4.2.9 Data synthesis

This section should describe how the data will be combined and reported in the systematic review. Essentially, in a systematic review of effectiveness there are two synthesis options: statistical synthesis (meta-analysis) and narrative summary (narrative synthesis). Details of the statistical models and methods and effect estimates that will be calculate and measures of statistical heterogeneity should be included (See Section 3.3). Authors should ensure that the effect estimates that will be calculate correspond to the type of data (dichotomous and/or continuous) they have suggested will be collected in their protocol (see Section 3.2.4.4). The review protocol should also explicitly specify the pre-planned approaches that will be used for the examination of publication bias, including the use of funnel plots and the use of statistical tests for the examination of publication bias (see Section 3.3.11).

The review protocol should explicitly specify that reviewers plan to use the GRADE approach for the reporting of the strength of evidence, including the reporting of the summary of findings table of evidence. The use of GRADE approach is currently endorsed by JBI and JBI reviewers must use it regardless of the synthesis approach employed, meta-analysis or narrative synthesis.