EBM Flashcards
usefulness =
(relevance x validity) / (work)
unacceptable loss to follow up
> 20%
NNT =
100 / (% treatment - % control)
summary review
- cover full breadth of topic
- overview of disease etiology, diagnosis, prognosis, management
- address a number of questions
- good for background questions
synthesis review
- defines 1 or 2 specific questions
- carefully finds all available evidence, evaluates validity, and reports answer to Q
- useful for foreground questions
meta-analysis
- statistical technique
- combines findings from independent studies
- treats data as if from 1 big study
- usually only done with RCTs
degree of inconsistency
<40% : not important (good)
30-60%: moderate heterogeneity
50-75%: substantial
75-100%: considerable
post-test probability =
pre-test probability, given the test result
THE SAME TEST WILL PERFORM DIFFERENTLY GIVEN BACKGROUND PROBABILITY
post-test odds =
pretest odds x LR
basic criteria
-test is sufficiently sensitive and specific
minimally useful
- test changes diagnosis
- label patient
more useful
-test changes treatment
very useful
test changes outcome
maximum benefit
test is worthwhile to patients/society
authority based guidelines
BOGSATS
- consensus guidelines
- low quality
evidence based guidlines
- evidence based
- BUT limited by lack of explicitness
evidence linked guidelines
-identified by a methods section explaining how the evidence was determined, graded, and used to inform the decision making guidelines
