendment. Rasmus agreed to revise their status on the IMSG web site to reflect this.
Action: Rasmus

Matters arising, not on this agenda

Item 4 - Publication of co-reviewers' contact details in The Cochrane Database of Systematic Reviews

Mike had been asked whether co-reviewers' contact details should be those at the time that the review was written, or whether the contact details should be up-to-date if possible. It was agreed that the details should be the most up-to-date available to the RGC. Mike agreed to respond to the RGC that had raised this issue.
Action: Mike

Item 12 - Dates in Cochrane reviews

Mike advised that Frances Fairman had prepared a document on the use and display of dates within Cochrane reviews. He agreed to discuss this with her and report back to the next RAG meeting.
Action: Mike

Item 18 - Chinese version of RevMan

Monica agreed to ask the Chinese Cochrane Centre for an explanation of what method they had used to translate the interface for RevMan. She will e-mail the RAG to see if anyone had concerns. Action: Monica

RevMan 4.2

Advice on using new features

Headings

Sonja reported that the Cochrane Style Guide Working Group had been asked by users of the Cochrane Style Guide to recommend a style for headings to be used in the text of RevMan, now that reviewers can use bold, italics and underline. Following recent discussions, the Cochrane Style Guide Working Group had suggested the following style for heading levels three to five. (Level one and two headings are currently fixed in RevMan and, therefore, cannot be altered by the RevMan user.)

HEADING ONE
Heading two
Heading three

It was agreed that there would be no additional spacing between these headings and the following text, but that there should be an empty line before the heading (in keeping with the display of fixed headings in RevMan). It was pointed out that the publisher can change the way in which RevMan's fixed headings appear when the review is published but, because the publisher will not be able to identify manually entered headings, they would be unable to change the appearance of the level three to five headings in the published review. These headings would, therefore, appear in the published review in the same style used to enter them in RevMan. There were concerns that the proposed level three heading would look strange if it appeared after the fixed heading two in RevMan, which uses bold with an initial capital only. If these additional headings are to be accepted, it was suggested that the publisher could change the appearance of the fixed headings when Cochrane reviews are published so that these are more easily distinguished. Sonja agreed to ask the Cochrane Style Guide Working Group to reconsider and revise their proposal for the level three to five headings and to send their proposal to Wiley for comment. Mike agreed to recommend to Wiley that the second fixed heading should be published in a larger font size than the body of the text (and, hence, these additional manual entered headings). It was agreed that there is a need for up to three additional levels of headings, but that the guidance should not be changed until after the module submission for Issue 3, 2003 of The Cochrane Library.
Action: Sonja, Mike

Italics

Sonja agreed to ask the Cochrane Style Guide Working Group to change the entry for 'The Cochrane Library', putting those words in the middle of a sentence, to show that 'The' should be capitalised. It was agreed that non-English words should be italicised, but Sonja agreed to check whether it was first use or every use within the review. She also agreed to discuss with the Working Group, the process for alerting people to changes made to the style guide.
Action: Sonja

Bold

It was agreed that reviewers should avoid using bold for emphasis, but that it may be used to indicate the letters used to form an acronym or abbreviation.

Underline

It was agreed that reviewers should avoid using underline, because it is often used to indicate an electronic link in documents (such as Internet links), and may cause confusion if an underlined word is not a link.

Superscript

The examples given in the style guide were agreed.

Subscript

The examples given in the style guide were agreed.

Help with using additional figures

Mike reminded everyone that RevMan 4.2 allows reviewers to insert additional figures and noted that concerns had been raised that some additional figures might be difficult to view in the published review. It was agreed that Wiley should check all additional figures before publication. Rasmus pointed out that if a large file holding an additional figure were inserted in RevMan, a warning screen would pop up within RevMan referring the reviewer to the guidelines. Monica (as convenor of the Information Management System Group) agreed to notify CRGs that Rasmus will provide advice on additional figures for a pilot period. This advice would focus on how a figure would appear in the published review and would not relate to editorial decisions about whether a figure is appropriate or suitable for the review. It was agreed that more written guidance is needed on the use of additional figures and Mike agreed to discuss this with Wiley.
Action: Monica, Mike

