Skip to Main Content
Library Hours

Systematic Reviews

Extract Data

​​Once you have selected the relevant studies to be included in your systematic review, a standardized data extraction form or template will help ensure you obtain all relevant information from each study, allowing you to examine and compare results.

As with the study selection process, it's a good idea to have two reviewers complete the data extraction independently, and to have a plan for resolving disagreement.

The Data Extraction Form

A data extraction form is essentially a template, tailored to fit the needs of your review, that you will fill out for each included study. The length and format of this form will vary based on factors such as the nature of your research question, the number of studies you are including, the types of studies you are including, and what types of analysis you are conducting.

How to develop a data extraction form for systematic reviews:

Determine what data items you need to collect.

The data that is relevant to the review will be different for every systematic review. Some categories of data that might be included:

  • Identification features (such as article title, author, citation, record number, type of publication, publication date, funding sources)
  • Methods
  • Participant characteristics
  • Intervention and Setting
  • Outcome/Results
  • Study quality factors
Determine the best way to collect that information.

Pre-specifying categories for form fields can help systematize the extraction process, and is especially useful when you plan on performing a quantitative analysis. When doing so, be sure to include, "other," "not reported," or "unclear" options.

At other times, it may make more sense to have an open field.
Select a tool to organize your data extraction.

Types of tools you might consider for your data extraction form:

  • Word Processors (e.g. Microsoft Word)
  • Spreadsheets (e.g. Microsoft Excel)
  • Surveys (e.g. Qualtrics)
  • Systematic Review tools (e.g. Covidence, RevMan, SRDR)
Run a pilot Try your form out on a sample set of studies and refine it as needed.