mjl1717 Forum Hero

Topics: 955 Posts: 5,450
| | 11/11/04 - 11:14 AM  
 
   
 
|   #1 |
In your own words then in mathematical terms can someone define sensitivity and specificity.
___________________ Smell the coffee! "Is That an Osler move??"
|
| VM Forum Junior
Topics: 4 Posts: 65
| | 11/11/04 - 12:13 PM  
 
   
 
|   #2 |
sensitivity is basically how "sensitive" ( how acurate in predicting) the test is in finding one with disease. if there are 100 peopel with disease(A+C) and you test all of them with your new test and 90 turn out positive(True Positive A) which means 10 are negative (false negativeC) then your sensitivity is 90%. or (A/A+C) ( think she is very sensitive.. any lil sound can wake her up.) Specificity .. think NO MEANS NO.. thats how specific you want the test to be... so if you test 100 non diseased patients(B+D) with your new test and the test is positive in 5 ( false positive B) and negative in 95 ( true negative) then your specificity is 95%( D/B+D) If you have a deadly disease you want your screenign test to be highly sensitive since you dont want to miss any cases.. so HIV ELISA or most cancer screening tests or any rapidly deteriorating conditions. But if the disease requires radical treatment options then you want to be SURE ( SPECIFIC) that the patient actually has the disease.. so western blot in HIV. I hope I helped you and not confuse you more.. VM
|
|
| |
| | |