4.3 RevMan 4.2 on CD-ROM

Monica explained that no requests for RevMan 4.2 on CD-ROM had been received and, following discussion, she agreed to notify RGCs that the RevMan team will not be preparing CD-ROMs for central distribution. It was agreed that CRGs should encourage their reviewers to download RevMan from the Internet, in order to get the most up-to-date version and any patches required. Reviewers who cannot download it from the Internet should ask their CRG or Cochrane Centre to prepare and send them it on CD-ROM. If any entity has difficulties in doing so, they should contact the Nordic Cochrane Centre. Monica pointed out that a revised permission for publication form will be included in the next version of RevMan and it is important that this version of the form should be signed by reviewers of new and substantially updated reviews. She advised that there is no timetable for the release of the next patch, and that solutions for as many bugs as possible would be included in the patch when it was released. Rasmus agreed to include a solution for the bug which shows 'no numeric data if '0' is entered against the standard deviation, (it should show 'not estimable').
Action: Monica, Rasmus

Search strategies

Listing in protocols

There is currently no explicit guidance for including search strategies in the protocols for Cochrane reviews. It was agreed that guidance should be made clearer. The Handbook Advisory Group will be asked to consider this.
Action: Mike

Listing in reviews

After discussion, it was agreed that search strategies for each database searched should be presented in single columns of Additional tables in RevMan which could be linked to the appropriate part of the review text, including search dates. Heather agreed to seek the views of TSCs and RGCs on this recommendation, and to send any comments to the RAG e-mail discussion list. Mike agreed to ask the Handbook Advisory Group to include this advice in The Cochrane Reviewers' Handbook, once views had been considered.
Action: Heather, Mike

Citations for other Cochrane reviews within a Cochrane review

Mike advised that Wiley are making progress for citations to be included in ISI. He explained that to have an appropriate impact factor, ISI will need to be able to link each citation to its target record. He pointed out that it is particularly important that other Cochrane reviews are cited correctly within Cochrane reviews. Monica agreed to use the Parent Database to assess the extent of incorrect citations. When this has been done, Mike will notify TSCs and RGCs about the importance of entering citations correctly. Monica agreed to liaise with Wiley who were discussing this issue with ISI. Mike said that he would soon send all entities a short report on a project carried out by the UK Cochrane Centre, in which citations of Cochrane reviews were searched for in the Science Citation Index. The initial findings showed that items of output of The Cochrane Collaboration has been cited at least 7000 times in the period since 1995.
Action: Monica, Mike

Needs Assessment Survey - Developing Country Survey

Elizabeth explained that the needs assessment survey of people in developing countries had been carried out by e-mail. She had received 80 returns (20%) from 335 sent, which was a higher response rate than the worldwide survey in 2001. Although plans had been made to translate the survey into languages other than English, only one such request had been received. There are responses from 26 countries, covering 22 different languages. 93% of respondents were comfortable completing the survey in English. 83% of respondents had a Cochrane Centre in their country. The main problem reported was the use of Cochrane software; covering issues such as its user friendliness and its inability to act like Microsoft Word. Respondents mainly used Microsoft Word for word processing. One person had no Internet access, although most had access at work. Predominantly, people used Microsoft Word to prepare protocols and reviews before transferring the text to RevMan. Elizabeth agreed to send a more detailed report to the RAG by mid-May 2003 and will work with Valeria on this. The UK Cochrane Centre had agreed to purchase one-year subscriptions to The Cochrane Library for two respondents as an incentive. These two people were randomly picked by members of the RevMan Advisory Group.
Action: Elizabeth, Valeria

New publishing arrangement for the output of The Cochrane Collaboration

Mike reported that the main impact of the new publishing arrangement on RevMan would be that of the new permission for publication form, which will be discussed in the upcoming IMSG meeting. He reported that the publisher is now John Wiley and Sons, and agreed to request that a technical representative from Wiley attend future RAG meetings. Discussion took place of whom to contact if some part of The Cochrane Library was not working. Mike advised that at present, users should contact Update Software who will re-route any enquiries. He reported that over the coming weeks, a message will be sent to all entities detailing contact people in Wiley.
Action: Mike

Wishlist for RevMan

Section F:
155 - Approved in principle for when RevMan 5 is developed.
156 - No desire for a new section - rejected.
157 - Structural change. New maximum of 50 characters agreed for RevMan 5. Rasmus agreed to consult the CCAG.
 ADD: It was agreed to add the item 'Study ID' with the status 'Rejected'.
158 - Rejected.
159 - Approved for RevMan 5 and include changing the font size to larger.
160 - Rejected.
161 - Approved as a display spell-check without automatic replacement.
163 - Rejected.
164 - Rejected. RevMan already has a row by row functionality.
165 - Rejected.
166 - Rejected. It will be available through the new IMS and editorial bases should have a tool to make .pdf versions. The publisher should be able to make a copy available to reviewers.
167 - Rejected.
168 - It was agreed that people should be able to zoom in and out of text to enlarge it in RevMan 5. The wish to change font size and style was rejected. Once the style of additional headings was agreed, it will be included in RevMan 5. Rasmus agreed to think about adding regular RevMan tips to CCInfo and Cochrane News.
169 - Remain as is - linked to RevMan 5.
170 - Julian agreed to raise this with the Statistical Methods Group to find a solution. Rasmus agreed to move this item to section D.
178 - Approved.
182 - Approved.
183 - Rejected.
186 - Rejected.
187 - Rejected.
188 - Approved.
189 - Approved.
190 - Approved.
191 - Structural change for RevMan 5. Rejected.
192 - Approved.
Action: Rasmus, Julian

Sources of support - providing additional information

The issue of whether the 'Sources of Support' part of a Cochrane review should be amended to store additional information (for example on the type of support) was deferred. It was noted that information of this type could be included in the Acknowledgements section. Rasmus agreed to add the item to the Wishlist.
Action: Rasmus

RevMan 5

Structure of a Cochrane review in RevMan 5

Rasmus explained the suggestion from the IMS team that RevMan 5 should move reviewers toward working with single RevMan files. This would remove the current problem that a problem with one review could corrupt the contents of other reviews. Concerns were raised about how this would work in practice and how versions of reviews would be identified but there were no objections to this suggestion in principle.

Version control in RevMan 5

Monica reported that the new IMS would include a version control system for reviews through the central server. It will be possible to open more than one version of a review simultaneously to enable copying and pasting. It would also be possible to mark reviews for publication using this central system.

Contact details in RevMan 5

Rasmus explained that contact details would be shared between CRGs in the new IMS and presented three options:
1. Maintaining the current system where RevMan stores a shared list of contacts identified by contact ID, and any conflicts with contact details stored in individual files are resolved when a review file is opened.
2. Including a separate unrelated set of contact details in each review file. This would mean that the same data would have to be entered more than once.
3. Not including contact details in reviews but linking to contact details within the central IMS.

Rasmus acknowledged there were more options than he had presented. Concern was raised about whether a central contact database would be compatible with all types of e-mail software. It was agreed that the preferred system was one where there is a link from the reviewer ID to the Contact Directory. It was agreed that there should be a unique identifier for each reviewer, rather than having these allocated by each entity. After discussion, the third option (above) was agreed. If the details for co-reviewers change, it was agreed that the RGC would be responsible for amending the Contact Directory.

Future plans for the Information Management System

Monica reported that there had been no further developments to discuss at this time.

Umbrella reviews

Rasmus will check that this item is on the Wishlist.
Action: Rasmus

Any other business

Remit of the RAG

Sonja reported that a revised remit is being finalised for the ModMan Advisory Group. Mike agreed to liaise with Sonja so that this can be used as a basis for the remit for the RevMan Advisory Group.
Action: Mike

List of journals

It was noted that RevMan 4.2 allows reviewers to build up a pick list of journals. Rasmus agreed to investigate further whether it would be possible to send a pre-prepared pick-list of journals to each RGC (for example by providing a dummy review).
Action: Rasmus

Date of next meeting

It was agreed that a special meeting would be convened in September 2003, in Oxford at which the main item of business would be the specification for RevMan 5. Mike agreed to canvass members for their availability. Rasmus will circulate background papers in advance of this meeting. The importance of having several reviewers at this meeting was acknowledged.
Action: Mike, Rasmus