-
Sensitivity
- finds disease ...best used to screen
- Excludes disease if test negative, Includes disease is test is positive
- TP/TP+FN
-
Specificity
- How good in normal population
- used to confirm disease
- TN/TN+FP
-
Reference interval
+/- 2SD = 95% of normal
-
Negative Predictive Value
- How likely the negative result is a true negative and not a false negative
- - with a 100% sensitive test = 100% b/c there are no false negatives
- -with a 100% specific test = could be TN or FN does not exclude disease
- -TN/TN+FN
-
Highly sensitive test
...increases prevalance of disease
-
Increasing upper limit of normal of a test reference interval
- increases specificity and PPV
- decreases sensitivity and NPV
-
decreasing upper limit of normal on test reference interval
- increases sensitivity and NPV
- decreases specificity and PPV
-
Prevelance
- Prevelance = incidence x duration
- TP+FN/ TP+FN+TN+FP
- -prevelance is how many people have the disease right now. It increases as the duration of a disease increases.
- -incidence is how many people develop a disease in a population over a set time, usually one year.
-
Positive predictive value
- how likely a positive result is a true positive
- -TP/TP+FP
- -100% sensitive test = result could be TP or FP includes people with disease but does not confirm
- -100% specificity = 100% b/c there is no false positives, confirms disease!
|
|