Literature reviews - services and support

Before you decide to do a systematic review, it is important to check what similar studies have been done in the past and how they differ from your intended review. Are there enough arguments to make a new one? Or can previous studies be used to inform your own? Since the protocols for systematic reviews recommend that you state "systematic review" or "scoping review" in the title of the following article, these studies are easy to find.

Within the family of systematic reviews, there are several different variants of literature reviews. A fully fledged systematic review is resource-intensive and places high demands on knowledge of the subject area. Therefore, different variants have been developed that are less resource-intensive (rapid review) or that aim to explore a subject area for a regular systematic review (scoping review). Other reviews focus on compiling results based on published reviews (umbrella review) and depending on how you compile the results, there are also variations (e.g. meta-analysis). Therefore, it is good to think about the project's conditions before starting. To support you in choosing an overview, you can contact the library or consult this quick guide from Cornell University library.

Different types of overviews take different amounts of time. Know early on how long the project can last. A systematic review usually takes between 6 months and 2 years to complete.

A characteristic feature of a work, with a systematic review as a method, is that it requires at least two people in the project. The idea is that the project participants, independently of each other, go through the material according to predetermined criteria in order to be included or excluded. This is to prevent arbitrariness and bias.

To manage and collect your references, we recommend that you use a reference management program such as EndNote or Zotero. Endnote includes a lot of storage of full-text files in the cloud, in Zotero this storage space is limited from the start but can be expanded if needed by connecting to different cloud services.
In the work with screening, i.e. in the process of reviewing the search results, there are several programs that support the work, e.g. Rayyan and Covidence. Recently, programs have also been developed that can shorten the review work with the help of AI, see ASReview.

A well-defined and delimited research question is fundamental to a systematic search. To structure and limit the research question, there are several frameworks available. On the University of Plymouth Library website, you can find information about the most common frameworks used for this purpose. They can be helpful in constructing a query so that it is possible to extract searchable main concepts.

For certain types of topics and research questions, it may be appropriate to supplement the database search with other search strategies. For example, it may be a matter of searching in certain selected journals, or doing a citation search.

In connection with systematic reviews, it is also recommended to do a so-called forward and backward search for the studies included in the review, i.e. that you go through citations and references to the included publications. There are several different tools and databases where you can search for citation data, some examples are Web of Science, Google Scholar and SpiderCite.

Grey material such as government reports, articles in industry journals, etc. does not have the same evidential value as research publications, but can still contain insights into the state of knowledge on an issue and highlight negative results (which are rarely published). It may therefore make sense to include this type of publication in your study. Determine early on whether this is relevant or not and create your inclusion and exclusion criteria with this in mind. For more detailed discussions about grey literature in literature reviews, see the websiteof the Karolinska Institutet Library.

The standard for reporting a systematic review is PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses). By following PRISMA in reporting, you ensure that the minimum requirements are met for the study to be repeatable. The PRISMA Statement consists of a checklist and a flowchart and can be supplemented with specific checklists for e.g. protocols and search.

See:

PRISMA checklist: A guideline with 27 items for reporting
PRISMA Flow Diagram: Visualises how references have been identified, included and excluded
PRISMA Extensions: For example PRISMA-P for protocols and PRISMA-S for searching