VU Systematic Literature Reviews: Define question

Define your question with a framework

The first step of the systematic review process is to develop a clear, well-formed, and focused question.  This will make it easier to apply the key concepts in your question to your search and will ultimately make searching for evidence more straightforward.

Watch the video below to learn more about developing a research question.

Video Length: 6:26

Key points from the video:

 

  • All good academic research starts with a research question.
  • A research question is an actual question you want to answer about a particular topic.
  • Developing a question helps you focus on an aspect of your topic, which will streamline your research and writing.
  • To create a research question:
    • Pick a topic you are interested in.
    • Narrow the topic to a particular aspect.
    • Brainstorm some questions around your topic aspect.
    • Select a question to work with.
    • Focus the question by making it more specific. Make sure your question clearly states who, what, when, where, and why.
  • A good research question focuses on one issue only and requires analysis.
  • Your search for information should be directed by your research question.
  • Your thesis or hypothesis should be a direct answer to your research question, summarised into one sentence.

Hall, L. A. (2019, June 3). How to write a research question for your literature review [Video]. YouTube. https://youtu.be/39bLneUxGwA

The way you frame your question is critical.

Watch the video below to hear Dr. Saravana Kumar, Senior Lecturer in Physiotherapy, UniSA, explain the different question types and the body of evidence required for each.

Video Length: 3:44

Key points from the video:

 

  • The success or failure of your systematic review is heavily influenced by the design of the question.
  • There are different types of questions.
  • When developing your research question, you need to know whether it is an intervention, etiological, prognostic, diagnostic or screening intervention type question.
  • Intervention questions look at the effect of a treatment on different outcomes.
       For example, What is the effect of taking Panadol on headaches?
  • Etiological questions look at the causative factors of a disease.
       For example, What is the effect of smoking on lung cancer?
  • Prognostic questions explore the likelihood of particular outcomes for patients with particular disorders.
       For example, What is the prognosis for men diagnosed with prostate cancer?
  • Diagnostic questions compare the accuracy and safety of diagnostic tests against the standard method. 
       For example, How effective is x-ray compared to MRI in diagnosing muscle or ligament tears?
  • Different question types are best answered by different types of evidence (i.e. study types).

University of South Australia (n..d.). Different types of questions [Video]. Panopto. https://unisa.au.panopto.com/Panopto/Pages/Viewer.aspx?id=a97be758-b03c-40fa-831e-aa6900308f15&start=0

Background questions

These ask for general knowledge about a condition or disease. They are:

  • not related to an intervention, but useful to develop a better understanding of a condition.
  • asked because of the need for basic information, not normally to make a clinical decision about a specific patient.

Foreground questions

These ask for specific knowledge to inform clinical decisions or action. They use a structured framework, such as: PICO(T), or PECO(T). Some examples:

  • Does pole dancing improve aerobic fitness in middle aged adults?
  • In older adults with knee osteoarthritis are NSAIDS more effective than exercise in reducing pain?
  • Is a paleo diet more effective than a Mediterranean diet for weight reduction in obese adults?
    • What is the evidence for the effectiveness of self management of diabetes?

Before you formulate your review question, it is important to consider the FINER criteria (Feasible, Interesting, Novel, Ethics and Relevant). This will help you fine-tune your question.

Select the headings below to read more about each of the FINER criteria.

  • Are you capable of addressing the question with the evidence available?
  • Have you done a preliminary search to identify a sensible boundaries for your review?
  • Will your review question potentially retrieve unmanagable quantities of information? 
  • Is your question answerable with existing evidence?
  • Do you find your topic interesting?
    You are going to spend a significant amount of effort and commitment in this review, so it is important that you are interested in the review question.
  • Does your question address a meaningful 'gap' in the knowledge?
  • Have you checked if there is a same/similar synthesis already published?
  • Have you checked if there is an on-going review in PROSPERO register?
  • Does your question have any political implications that may potentially broaden the health inequality?
    For example, have you explicitly described the effect of the intervention not only on the greater population, but also on disadvantaged groups or/and health inequality within the community?
  • Have you consulted with stakeholders in your field and defined the focus of your question?
  • Are the findings of your review relevant enough to facilitate others informing their decisions?

The PICO framework is useful for developing an answerable clinical question.

For example: Is a paleo diet more effective than a Mediterranean diet for weight reduction in obese adults?

P Population (patient)
What are the characteristics of the patient or population? OR
What is the condition or disease you are interested in?		 obese adults
I/E Intervention or Exposure
What do you want to do with the patient (e.g. treat, diagnose, observe etc.)?		 paleo diet
C Comparison
What is the alternative to the treatment (e.g.placebo, different drug, surgery)?		 mediterranean diet
O Outcome
What are the relevant outcomes (e.g. morbidity, complications)?		 weight reduction

Use the PICO worksheet and template to get started with your question. If you are using another framework, adapt it.

