Flinders University Library
Search Smart

Evidence Based Medicine (EBM)

This guide is designed to walk you through the Evidence Based Medicine process: the elements of a well-formulated clinical question, types of studies, and the key critical appraisal questions that help determine the validity of evidence.

Harm/Aetiology question

Clinical scenario

Sally, an 18 year old female, has just been diagnosed with a deleterious mutation of the breast cancer susceptibility gene 1 (BRCA1). She has recently started taking oral contraceptives. She asks you if taking oral contraceptives will increase her risk of early onset breast cancer and whether she should change to another form of contraception.

PICO

  • P: Young females with an inherited BRCA1 gene mutation
  • I: Oral contraceptives
  • C: No oral contraceptives or another form of contraception
  • O: Increased risk of early onset breast cancer

Clinical question

In young women with an inherited BRCA1 gene mutation, will taking oral contraceptives increase the risk of early onset breast cancer?

Logic Grid

Concept 1:
Young women
Concept 2:
BRCA1 gene mutation
Concept 3:
Oral contraception

 

  • Young female(s)
  • Young women

 

  • BRCA1 (hereditary/germline) mutation
  • Oral contraception
  • Oral contraceptives

 

Medline search strategy

Search uses both MeSH terms and keywords and applies the following search limits:

  • English language
  • Reviews (maximises specificity) Clinical Queries filter
  • Causation-etiology (maximises specificity) Clinical Queries filter

PubMed Clinical Queries search strategy

Go to Clinical Queries and enter search string: (BRCA1 AND breast AND (cancer* OR carcinoma* OR neoplas*) AND contracept*)

Choose Etiology category and Broad scope for best sensitivity where there appear to be few studies.

Important to note:

  • Use brackets to separate strings of synonyms combined with OR from the AND operator
  • Operators AND/OR must be in uppercase in PubMed and PubMed Clinical Queries
  • You can exclude citations already found in Medline by adding NOT Medline[sb] to the end of the search string. This reduces the search set to newly published studies and reviews possibly not yet available in the Medline database. 

E.g. (BRCA1 AND breast AND (cancer* OR carcinoma* OR neoplas*) AND contracept*) NOT Medline[sb]

This only finds one additional clinically relevant study: Risk-benefit assessment of the combined oral contraceptive pill in women with a family history of female cancer.

Scopus search strategy (will also work in Web of Science)

Enter search string in Scopus search box: BRCA1 AND breast AND (cancer* OR carcinoma* OR neoplas*) AND contracept*

Important to note:

  • Double quote marks are needed around phrases in Scopus to keep words together
  • Deselect subject areas Physical Sciences, Social Sciences & Humanities, and Life Sciences but keep Health Sciences selected for more precise results.

The search found several promising systematic reviews:

These will need to be critically appraised using the following appraisal questions for systematic reviews:

1. Were the methods used in the review valid?

  • Did it address a clearly focused question in terms of participants, interventions, comparisons, and outcomes (your PICO). Did it have clearly defined eligibility criteria?
  • Did the review include high quality, relevant studies
    • of robust study design (for example, randomised controlled trials)?
    • that addressed relevant questions?
    • selected by more than one independent reviewer?
  • Is it unlikely that the review missed important, relevant studies? 
    • Is the search strategy reproducible?
    • Did the search strategy include synonyms for all concepts?
    • Is the search strategy sufficiently comprehensive? 
  • Did the review include an assessment of the risk of bias of included studies and was this assessment incorporated into the review finding?
    • Is the method for assessing risk of bias reproducible and accurate?
    • Was the risk of bias assessment conducted by more than one independent assessor?
    • Was it incorporated into the review findings?
  • Did the review combine the results from studies and, if so, was it reasonable to do so?
    • Were the results similar from study to study (was an examination fo heterogeneity undertaken)
    • Were reasons for variations in results discussed? 

2. What are the results of the review?

  • What are the overall results of the review? 
  • How precise are the results? 

3. How relevant are the results to me? 

  • Were all important outcomes considered (from the points of view of individuals/patients, policy makers, healthcare professionals, family/carers, the wider community)?
  • Can the results be applied to my patient/s? Consider how well your patient(s) and practice setting compare with the review population.


Several individual harm
studies, such as the following case-control study, were also identified

In appraising a study about harm or aetiology, it is important to consider the following questions:

  1. Were there clearly defined groups of patients, similar in all important ways other than exposure to the treatment or other cause?
  2. Were treatments/exposures and clinical outcomes measured in the same ways in both groups?
  3. Was the follow-up of the study patients sufficiently long for the outcome to occur and complete?
  4. Do the results of the harm study fulfill some of the diagnostic tests for causation?
    • Is it clear the exposure preceded the onset of the outcome?
    • Is there a dose-response gradient?
    • Is the association consistent across studies?
    • Does the association make biological sense?

Reference

Straus SE, Glasziou P, Richardson WS, Haynes RB. Evidence-based medicine: how to practice and teach it. 4th ed. Edinburgh: Churchill Livingstone Elsevier; 2011. 293 p.