Calculating Flashcards
Lectures: -L10 - Diagnostic testing
What are reasons that real life test performance can differ from test performance reported by the manufacturer? (5)
- Bad quality product
- Bad lot of products
- Performance measured on a specific population/time point
- Different sample type used
- Inter-institute variation
How can a bad lot quality be detected?
Whenever new products come in, ISO-accredited labs determine test characteristics & performance
What is the definition of test sensitivity?
The ability of a test to correctly identify those with a disease
What is the calculation for test sensitivity?
TP/TP+FN
What is the definition of test specificity?
The ability of a test to correctly identify those without the disease
What is the calculation for test specificity?
TN/TN+FP
What do the positive predictive value/negative predictive value depend on?
Disease prevalence within the population
What is the positive predictive value?
Proprotion of positive results that are true positives
What is the negative predictive value?
Proportion of negative results that are true negatives
What happens to PPV when disease prevalence increases? Why does this happen?
PPV increases
The higher the disease prevalence, the more people are ill -> even low sensitivity tests will pick up quite a lot of people
In which situations can tests with (relatively) low sensitivity be used?
In diseases with a high prevalence, because PPV will still be high
What happens to NPV when prevalence increases? Why?
NPV decreases
Tests with a high specificity still misdiagnose TN people as FP when they are run often enough
