Systematic Reviews: A How-To Guide

Data Extraction

The next step is to extract data from each study, data refers to information which is relevant to your review/thesis question.

You will need to extract data from relevant studies in order to examine and compare results. While the data is being extracted, it is very important to employ good data management practices. Proper data management should begin as soon as you start extracting data, and may even dictate which types of data you decide to retain. Some reviews require the use of software to help with extracting data.

The links below include can help you with this task.

• Covidence

  Subscription to unlimited reviews provided by UVic Libraries. Recommended by Cochrane. Sign up here.

• EPPI-Reviewer 4

  Fee-based; offers one-month free trial. Features include data extraction, coding, and meta-analysis.

• Distiller

  Fee-based; offers special pricing for students (free for 4 months; $15USD/mo after that) and Cochrane Review Groups. Available in two versions (DistillerSR and DistillerCER) with varying features.

• JBI SUMARI tool

  Free with JBI subscription; requires login. JBI SUMARI supports the entire review process, from drafting your protocol, study selection, critical appraisal, data extraction and synthesis.

• Systematic Review Data Repository (SRDR)
  Free; requires login. This systematic review repository also acts as a data extraction tool.

Adapted and modified with gratitude from Dalhousie Libraries Data Extraction and Management and City University of London's Doing Post Graduate Research research guide.

Good Data Management Practice

The NYU Health Sciences Library has put together a short video about best data management practices. The video outlines four data management tips:

• Retain component variables instead of compound variables whenever possible
• Clarify all aspects of your team's workflow expectations
• Develop a naming convention for data to ensure clarity between team members
• Make frequent backups of your work, housed in different locations

Extracting Data

The extraction process should be tracked using a standardized data extraction form (see examples below). Data can also be coded for computer analysis.