1.) Screening Tests
a.) Validity- Measure of ability of a test to correctly distinguish who does and does not have the disease. (a test that has sensitivity and specificity at 80%)
- Sensitivity: Ability of a test to correctly identify people who have the disease. ( correctly include everyone with the disease)
Sensitivity: Number of true positives/ Number of true positives and false negatives
- Specificity: Ability of a test to correctly identify people who do not have the disease. (exclude people without the disease)
Specificity: Number of True Negatives/ Number of true negatives and false positives
Test Results
True Positive: Have the disease and test positive.
False Positive: No disease, but test positive.
False Negative: Have the disease, but have a negative test.
True Negative: No disease and negative test.
Unimodal: 1 peak (traditional bell curve)
Bimodal: 2 peaks
2.) Issues with continuous variables and breakpoints:'
a.) Continuous variable:It can be divided multiple times; There is no clear break point (ex: height) (categorical: You have specific options; ex: male, female or small, medium, large)
3.) Sequential and Simultaneous Tests:
a.) Sequential: Do an initial test and then an additional test to follow up on the positive test results. You will see a reduction in the sensitivity of the test and an increase in specificity.
b.) Simultaneous: More than 1 screening test for same disease is performed at the same time. This leads to an increase in sensitivity and a decrease in specificity. (If you test positive on either test, then you are considered positive)
4.) Predictive Value: How confident are you that your results are correct?
- This is the chance that you really have a disease, given that you had a positive test or that you do not really have the disease given that you had a negative test.
a.) + Predictive Value: Chance that you have a disease, given that you had a positive test.
Number of true positives/ total number of positive tests (TP and FP)
b.) Chance that you don't have the disease, given that you had a negative test.
Number of true negatives/ TN and FN (total number of negative tests)
5.) Intra-Subject Variability: Amount of variation in repeated measurements in the same subject over time. (Lead to different results in the same individual) Consider conditions under which the test was performed.
6.) Inter-Subject Variability: Amount of variation amoung subjects.
7.) Kappa: Determines how much rater agreement can be explained by chance alone.
- Calculate Kappa: Determine chance of each rater assigning a particular score.
Kappa stat= (observed % agreement) - (% agreement expected by chance alone) / 100% - (% agreement expected by chance alone)
- Overall % agreement: Add the number in all cells in which people agreed upon and divide that sum by the total number of tests and multiply by 100.
- % Agreement: Total number of normal and abnormal tests that they both agree on/ Total number of individuals tested.
