Sensitivity- Ability of test to correctly identify those with disease x
What is the equation for sensitivity?
(a/a+c) x 100
(a/a+b) x 100
Specificity
Ability of new test to correctly identify those without disease
Test that best predicts disease prevalence
Probability that person with +ve result has disease
The gold standard is the test that best predicts disease prevalence
Negative predictive value (probability that person with -ve result has the disease) =?
(d/c+d)
(d/b+d)
When identifying people with a very high risk disease, what tests do you use, and in what order?
First one that's highly sensitive (so we don't miss anyone who does have it) and then one that's highly specific to remove those that don't
First one that's highly specific (to not worry people who don't have it unecessarily) then one that's highly sensitive
Prevalence = Those with disease/Total population