Cardiology Board Review statistics

Editor-In-Chief: C. Michael Gibson, M.S., M.D. [1]

Bayes Theorem[edit | edit source]

• The pretest probability of disease affects utility of the test
• Low pretest probability of disease: Negative result virtually excludes disease. Positive test not that helpful, triggers more tests.
• High pretest probability: Negative test not helpful.
• Intermediate Range Probability of Disease: Tests are most useful in identifying patients with the disease.
• Likelihood ratio (LR) = sensitivity / 1 - specificity
• Pretest probability X LR = usefulness of the test

Incidence[edit | edit source]

New cases per year or over a period of time

Prevalence[edit | edit source]

How many people have the disease at a snapshot in time

False Positive[edit | edit source]

Don't have the disease, but test indicates they do

False Negative[edit | edit source]

Have the disease, but tests says they don't

Positive Predictive Value[edit | edit source]

Proportion of people with disease who have a positive test. If you have a low prevalence of disease, this may be low. If you have a prevalent disease, this will often be positive.

Negative Predictive Value[edit | edit source]

Proportion of people with a negative test don't have the disease. If the prevalence of disease is low, this is helpful (atypical CP and a negative stress test rules out the disease).

Metaanalysis[edit | edit source]

• NNT, number needed to treat, number need to treat to save one life, or number of lives saved in treating a thousand patients

References[edit | edit source]