Systematic reviews: The question

• Has the question already been answered?
• What is the purpose of asking this question?
• What will this question add to the existing literature?
• Is it helpful to use a question framework such as PICO or SPIDER? 
• How will concepts in the question be defined? For example:
  • How will adult be defined? What age range is this? 18 to 64? 18 plus? 21 plus?
  • How will diabetes be defined? Does it include type 1, type 2, pre-diabetes?
• Does the question limit to a particular type of study, such as RCTs?
• Does the question involve quantitative, qualitative or mixed method publications? 
• What is in or out of scope? What will the inclusion and exclusion criteria be?
• Does the question involve a specific timeframe (limiting by date)? If so, why?

It is not essential to follow a question framework, but it may be useful. Remember that every concept in a framework is not always replicated in a search strategy, as it can make the search too tight. 

Some question frameworks:

• PICO - Patient | Intervention | Comparison | Outcome
• PECO – Population | Environment | Comparison | Outcome
• SPIDER - Sample | Phenomenon of Interest | Design | Evaluation | Research type
• SPICE - Setting | Perspective | Intervention | Comparison | Evaluation

Learn more:

• Framing a research question - University of Maryland Library
• Systematic Reviews for Health: 1. Formulate the Research Question - University of Tasmania Library

Library staff can assist with searching for existing review papers. Please note that this is a quick search where strategies are not provided. It is not a structured search for the systematic review itself.

• Request a standard literature search (SVHM staff only)

Many systematic reviews don't make it over the finish line. These are some of the reasons we have experienced as co-authors:

Review team cannot agree on details or keeps expanding the question

The problem:

• Can create a "fluffy" topic that is too wide, badly defined and understood, and which changes.
• Even slight shifts in the question are likely to require new search strategies and starting over.
• Consumes a great deal of time as each version of the question is explored.
• As the review becomes too difficult the team may lose interest in proceeding.

How to avoid:

• Don't start work on the review proper (including searches) until consensus has been reached. 
• Make a decision and stick to it.
• Write and register a protocol. Lock it in!

The question has already been answered before you publish

The problem:

• Another systematic review on the same topic may be discovered during the searches - either initially or during updates.
• If there are long delays (especially with screening) another review might be published first.

How to avoid:

• Carry out an initial scoping search for existing reviews before writing a protocol.
• Search PROSPERO for an existing protocol.
• Ensure that the team is committed to screening in a timely way to avoid delays.
• If another review is discovered towards the end of the process, consider whether there is anything new to add that is likely to be accepted by a journal. All may not be lost!