Skip to Main Content

Evidence-Based Medicine (EBM)

Research material

Outline

A common evaluation tip for journal articles is to check whether or not the paper has been subjected to the rigour of a Peer Review Process. This means that the content of the paper has been subjected to an evaluation and checking process before publication.

There are two types of Peer Review, Open and Closed.

  • Open - this means that identities of authors and reviewers are known to each other.
  • Closed- there are two categories of closed review:
    • single blind peer review - where the identity of the author is known to the reviewer, but the identity of the reviewer is not known to the author
    • double blind per review - the identity of the author is not known to the reviewer, nor is the identity of the reviewer known to the author

Journals - peer review process pathway

 

Peer Review - Strengths and Limits

Whilst peer review indicates that a paper has been carefully scrutinised externally, it is not a guarantee of the quality of the research concerned.  Further checks in the form of critical appraisal should be applied.  See the later section in this guide on Reporting and Quality (risk of bias).

With acknowledgements to Paula Anne Goodall of Library Services, University of Birmingham, and Dr Clare Davenport, Institute of Applied Health Research, University of Birmingham.

Primary (unfiltered) and Secondary (filtered) information

In terms of literature searching to support EBM, it is important to note the difference between primary (or unfiltered) and secondary (or filtered) information. 

  • Primary material may be an article writing up the details and results of a trial for example, whereas secondary material may involve identifying or analysing the article alongside others on the same or a similar topic
  • Systematic reviews are regarded as being the most authoritative secondary material, as they involve an exhaustive search and study of literature on the topic concerned
  • Randomized controlled trials (RCTs) likewise are foremost in the primary literature, as they are about trials carried out with carefully-controlled methods and contexts

Filtered or secondary information: 

  • Systematic reviews; critically-appraised topics; critically-appraised individual articles
    • (See CASP (2020) for information about critical appraisal)

Unfiltered or primary information:

  • Randomized controlled trials; cohort studies; case-controlled case series, or reports; background information or expert opinion

The diagram below shows these different types.  

Research design and evidence:hierarchy or pyramid of evidence in medical literature.

Pyramid of evidence

CFCF (2015)

References

CASP (2020) Critical skills appraisal programme.  Available at: https://casp-uk.net/ (Accessed 12 October 2020)

Image on this page

CFCF (2015) Research Design and Evidence. Available at: https://commons.wikimedia.org/wiki/File:Research_design_and_evidence_-_Capho.svg (Accessed 8 December 2023).  CFCF, CC BY-SA 4.0 CC By SA 4.0 licence

 

The sort of question you are asking determines the type of research to look for.

Different study designs are used to help answer different types of questions.

Clinical practice requires answers to a wide variety of questions, not just those about the effectiveness of interventions.  

For example, we may want to determine:

  • How common is a condition in order to plan services? (e.g. how many births are there per year in Birmingham?  How many midwives do we need for the population of Birmingham?)
  • What causes a condition in order to prevent it occurring or to treat it? (e.g. high blood pressure causes stroke and myocardial infarction)
  • How can we detect a condition early in order to offer timely treatment? (e.g. What is the sensitivity and specificity of mammography screening for detecting breast cancer?)
  • What is the prognosis associated with a condition in order to inform patients or plan services? (e.g. how many people survive stroke and how many survive with a disability?
  • What is the experience of patients using a service in order to improve the service? (e.g. Why don’t all eligible women attend for cervical screening?).

Each question will be best answered by a different type of study design.  

  • The examples on the next page do not represent an exhaustive list of the different types of studies you may come across but act as a guide to the types of study design most appropriate to different questions
  • You will note that systematic reviews of research are nearly always preferred – that is, a systematic collection of the best evidence available on a topic rather than a haphazard selection of one of more studies
  • However, it is not uncommon for the best type of study design not to exist for a particular question.  In these circumstances we are forced to rely on evidence from a study that is less appropriate or less valid for answering our research question

Acknowledgement: thanks to Dr Clare Davenport of the Institute of Applied Health Research, University of Birmingham, for content on this page.

Appropriateness of study designs to answer different questions

Ranked from A (best) to C.  Unsuitable designs for each question are excluded.

How common is a condition?

  • A: Systematic Review of cohort or cross-sectional studies
  • B: Cohort Study, Cross-sectional study

What causes this condition?

  • A: Systematic Review of cohort or case-control studies
  • B: Cohort Study
  • C: Case-control study

What is the prognosis of this condition?

  • A: Systematic Review of cohort studies
  • B: Cohort Study

What is it like to live with this condition?

  • A: Systematic Review of qualitative research
  • B: Qualitative Research

How do I best diagnose this condition?

  • A: Systematic Review of test accuracy studies
  • B: Special type of Cross-sectional Study (a test accuracy study)
  • C: Randomized Controlled Trial of test accuracy

Is treatment ‘x’ effective for this condition?

  • A: Systematic Review of Randomized Controlled Trials and other Controlled Trials
  • B: Randomized Controlled Trial
  • C: Controlled Trial

Are treatments for this condition acceptable to patients?

  • A: Systematic Review of Randomized Controlled Trials, other Controlled Trials and Cohort Studies
  • C: RCT, Controlled Trial, Cohort Study

Adapted from Muir Gray, J. A. (2009). Evidence based healthcare. London: Churchill Livingstone. 3rd edn.

Concise definitions of Study Designs can be found in the following:

Georgia State University Library (2022) Evidence-based medicine: study designs. Available at: https://research.library.gsu.edu/c.php?g=115558&p=1966291 (Accessed 6 October 2022)

Institute for Quality and Efficiency in Health Care (2016) What types of studies are there? Available at: https://www.ncbi.nlm.nih.gov/books/NBK390304/ (Accessed 6 October 2022)

For more detailed reading:

Greenhalgh, T. (2019) 'Chapter 3. Getting your bearings: what is this paper about?' In How to read a paper: The basics of evidence-based medicine and healthcare.  Oxford: Wiley.  [UoB: available on FindIt@Bham as ebook]

Muir Gray, J. A. (2009) 'Chapter 5. Appraising the quality of research' In Evidence-based healthcare and public health. 3rd ed. Edinburgh: Churchill Livingstone. [UoB: available in print at Barnes Medical Library].
 

Accessibility statement