The flashcards below were created by user
on FreezingBlue Flashcards.
What's Pre-test probability?
probability of the disease before a test is done (basically the prevalence)
What's Post-test probability
- based on outcome of the test, the likelihood that the
- patient has the disease (rule in/out)
What's the table look like for Likelihood Ratios?
What's a positive likelihood ratio?
ration or proportion of those with disease & test positive: without disease with positive test result (1-specificity)
What's the calculation for the positive likelihood ratio?
What's a Negative likelihood ration?
Ration = those with disease & negative test result : without disease and negative test result (specificity)
What's the equation for the negative likelihood ratio?
What are the components of framing a clinical question?
- P - patient/population being addressed
- I - intervention in question
- C - comparison (when relevant)
- O - outcome of outcomes of interest
What's the significance of LR >10 or <0.1?
- Generate large and conclusive changes from pre-test and
- post-test probability.
What's the significance of LR of 5-10 and 0.1-0.2?
Generate moderate shifts in pre-test to post-test probability.
What's the significance of LR of 2-5 and 0.5-0.2?
Generate small (but sometimes important) shifts from pre- to post-test probability.
What's the significance of LR of 1-2 and 0.5-1?
Alter probability to a small (a likely insignificant) degree
What's a decision analysis?
Formal structure integrating evidence (beneficial/harmful effects) with values, preference, probablilitys, costs associated
What do squares, circles, and triangles represent on a decision analysis?
- Square- decision
- Circle - chance events
- Triangle - outcome nodes
When are dignostic tests useful?
When there are large changes between pre-test and post-test probability; validity compared to 'gold standard'
What factors affect predicitve valies the most?
prevalenceof disease, sensitivity, specificity
What are the advantages of LRs?
- Contribution of test result as a single number
- Determining post -test probablilty given a result
- Summarizing information
- Describing probability of disease
What's the effect of having a low cut-off level on sensitivity/specificy?
What's the effect of raising the cut-off level on specificity/sensitivity?