ing Forum Guru

Topics: 45 Posts: 435
| | 03/12/06 - 06:20 PM  
 
   
 
|   #1 |
This is from Kaplan. Positive predictive value is mainly affected by two variables, namely the accuracy (combination of sensitivity and specificity) and prevalency of the disease. For example, if a test has a sensitivity of 99% and specificity of 99% (i.e., a high degree of accuracy), and the disease has a prevalence of 1/10,000, the positive predictive value will be only 1%. On the other hand, if the same test is applied to a population in which the disease has a prevalence of 25/100, the positive predictive value will be 97%. Anybody know how the positive predictive value 1% and 97% caculated?
___________________ Never give up!!
|
| againusa Forum Newbie
Topics: 2 Posts: 11
| | 03/21/06 - 12:19 PM  
 
   
 
|   #2 |
I learned the following calculation method from this forum: 1)calculate the positive likely hood ratio which is sensitivity 1-specificity) in this case is 99:1 2)in the case of 25/100 prevalence, transform into disease:normal=25:75 3)multiply the ratio in 1) and 2), you will get 33:1, it is 33/34 about 97%
|
| hrishi Forum Newbie
Topics: 0 Posts: 17
| | 03/27/06 - 02:28 PM  
 
   
 
|   #3 |
cant really make out .please explain.
|
| pearljam59 Forum Guru

Topics: 120 Posts: 273
| | 04/07/06 - 09:24 AM  
 
   
 
|   #4 |
How about I give you the formula for PPV: True Positives / # of people who tested POSITVE for a disease Does that help?!

|
| robin082006 Forum Hero

Topics: 471 Posts: 5,125
| | 04/07/06 - 10:26 AM  
 
   
 
|   #5 |
For a question like this one, it's very simple if you fill a table. For example, prevalence = 25% .................Disease .............Non-disease ..............Total Positive ......24.75(0.99*25) ......0.75 ......................25.5 Negative .......0.25 ....................74.25(0.99*75) ......74.5 Total ............25........................ 75 ........................100 PPV=true positive/total positive = 24.75/25.5=97% Firstly, you write down number of diseases and non-diseases based on prevalence, E.g. prevalence 20%, disease is 20 and non-disease is 80, for easy calculation, total should be 100. Then based on Sensitivity and Specificity, write down number of true positive and true negative. False Negative and false positive then are calculated by substracting from total disease-true positive and total non-disease-true negative, respectively. then calculate PPV = true positive/total positive NPV=true negative/total negative
___________________ The Key to Succeed is Patience.
|
|
| |
| | | | | |