FAQs

Who needs to know about SoF tables?
Each Cochrane Review should contain at least one SoF table and review authors will be responsible for creating an SoF for their review. However, editors and peer reviewers will need to understand the tables and how they are created in order to provide informed comments. In addition, Review Group Coordinators will need to identify whether the SoF has been done properly and identify the need for additional support. The Applicability and Recommendations Method Group offers some limited support with the creation of SoF Tables (contact schuneh@mcmaster.ca)

Who is responsible for creating a SoF table for my review?
Review authors are now strongly encouraged to create at least one Summary of Findings table for their review. With the help of the GRADEpro software, review authors will be able to create their SoF table. Support to create these tables is provided to authors. The SoF table will become a part of your review, published in the review, and therefore authors are also responsible for its content.

After I create my SoF table who will check it?
The SoF table is like any other part of a review – it will need to be peer reviewed and approved for publication by the editorial team of the responsible Cochrane Review Group.

When should I start thinking about creating a SoF table for my review?
Planning for the SoF table begins with the selection of the outcomes to be included in the review in the protocol. You should consider what outcomes are important and should be included in the SoF table. The outcomes included in the SoF should not depend on whether information about that outcome was found. Instead, an outcome should be included if it is a primary outcome or important to people making health care decisions, such as clinicians and patients.

What if I want to include more than 7 outcomes in the SoF table?
It is strongly recommended that the final SoF includes no more than 7 outcomes. This limit was set because the SoF is meant to present key information from a review that is important for people making health care decisions.

How do I create a SoF table for my review?
There are 4 options for creating a SoF table:

1. create a table using GRADEpro by importing data from RevMan 5, preparing the table, and importing it into RevMan 5.
2. create a table using the template provided in RevMan 5.
3. create a table without a template and using the table function in RevMan 5.
4. create a table in another programme and save it as a figure to import into RevMan 5 as a figure.

What do I have to do to create a SoF table using GRADEpro?
GRADEpro will facilitate the creation of a SoF table. The overall tasks are presented in the HELP file. The basic steps are:

Using data from a RevMan file

1. Open GRADEpro and select and name a New Profile.
2. Import data and analyses from the RevMan file.
3. Select a comparison/profile. Add or edit the descriptive data about the comparison.
4. View and then choose a maximum of 7 patient important outcomes to present in the table. Move, delete or add outcomes.
5. Select an outcome. For each outcome there are 2 sections: Summary of Findings screen and the Quality Assessment screen.
6. Select Summary of Findings screen. Confirm data in the Summary of Findings screen and enter additional information (e.g. length of follow-up, assumed (control) risks, measurement scale descriptions,footnotes).
7. Select the Quality Assessment screen. Complete this screen by first confirming the number and type of studies, then assess the quality of evidence for the outcome by deciding whether to downgrade for each criterion that might decrease the quality of evidence (and, if relevant deciding whether to upgrade the quality of evidence) and enter footnotes expaining the reasons for downgrading (or upgrading).
8. Repeat (#6 - #8) for each outcome in the SoF table.
9. If, there are more than 1 comparisons of importance in the review, repeat (#4 to #9) to create a SoF table for each of the main comparisons.
10. Preview the SoF table, check it and edit if necessary. The SoF table can be previewed at any time as you create the SoF table by clicking on the ‘Preview SoF table’ button at the top of the screen. It should be checked carefully before exporting it.
11. Export the SoF table.
12. Import the SoF table into RevMan by ‘Summary of findings tables’ under ‘Tables’ in the tree structure and then clicking on the ‘Add Summary of Findings Table’ button in the screen to the right.

Are the Risk of Bias tables and SoF tables somehow linked?
Risk of Bias tables are created with a tool to help review authors assess the internal validity (or whether a study answers its research question correctly or in a manner free from bias). The information entered into the tool and presented in the Risk of Bias tables is important when considering the quality of the evidence for each outcome in the Summary of Findings Table.

When assessing the quality of the evidence using the GRADE approach in the GRADEpro, authors consider 5 factors which can impact the quality of the evidence. One of those factors is called ‘Limitations in design’. It includes the same threats to validity assessed in the Risk of Bias tool. Therefore, review authors should use the judgements they made when completing the Risk of Bias tables to make a decision about whether the quality of the evidence for a particular outcome should be ‘downgraded’ for ‘Limitations in design’ in GRADEpro.

There are so many systems out there for grading evidence. Why do we use GRADE?
Some grading systems are based on study design alone without explicit consideration of other important factors in determining quality of evidence. Some systems are excessively complex. No other system has focused on evaluating the quality of evidence for individual outcomes. An analysis of current grading systems has shown that these and other shortcomings have not been adequately addressed by any one system to date. See: How GRADE compares to other systems.

There is a need for a common, widely used system to reduce the confusion arising from wide variation in the use of multiple systems, with important shortcomings. The GRADE approach has been developed to address this need through an international collaborative effort of people from a wide range of organisations that produce clinical practice guidelines, health technology assessments (HTAs) and systematic reviews. It has become the official system of numerous organisations, many of which have contributed to its development, including the Cochane Collaboration, the World Health Organization, the National Institute for Health and Clinical Excelence (NICE), and numerous other organisations that prepare or publish guidelines, HTAs or systematic reviews.

Do I need special skills to use the GRADE approach to grade the quality of the evidence in my systematic review?
If you are preparing a systematic review, you should have the skills necessary to GRADE the quality of the evidence. When grading the evidence, you will need to consider factors related to study design (such as allocation concealment), meta-analysis (such as heterogeneity and precision), and publication bias. In particular, you will need to consider these factors and their impact on how sure you are about the results of your review for each important outcome. The GRADE approach is designed to consider these factors in a systematic and transparent way, and GRADEpro will help you to do this.

If you would like to improve your skills, there are many resources available.

For more FAQs related to the GRADE approach visit http://www.gradeworkinggroup.org/FAQ/