Not Logged In     Login  
 Release 22.0, Sept. 2008
 
The Medical Algorithms Project
Chapter : ch39. Decisional Analysis Section : Performance Measures
  Percent of Patients Misclassified by a Test in a Population with a Given Disease Prevalence

  Online Excel Reference
Copyright (c) 2008-2010, Institute for Algorithmic Medicine, Houston, TX, USA. All rights reserved.

Overview :

If a test with a certain sensitivity and specificity is used in a population with a certain disease prevalence, then a certain percentage of patients in the population will be misclassified.

 

NOTE: This is a variation on the content in 39.01.13 (above).

 

Population

Expression

percent true positives

(percent prevalent) * (sensitivity)

percent false negatives

(percent prevalent) * (1 - (sensitivity))

percent true negatives

(100 - (percent prevalent)) * (specificity)

percent false positives

(100 - (percent prevalent)) * (1 - (specificty))

 

percent of patients in a population misclassified =

= (percent false negative) + (percent false positive)

 

ratio of true positives to false positives =

= (percent prevalent) * (sensitivity) / ((100 - (percent prevalent)) * (1 - (specificity)))

 

ratio of true negatives to false negatives =

= (percent prevalent) * (1 - (sensitivity)) / ((100 - (percent prevalent)) * (specificity))

 

  References:

 

 
   Online Excel | Reference TOP
     Pubmed Search For
Copy and paste the article title, or authors names into the search box