Basics of Systematic Reviews

A guide to evidence synthesis

Step 1a: Develop Your Research Question

Research questions for systematic reviews are specific and answerable. They should be structured so that clear criteria for locating eligible studies can be identified.  

Research question frameworks can be very useful in helping your team hone and narrow the research topic.  These models help structure the inquiry and therefore aid in searching for articles in later steps.  Clinical and health science-focused systematic reviews typically use the PICO framework (population, intervention, comparison, outcome).  However, there are many different types of frameworks that can be used for a variety of disciplines:

Research Question Frameworks

Description of the PICO framework

Description of PEO Framework

Description of SPIDER framework

The SPICE question framework may be useful for qualitative research topics that evaluate the outcomes of a project, service, or intervention.

Description of the SPICE Framework

PCC is often useful for both quantitative and qualitative or mixed methods topics, and is commonly used in scoping reviews.

Description of the PCC question framework

The ECLIPSE framework can be useful for qualitative research topics that investigate outcomes of a service or policy. 

Description of ECLIPSE question framework.

If these question frameworks don't fit your research topic, don't panic! There are dozens of other question frameworks that you can use:

Step 1b: Build Your Team

Identify other individuals who will assist with the systematic review.  It is highly recommended to have 3 reviewers at minimum for a systematic review team.  Since systematic reviews typically take 12-18 months, it's important for your team members to understand the time commitment necessary for this undertaking.

Some other considerations include:

  • Determine if those on your team speak a language other than English; this can be useful for when it comes time to determine eligibility criteria and whether you will include articles in other languages within your review.
  • If you intend to pursue a meta-analysis in your systematic review, it may be useful to have a statistician on your team to assist with data extraction and statistical analysis.
  • If a librarian is assisting with more than the development of the search strategy (i.e. running/conducting the literature searches and helping to write the methods section), they will need to be credited as a co-author.  However, if the librarian is just assisting with the search strategy, only an acknowledgment at the end of the review is requested.

Step 1c: Review the PRISMA Checklist

Before starting the systematic review process, review the PRISMA 2020 Checklist.  The PRISMA Checklist is an evidence-based minimum set of reporting items for systematic reviews and meta analyses.  The checklist is intended to improve transparency in systematic reviews.  While primarily associated with the health sciences, its framework can be applied to fields outside of biomedical disciplines as well. 

Step 1d: Select a Citation Manager

Systematic and scoping reviews require a large amount of articles to be exported and deduplicated in preparation for the screening process.  A citation manager should be used to expedite this step.  Citation managers are databases that allow you to store, organize, and share your research and citations.  UTSA has several free options available.  You can view our research guides on various citation managers and download the applications using your UTSA credentials through the below links.

Step 1e: Explore Screening Tool Options

One of the most important later steps when conducting a systematic review is screening, which is the process of identifying studies from the literature search for inclusion in the review.  Two of the most common tools used during this step are Covidence and Rayyan.

The reason it's important to consider screening tool options early on is the difference in cost.  Rayyan offers a free plan with basic functions, as well as subscription plans that begin at $8.33 a month.  Covidence's individual plan costs $289.00 per year and allows for unlimited collaborators (other reviewers).  We recommend consulting with the research administration at your department or college to determine if the cost of Covidence may be written into your grant (if the project is funded).

Step 1f: Specify your Review's Eligibility Criteria

Eligibility criteria (also referred to as inclusion/exclusion criteria) are the characteristics of an article found using your search strategy that define whether it can help you answer the research question.  The criteria decided by you and your team will be used to decide whether you will cite an article in your review or not.

Some examples of eligibility criteria can be found below:

Inclusion/Exclusion Criteria Examples
Type of Criteria Example
Study participant types Eligible articles may be limited to specific groups of people or age ranges. (ex: adult women)
Intervention Eligible articles may include interventions of interest and excludes any others (ex: CBT therapy)
Setting Eligible articles may be limited to a specific setting (ex: inpatient, ambulatory, classroom setting, etc.)
Study Design Eligible articles may include specific study designs and exclude others based on which studies best answer the research question. (ex: clinical trials, observational studies, etc.)
Outcomes of interest Eligible articles include outcomes of interest and may exclude studies that report outcomes not of interest (ex: mortality rate, adverse events, patient satisfaction, test results, etc.)
Publication Type Reviews, editorials, commentaries, and letters are often excluded from systematic reviews.
Publication Dates Date ranges are typically only applied when updating a systematic review or when a start date is specific to an intervention or therapy (ex: a systematic review focused on a specific type of radiation therapy would have a date range beginning when that therapy first appeared as a treatment).
Language Non-English articles are frequently excluded from reviews; however, this may allow language bias to affect the quality of your review and should be acknowledged as a weakness during the write-up.