This guide has been developed for Barwon Health staff who are thinking about conducting a systematic review. The guide is not a comprehensive step-by-step set of instructions on how to conduct a systematic review. The guide contains links to resources that provide information about systematic review methodology and resources.
Barwon Health Library staff can meet with you and provide guidance on how to develop a literature search strategy. To meet with a Librarian, please complete the online form: Book A Consultation or contact the library for more information.
The video above was prepared by the Cochrane Consumers and Communication group. It provides an overview of systematic reviews (duration 3min 24 sec).
"Systematic reviews seek to collate evidence that fits pre-specified eligibility criteria in order to answer a specific research question. They aim to minimize bias by using explicit, systematic methods documented in advance with a protocol."
- Cochrane Handbook for Systematic Reviews of Interventions
Learning Opportunities
Below are links to learning opportunities that Barwon Health staff highly recommend:
Courses at Barwon Health for staff - keep a lookout in the What's On; Research and Grand Round Newsletter and Library e-newsletters:
This ebook, developed by Librarians at Federation University, Erin Roga and Dr Karen Pruis, leads readers through the stages of conducting a review: Introducing scoping and systematic reviews.
Ask the following questions before you invest time in thinking about or planning a systematic review:
If you have been given the go ahead, you will also need to ask yourself if your research question lends itself to a systematic review or is it more appropriate for a scoping review or literature review?
To help you answer this question, read the article:
If you have decided to conduct a systematic review, we suggest you read this article before you start formulating your research question to work out which kind of systematic review you will conduct:
According to Munn et al. (2018), many authors try to force their research question into the PICO (population, intervention, comparison, outcome) framework, which may not suit the research question, and confound the rest of the review process. Table 1 (p. 3), provides a list of typologies, the aim of each typology, the relevant question format and a question example. Each typology is explained in more detail within the article, with a brief description of relevant tools to assist the review process (as applicable).
Is the answer already out there?
When you have decided on a topic, search in PROSPERO or PubMed (use the SR filter) to find systematic reviews on your topic.
PROSPERO is a database of prospectively registered systematic reviews, containing review protocols.
If you find that a systematic review is being conducted on the topic you have chosen, consider whether you need to replicate it. The following article provides consensus based guidance on when to replicate or not (including a 12-item checklist):
If you haven't already done so, get other people involved. Recruit at least one other person to assist you to screen studies. Having at least two team members screen, helps to minimize bias and ensure the reliability of the review process.
A protocol outlines the methods and procedures you will follow, such as:
Inclusion and exclusion criteria for studies,
Search strategy
Data extraction process
Methods for assessing the quality of included studies.
For more detail, view systematic review protocols in PROSPERO or view this example.
You may already have a decided to follow a particular systematic review methodology or guideline, such as Cochrane, Campbell Collaboration, JBI (Joanne Briggs Institute). Depending on your methodology, you may use PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidance, which includes a checklist and flow diagram to help with systematic review reporting.
Once you have finalised your research question, and after you have developed a protocol or draft protocol, you will need to develop a search strategy and conduct a comprehensive literature search. Multiple databases should be used, including clinical trials registries (as relevant) and grey literature sources. You should have a set of 'gold standard' articles and look at the references in each of these, as well as entering the gold standard article titles into Google Scholar or PubMed to see who has cited these articles. PubMed also has a feature where you can see 'Similar articles'.
For more information about the requirements of a comprehensive literature search, please see the
Cochrane Handbook for Systematic Reviews of Interventions, Chapter 4 'Searching & selecting studies'. The information in this chapter is relevant no matter which type of systematic review you conduct.
If you need assistance in developing your search strategy, please contact the Library or complete the online form: Book A Consultation. A Librarian can tailor the consultation to your needs', for example, show you to use databases effectively, and provide guidance for key terms, phrases and relevant subject headings to include. An initial consultation usually runs for an hour and can be conducted face to face at the UHG Library or the McKellar Library; alternatively, online via Teams or Zoom. Please consider attaching your draft protocol to the online booking form.
The most efficient way of managing your citations is to use citation management software such as RefWorks or EndNote.
The Library has a RefWorks subscription – for more information see RefWorks or contact the library.
The point of a systematic review is to find as many relevant studies as possible. Your search strategy should be broad to ensure that relevant studies are found, and therefore a number of irrelevant studies will also be found.
All team members should start off by screening studies at the title and abstract level to identify relevant studies according to the inclusion and exclusion criteria set out in your protocol. Studies considered relevant will need to be assessed by examining the full text of the article, using the inclusion and exclusion criteria. There are online tools that can assist with screening, for example:
When you have your set of included studies, you will need to extract the relevant data in order to synthesise the evidence and write a report. Data extraction will depend on the type of systematic review you are conducting. Data should be recorded in a standardised format.
The Barwon Health Biostatistics Support Service provides statistical support to Barwon Health researchers. The service provides assistance with data analysis, interpretation and reporting.
Publication in a peer-reviewed journal is not the only way to disseminate your findings, here are some other ways:
For more ideas see the following article: