Skip to Main Content
It looks like you're using Internet Explorer 11 or older. This website works best with modern browsers such as the latest versions of Chrome, Firefox, Safari, and Edge. If you continue with this browser, you may see unexpected results.

Good review practice: a researcher guide to systematic review methodology in the sciences of food and health

Data extraction

Data Extraction 

After the quality assessment is completed, the data is extracted from the included studies to make a conclusion about the outcome.

What data to extract for synthesis?

The review protocol defines which relevant information is to be collected from the included studies.  Regardless of the topic, these include the following key information of the primary studies:

  1. Study characteristics that are important to assess the internal validity and external validity (applicability) of the primary studies. For example, specific characteristics of the subjects/participants of the primary studies that might limit or extend the applicability of the results or causes heterogenicity of the outcomes. 
  2. Study methods, conduct and limitations which are likely to be a potential source of bias in the estimation of the study results. For example, analytical methods used to estimate prevalence of a disease cause or detection limits of diagnostics accuracy tests. 
  3. Study findings including numerical and descriptive information such as study sizes, sample sizes or providing a mean/median plus a standard deviation for the quantitative data or a base line for qualitative data. 

Data extraction procedure

Data Extraction Procedure

Creating Extraction Forms 

All data collected from included studies at this stage will be used for data synthesis and for concluding the outcomes. Tables and tabulated forms are used to group or categorise studies based on those aspects important to the review topic and should be planned and structured at the protocol development step. Using tables to compile data is also necessary to keep data extraction consistent across all studies. (See Appendix A)  to choose from examples of data extraction template, designed by the first author to download)

Grouping and organising data 

The underlying principles of the topic and the aim of the review determine the specific characteristics by which studies should be grouped for synthesis. These characteristics are predetermined and the methods of organising studies into subsets and groups should stay consistent with the methods proposed and documented in the review protocol. 

The study characteristics and key information of included studies can be then organised into explicit categories in the tables and forms or even summarised. Summaries describe studies in short which may include sample size, study design, setting, location, and authors and year of publication. Summaries are usually created for collaborative reviews, descriptive and qualitative studies. They can be interpretive and use specific keywording strategies, also known as keywording process to construct descriptive maps of literature in social contexts. The results of quality appraisal or weighting processes and the ratings of studies are added to the tables or summaries at the end.

Preparing data for statistical analysis 

Numerical data should be extracted from included studies in a format that is consistent with the one used to structure study characteristics to examine and prepare them for data analysis. Data types vary depending on the nature of studies and methods of analysis.  

Data extraction templates