Evidence Based Medicine Flashcards
EBM
integration of clinical expertise, patient values, and the best research evidence into the decision making process for patient care
clinical expertise
clinician’s cumulated experience, education, and clinical skills
steps in practicing EBM
- define the clinical question
- find the evidence
- analyze the evidence
- summarize & apply the evidence and assess the outcome
EBM practice requirements
Assess Ask Acquire Appraise Apply
assess the patient
a clinical problem or question arises in the care of the patient
ask the clinical question
utilize PICO
PICO
P - patient or population (age/sex/race)
I - intervention (tests, therapy, etc)
C - comparison group (compare to nothing, to a placebo, or another intervention method)
O - outcome of interest (what do you hope to accomplish? what is the clinical effect?)
acquire the evidence
select the appropriate resources and conduct a search
appraise the evidence
for validity and applicability
apply the results
integrate evidence with clinical expertise and patient preferences and apply it to the practice of medicine
benefits of a well formulated question
save time by finding evidence that is relevant to patients’ clinical needs and directly address our knowledge needs
allows high-yield search strategies
clearer communication when sending or receiving a patient in referral
evidence pyramid
1 - background information 2 - case controlled studies 3 - cohort studies 4 - randomized controlled trials 5 - critically-appraised individual article 6 - critically-appraised topics 7 - systematic reviews
case controlled studies
are retrospective observational studies of a particular diagnosis or exposure
can establish correlations but cannot prove causation
cohort studies
compare two groups of subjects over time, one having a particular condition or receiving a particular treatment, the other not
used when a randomized controlled trial would be unreasonable or unethical
randomized controlled studies
randomly assigns subjects into treatment group or a control group to compare efficacy of a treatment, diagnosis, or drug
critically appraised topics
synthesize results from several studies on a particular topic and provide the best available evidence to a specific clinical question
systematic reviews and meta-analysis
comprehensively synthesize and critically appraise information from various articles that meet specific selection criteria
provide recommendation based on the best evidence available
bayes’ theorem
calculates the probability of a future event given already occurred events (IE: the prob that a bucket of balls is all red (A) given the first three balls drawn were red (B))
p(A|B) = p(A + B)/p(B)
statistical power
ability of a study to detect a (statistically significant) difference, if it really exists
poorly designed study = low power study
regression to the mean
In an experiment, if the first value is extreme, the next measurement will be closer to the mean (or vice versa). This seems to show a significant change (wrt studies on treatment potential), but is actually completely random
research questions should be
specific, answerable, novel, and relevant
research q types
descriptive
evaluative
explanatory
interventional studies
- test the effect of treatment or method of care
- controls not always used, but allow for comparative efficacy
- randomization is the best way to ensure comparisons are concurrent and unbiased
why randomize?
- removes bias of subject characteristics
- provides for fair conclusions
- allows blinding
- prevents sicker patients from being given particular treatment