Watch the following video (created by the Centre of Evidence-Based Practice, Oxford University) to see PICO mapping in practice.

Video Length: 2:52

  • Consider the following scenario:

A GP with a special interest in smoking cessation wants to encourage his practice to engage with teenagers to stop them smoking. In order to persuade his colleagues, he wants to see what evidence is available, particularly on the effectiveness of brief intervention techniques.

  • PICO is a way of formulating your question and identifying potential search terms to retrieve the best set of results possible.
    • P stands for patient, population, or problem.
    • I stands for intervention.
    • C stands for comparison or control.
    • O stands for outcome.
  • Authors use different phrases to describe the same concepts, so it is necessary to think of alternative words to include in your search in order to capture all of the relevant literature.
  • In the example above, the P element is teenagers. They may also be described as adolescents or young people.
  • The I element is 'brief intervention'. This can also be called brief advice, brief counseling, or minimal advice.
  • The O element is smoking cessation, but authors may also refer to patients who stop smoking or quit smoking.
  • It's not always necessary to use all four arms of the PICO framework. This example does not use a control or comparator. 
  • Bringing all the elements together, we get the following focused question:

Can brief intervention be used as an effective smoking cessation technique with teenagers?

  • It is important to keep this question in mind at every stage of the search process. It is the key to retrieving a focused set of results for your search.

 

Additional videos:

A variant of PICO could be when you do not have a comparator:

For example: What is the evidence for the effectiveness of self management of Diabetes Mellitus Type II?

P Population (patient)
What are the characteristics of the patient or population? OR
What is the condition or disease you are interested in?		 Diabetes Mellitus Type II
I Intervention or Exposure
What do you want to do with the patient (e.g. treat, diagnose, observe etc.)?		 self management
O Outcome
The outcome is normally implicit so it may not be necessary to search for it.		 effectiveness

 

You can use PICO for an diagnosis question.

For example: In adults with lateral ankle sprains, how useful are x-rays in confirming the diagnosis?

P Population (patient)
What are the characteristics of the patient or population? OR
What is the condition or disease you are interested in?		 adults with lateral ankle sprains
I Intervention or Exposure
What do you want to do with the patient (e.g. treat, diagnose, observe etc.)?		 x-rays
O Outcome
The outcome is normally implicit so it may not be necessary to search for it.		 confirming diagnosis

If your research question is qualitative you may need to focus gathering of research which describes a particular phenomenon or topic of interest. You can adapt people in the following way: 

PICo

P=Population, I=Interest area, Co=Context. Useful for qualitative questions.

Qualitative example: What are women's experiences with bullying in the workplace?

P Population (patient)
What are the characteristics of the patient or population?		 women
I Interest
What is the phenomena that relates to a defined event, activity, experiences or process.		 experiences with bullying
Co Context
Refers to the context or the environment.		 workplace

Here are some different frameworks you may want to use:

PICO(T) Population (patient), Intervention, Comparison (control) and Outcome. Add a Timeframe if required. Used particularly for treatment type questions.
PECO(T) A variation of PICO where E= Exposure and T=Timeframe if required.
PCC Developed for a Scoping Review, where P= Population (or Participants), C=Concept and C=Context.
PIPOH Developed in the context of practice guideline adaptation. Includes P= Professionals/Patients, O= Outcome and H= Healthcare Setting.
SPICE S= Setting (where), P= Perspective (for whom), I= Intervention (what), C= Comparison (compared with what), E= Evaluation (Booth 2006).
SPIDER S= Sample, P= Phenomenon of interest, D= Design, E= Evaluation, R= Research type. Useful for qualitative or mixed method studies (Cooke, Smith and Booth 2012).
ECLIPSE E= Expectations, C= Client group, L= Location, I= Impact, P= Profession, SE= Service (Wildridge & Bell 2002).
PESTLE P= Political, E= Economic, S= Social, T= Technological, E= Environmental, L= Legal (CIPD 2010).

Preliminary searching

Before you finalise your question:

  1. Conduct some scoping searches to help define your concepts and identify additional terminology
  2. Check to see if a systematic review has been done before on your topic before
  3. Conduict preliminary searches to confirm that there is enough evidence (studies) to answer your question

Preliminary (or scope-out) searches will help you define your concepts, develop, focus and refine your question, and identify additional terminology used in the literature. 

You can search for existing systematic reviews in the locations linked below:

Watch the video below to learn more about preliminary searching and finalising your question:

Further resources

Test your knowledge

Test your general knowledge of systematic reviews by moving through the slides in the activity below: