Cochrane review advice

The advice on this page is some practical stuff about taking on a Cochrane review - though to some extent this could be applied to undertaking any systematic review. It has been compiled by Sally Collins, an Oxford medical student, though previously a research assistant at Pain Research in Oxford and an author of many high-quality reviews. This is, of course, a personal view, and not formal Cochrane Collaboration policy.

A: Background knowledge
Chosen subject			If 'No'
Do you have sufficient knowledge of the subject area to generate an important, appropriate and potentially answerable research question?	Yes	No	Seek advice from your RGC to discuss recruiting additional reviewer(s) with appropriate knowledge to the team.
PaPaS can provide tools to facilitate the writing of a review but the reviewer(s) must have a sufficient level of understanding of the topic to enable meaningful analysis and relevant clinical comment
RCT Methodology
Do you have a good understanding of the validated methodology for high quality RCTs in the chosen topic?	Yes	No	Seek advice from an expert, preferably someone actively involved in designing and running RCTs
Are you aware of common design errors in trials in your topic area (such as testing an analgesic in patients who have an insufficient level of pain)?	Yes	No	As above
B: Skills
Database handling
Are you confident about using an information retrieval system for searching electronic databases, eg, Silver Platter (WinSPIRS) or Knowledge Finder?	Yes	No	Training may be available through your local medical library, or contact your RGC for advice
Do you have an understanding of how search engines operate, including Boolean operators, in order to formulate a sensitive search strategy?	Yes	No	As above
Computer skills
Do you have basic computer/word processing skills? These are essential.	Yes	No	Seek advice from your RGC. Cochrane review-writing software (RevMan) is reasonably intuitive, and training courses are available through most Cochrane Centres
Basic Statistics
Do you have a reasonable level of statistical knowledge? This is necessary: To assess whether the original study has used an appropriate test for statistical significance. To decide what data can be extracted and used for analysis with RevMan. To decide whether any non-RevMan analyses are suitable for your review, eg to generate numbers-needed-to-treat (NNTs)	Yes	No	Most educational establishments offer training courses in basic statistics, and there are many good textbooks on medical statistics
Critical Appraisal Skills
Do you have experience in appraising papers and extracting relevant data? Are you able to identify whether an appropriate methodology and appropriate quality scales have been used?	Yes	No	Courses, such as CASP workshops, may be available. Also try Trisha Greenhalgh's book 'How to read a paper' 1998, BJM Publishing. ISBN 0-7279 1139-2; and this website: www.update-software.com/CASP.htm
If English is not your first language
Cochrane reviews are written in English and have a wide audience. Can you write clearly, or do you have access to help from a medically-oriented translator?	Yes	No	You may need to find a co-reviewer/ translator to help.
C: Resources - essential
Time
Have you set aside sufficient time to prepare the review? Reviews can take 1-2 years to complete on a part-time basis, and are difficult to undertake in half-hours snatched from an already busy schedule.	Yes	No	Consider whether this is the right time for you to be taking on significant extra workload
Cochrane reviews require updating once a year and it is your responsibility to do this, or arrange for a co-reviewer to do it. Are you able to make this level of commitment?	Yes	No	Consider whether the Cochrane Library is the right place to publish your review
Computer facilities
Do you have sufficient and convenient access to a computer? RevMan is used at the protocol and review writing stages, and a PC (or Apple Macintosh with a PC simulator) which uses Windows 95, Windows 98 or Windows NT 3.51 (and upwards) is essential. For reasonable speed a Pentium with 32 MB of RAM is recommended.	Yes	No	Seek advice from your RGC
Access to databases
Do you have access to the databases needed to undertake a thorough search? As well as the Cochrane Library and Medline, there may be others which are relevant to your topic, eg, Embase, Cinahl, PSYCLIT	Yes	No	Seek advice from your university or medical librarian, or RGC
Library support
Do you have library support and resources to obtain all the required papers? You may need papers that are published in journals not held by your local library. Will you be able to obtain/translate these? It may prove expensive.	Yes	No	Seek advice from your librarian or RGC
Co-reviewers
Do you have a co-reviewer to work with? All papers for inclusion in the review and data extraction should be agreed/undertaken by at least two people independently of each other	Yes	No	Seek advice from your RGC
D: Resources - useful
Reference manager
Do you have a reference manager for handling the citations generated by your search (eg, Procite or Reference Manager)? This can help you to keep track of possible RCTs. Especially important when considering a topic with a lot of related published information	Yes	No	Investigate the possibility of getting software which has this facility. Seek advice from your librarian or systems administrator, as many information retrieval systems have this capability
Can you download your search results to disk? It can be hard to cope with search results if you can only print them off, and is essential if you want to transfer the records to a reference manager	Yes	No	As above
Internet/e-mail access
Do you have e-mail? This is not essential but could prove useful for maintaining contact with other reviewers and the PaPaS editorial base.	Yes	No
Sally Collins/fsf/2000
Glossary
CASP: Critical Appraisal Skills Programme
RCT: Randomised controlled trial
RevMan: Review Manager - Cochrane Collaboration software for writing reviews
RGC: Review Group Co-ordinator
PaPaS: The Cochrane Pain, Palliative and Supportive Care Group