Campbell Collaboration

Formulating the research question

—

by

Campbell Collaboration

By Ciara Keenan
This article was originally posted on the Meta-evidence website on 7 March 2018.

Any systematic review should begin with a well formulated and answerable research question. This question might be formed naturally through a reviewer’s knowledge of an unanswered issue in their field, or through the strategic use of evidence and gap maps which will identify where robust evidence is lacking in a particular area.

Either way, translating this gap in evidence into a well-defined research question is a necessary skill. In a systematic review methodology, there are no hard and fast rules on whether the research question must be narrow or broad in focus, and there are advantages and disadvantages to both. Usually the degree of focus will depend on the answer you seek, the amount of primary research available for review, and the resources available to the review team.


Focused and broad research questions

Using PICO to define the question

The research question should be presented in a way that will help the reader identify the key components of the study. The most widely used process to identify important and appropriate components in a research question is the mnemonic PICO. Although permutations exist (alternatives provided in parentheses), the usual variables which make up PICO are problem (or population or patient), intervention (or exposure), comparison, and outcomes. Some researchers may encounter questions which are not easily adapted to the PICO framework, Andrew Booth has provided a useful list of alternatives.

An example research question

A question that is clear, structured, and objective will also allow the reviewer to build a search strategy with reduced bias and helps to avoid missing potentially relevant studies.

Check the question is new

To avoid duplication of effort, you should always ensure the research question has not been addressed previously through a systematic review. To determine this, researchers can check the following sources (this list is not exhaustive):

If you find that your question has been answered previously by a systematic review you may be justified in conducting a similar review if there is the potential of your review finding additional papers due to:

  • Previous reviewers using a different setting of interest
  • Your inclusion of additional study designs
  • The quality of search strategy used on previous reviews was poor

Register the title

Registering your research question is an important part of the systematic review process. If the title is registered and published early in the project it will help reduce duplication of effort, and will ensure that the editors of the coordinating group can intervene early if the research question is not appropriate for systematic review, thus reducing wasted resources. The title registration phase of a review will ask for details around the scope of the review, the review team and their expertise, and an outline of the methods to be used.

Top tips

Tip 1: Understand the gaps that exist in your area of expertise

Tip 2: Use PICO to define the research question

Tip 3: Check that the question is unique

Tip 4: Register the title

Meta-evidence is a blog for interviews and tips on evidence synthesis brought to you by Campbell UK & Ireland.