Systematic Reviews

Developing a question will help define the scope of the review and it will help with developing key concepts to be used in your search strategy.  The question should be open rather than a statement and not the question that would lead to a ‘yes’ or ‘no’ answer. The question should be specific and have three or four elements as suggested by frameworks outlined below.

This reading explains the PICO, PICOS and SPIDER frameworks with examples:

PICO, PICOS and SPIDER: a comparison study of specificity and sensitivity in three search tools for qualitative systematic reviews (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4310146/)

Here are some frameworks for developing a research question:

PICO is a helpful framework for developing a systematic review question. It is used in health and medicine disciplines. 

P -  Patient or Population: What are the defining characteristics of the patients or population?

I -  Intervention:  What is the treatment for the patient or population?

C - Comparison: Is there another treatment method that you would like to compare the intervention to?

O - Outcome:  What is the outcome of the intervention?

SPIDER framework is focused on qualitative research.

Sample - group of people being studied

Phenomenon of Interest - looks at the reasons for behavior 

Design - type of research such as interview or survey

Evaluation - outcome

Research Type - qualitative, quantitative or mixed methods

Creating a search strategy is an important part of the systematic review and is typically reported in the methods section of the paper. A strong search strategy will retrieve the majority of the studies you will assess for eligibility and inclusion. The quality of the search strategy is important because it can affect literature that may have been missed.

Search Terms and Subject Headings

A good search strategy should include a combination of keywords and subject headings in order to maximize the number of studies found.

Keywords help broaden results. A search in a database will look for keywords in journal titles, abstract and journal text.

Subject Headings help focus a search. A search in a database will look for a specific subject term applied by the indexer.

Combining Search Terms Using Boolean Operators

Search Strategy Example

An example of a search string for one concept in a systematic review. The two examples shown are from a PubMed search. [MeSH] means Medical Subject Headings and [All Fields] searches for the concept in the full text of the article. This is just an example to show that as many variations of a concept should be included in a search string in order to find all the relevant literature.

Concept 1: Cognitive behavior therapy

"cognitive behaviour therapy"[All Fields] OR "cognitive behavioral therapy"[MeSH Terms] OR ("cognitive"[All Fields] AND "behavioral"[All Fields] AND "therapy"[All Fields]) OR "cognitive behavioral therapy"[All Fields] OR ("cognitive"[All Fields] AND "behavior"[All Fields] AND "therapy"[All Fields]) OR "cognitive behavior therapy"[All Fields]

Concept 2: Depression

"depressive disorder"[MeSH Terms] OR ("depressive"[All Fields] AND "disorder"[All Fields]) OR "depressive disorder"[All Fields] OR "depression"[All Fields] OR "depression"[MeSH Terms]

Search Filters, Inclusion, Exclusion Criteria

Databases have search filters allowing you to narrow results in order to retrieve articles most relevant to you (e.g. publication dates, species, language, sex, subject, ages, etc.). In a systematic reviews search, filters should be used with care as you may lose relevant research unintentionally.

For further information, check out the "Introduction to Search Strategies" chapter from the Cochrane Handbook for Systematic Reviews and Interventions (https://training.cochrane.org/handbook/current/chapter-04#section-4-4).