Professional Documents
Culture Documents
Searching Clinic
Searching Clinic
How?
Asking Answerable Clinical
Questions
Patient
(For which patient/population or problem do you need
information?)
Intervention
(What medical event do you need to study the effect of?)
Comparision
(What is the evidence that the proposed intervention
produces better or worse results than no intervention at
all, or a different type of intervention?)
Outcome
(What is the effect of the intervention?)
Where to find the evidence for
your question?
Books
Library databases
Online bibliographic databases
Cohrane Database of Systematic Reviews
Searching for the Evidence
US National Library of Medicine:
24,000 journals
4,000 indexed in MEDLINE
worry that
significant evidence may be outside major databases
Missed key document may downgrade usefulness of review
Electronic sources of evidence
Source Availability Advantages Disadvantages
Critically http://cebm.jr2.ox.ac.uk Pre-appraised Only one study per
appraised www.eboncall.org summaries for a CAT; time limited;
topics (CATs) clinical question quality control
• Thesaurus searching
• Textword searching
Define keywords
MeSH terms
Textwords
• Non-Hodgkin’s Lymphoma
• • Important keywords not
Non-Hodgkin Lymphoma
included in the MeSH
• Lymphoma, non-Hodgkin tesaurus
• Lymphoma, non-Hodgkin’s
Check whether your keywords
are listed in MeSH
– AND
– OR
– NOT
Strategy of using Boolean operators:
first increase sensitivity, then specificity
1. One keyword concerning the patient
2. Another keyword concerning the patient
3. #1 OR #2
4. One keyword concerning the intervention
5. Another keyword concerning the intervention
6. #4 OR #5
7. One keyword concerning the outcome
8. Another keyword concerning the outcome
9. #7 OR #8
10. #3 AND #6 AND #9
Sensitivity and Specificity
vaccin*
vaccine, vaccines, vaccination
seizure$
singular and plural
Wildcards
gyn?ecology
gynaecology, gynecology
randomi?*
randomisation, randomization,
randomised, ...
How to increase specificity?
1. Use thesaurus to identify more specific
headings.
2. Use more specific terms in textword
search.
3. Use Boolean AND.
4. Limit the search by publication type, year
of publication, patient’s age, etc.
Limits
• Type of publication
• Age group
• Gender
• Year of publication
• Language
Identify the type of article
• Diagnosis
• Prognosis
• Therapy/prevention
• Harm/etiology
Identify the type of research that is most
likely to answer your question
Diagnosis cross-sectional
Prognosis cohort
Etiology/harm 1) cohort
2) case-control
Therapy/prevention 1) randomized control trial
2) non-randomized control trial
3) cohort
4) case-control
Patients' experiences qualitative studies
BASIC STEPS
?
Questions?
I. FORMULATION OF
SEARCH QUESTION
… helps to
obtain relevant answer
understand question & concentrate mind
identify relevant path of sources to find
information
better formulated the question,
more relevant the answer
e.g …….CLINICAL QUESTION
DEFINING THE QUESTION-
BREAKING INTO CONCEPTS
Patient / population (patient with condition x)
Intervention/exposure/prognostic factor
Comparison (usual care/no intervention)
Outcome (did it work? which works better?)