Intro Slides Flashcards
What is prevalence?
why is it useful?
proportion of the population that have a condition
Can be useful for establishing a baseline estimate of the probability that the patient in front of you has a condition (i.e. “pre-test probability”)
Does prevalence = incidence?
not the same as incidence (new cases of a condition in a given period of time)
Prevalence: proportion of the population that have a condition
Sensitivity?
Sensitivity: In patients who have the condition, the probability that the test will be positive.
Specificity:
Specificity: In patients who don’t have a condition, the probability that the test will be negative.
What are likelihood ratios?
A means of adding evidence to our pre conceived notions about a patient condition
Number representing diagnostic usefulness of a test
Range: 0 to infinity
1: useless
> 1: rules in (the higher the better)
<1: rules out (the lower the better)
LR+: test is positive
LR-: test is negative
Special note about “+/-“ in LRS
Note: in “LR+” and “LR- “ the “+” and “-” signs are technically unrelated to whether the probability increases or decreases
no negative numbers
How to calculate LR +
LR+= sensitivity/(1−specificity)
How to calculate LR -
LR^−= (1−sensitivity)/specificity
The best ways to use liklihood ratios
In a case where there is a score, say 0-5 and if the patient has 3/5 symptoms we get an LR that reflects the probability to rule in or out a condition
combining clinical decision rules and LRs
Causal reasoning?
**look this up in notes (handwritten)
What is it best for?
Generating vs. refining and assessing hypotheses
Threshold model of treatment
deciding at what point you are comfortable treating or not treating a patient depending on the liklihood that they have a condition
this is personal
Premature closure
accepting a diagnosis before it is appropriately verified
Vocab :
Differential diagnosis (DDx)
“Red flag”
Emergent
- look up in handwritten notes
