This document outlines the steps of the evidence-based medicine process which includes assessing the patient, asking a question, acquiring resources, appraising the evidence, and applying to the patient. It discusses how to formulate a well-built clinical question using PICO(TT) and describes the different types of study designs and evidence hierarchies. Guidelines are provided for evaluating the quality of individual studies, such as using the FRISBE checklist, and for critically appraising systematic reviews and meta-analyses.
2. The Steps in the
Evidence Based
Medicine Process
ASSESS the patient
ASK the question
ACQUIRE the resources
APPRAISE the evaluation
APPLY to the patient
4. PICO or PICOTT –
How to ask the question
P – Population, patient, or patient problem
I – Intervention, treatment, or exposure
C – Comparison of a 2nd intervention,
treatment, exposure
O – Outcome
TT – What type of question do I have, what
type of study do I need?
6. Study Types
•Etiology – Causes of a health issue
•Diagnosis – Identification of a health issue
•Therapy - Treatment of a health issue
•Prognosis – Probability of recovery
7. Mining the Article
Databases
With a well-formed PICOTT question, now
you are ready to start searching for and
collecting groups of results
9. Evidence Hierarchy
•Systematic review
• extensive literature search
•Randomized, controlled clinical trials
•randomly assigns exposure and moves forward; harder to
setup/manage – ethical issues
•Cohort study
• large population and follow forward from exposure time
10. Evidence Hierarchy
•Case controlled study
•retrospective comparison of exposures of patients
with control group; patients have the outcome and
have to reflect/remember back
•Case report
•reports of treatment of individual patient(s) without
control groups
11. Evaluating the Quality of
Individual Articles
With a smaller list of key articles, now you
need to use a fine tooth comb to evaluate
if the article results are reliable and
applicable to your patient
12. FRISBE –
How to Check for Quality
F – Follow-up
Does the study begin and end with the same
number of participants? If not, do the authors
acknowledge why there is a change? These factors
may skew stats
R – Randomization/concealed allocation
Did everyone in the study have an equal chance of
getting into the control or the test group? The
randomization process should be concealed from
researchers to avoid any bias
13. FRISBE –
How to Check for Quality
• I – Intention to treat
• Although some patients may refuse or be
noncompliant with treatment (real or placebo), you
still need to count them in the results and note the
intention
• S – Similar at baseline
• Patients should be similar so you don’t compare
very different populations (imagine athletes vs
morbidly obese people testing a diet pill)
14. FRISBE –
How to Check for Quality
• B – Blinding
• In the healthcare setting, the health care providers must
be blinded from knowing who is getting the test and
who is getting the placebo to avoid bias
• E – Equal treatment
• Treatments (other than the test) should be the same
across groups, otherwise you can’t determine if that
specific intervention did or didn’t work or if other
factors affected the results
15. Questions for
Systematic
Reviews/Meta-Analysis
Was the question focused?
Was a thorough review of the literature
(including grey literature) completed?
Do they have clear validity criteria?
Are the assessments reproducible?