Review Guidance

Data Extraction and Synthesis

Once the final set of papers has been selected, specific data will need to be extracted and synthesized for the review. The review team needs to decide which pieces of information need to be collected based on the review question. In general, study characteristics are extracted along with specific data related to outcome measures and results.

For our reviews, we have used simple worksheets to extract data by listing the items of interest in columns and the papers in rows. The example template below includes some items we were interested in for a particular review. There can be numerous columns, depending on the review question and intended analysis. Therefore, it is good to plan the synthesis methods as well when developing this worksheet so that all needed data can be extracted at one time and in the format needed to synthesize it.

Template:

In terms of synthesis or meta-analysis, there are several methods which can be employed. Kastner et al. (2012) reviewed a range of knowledge synthesis methods and provide descriptions of each in their paper (available here). Both qualitative and quantitative comparisons can be done. Quantitative meta-analyses should only be done when the studies are similar with respect to population, outcome, and intervention (Green, 2005, p. 272). For large sets of papers that are too heterogeneous, the reviewers may wish to group the papers according to shared characteristics for sub-analyses. Many resources exist for quantitative meta-analysis such as Egger, Smith and Altman’s “Systematic Reviews in Health Care: Meta-Analysis in context” (2001).

Back to Review Guidance Home
Search Strategy
Paper Selection
Quality Assessment
Data Extraction and Synthesis