How to calculate sensitivity, specificity, positive predictive value and negative predictive value.

Medical students need to be dexter in statistics. It is really needed though we often find it boring and hard to get through. These basic things apply in everyday medicine. So ready ?

stats dexter

 

Lets assume we have 1000 patients for sample.

Now lets say 100 patients actually have Diabetes not under treatment. So, 900 people have no Diabetes .

We want to see if Random Blood Sugar(RBS) can be used as screening test to detect Diabetes. All 1000 people under go RBS test.

Among 100 people with diabetes, 95 had positive screen ie high RBS , 5 had normal RBS.

 

Therefore-

People with diabetes AND tested positive are the true positives (TP)= 95

People with diabetes AND tested negative are the false negatives (FN)=5

People without diabetes AND tested positive are the false positives (FP)=90

People without diabetes AND tested negative are the true negatives (TN)=810

 

Now applying the Formula,

Sensitivity= True Positive/(True Positive +False Negative)  X 100%

Here, Sen = 95/95+5 X 100= 95%

This means, RBS is able to detect diabetes in 95% cases who have diabetes.

 

Specificity= True Negative/ (True Negative+ False Positive) X 100%

Here, Spec = 810/810+90 X 100=90%

This means the test has ability to rule out diabetes in 90% patients without diabetes.

 

Positive predictive value=True Positive/( True positive + False positive) X 100

Here, PPV= 95/(95+90) X 100= 51.40%

This means the test has ability to tell us that if the test is postive by RBS, 51.40% will have diabetes.

 

Negative predictive value=True Negative/( True Negative + False Negative) X 100

Here, PPV= 810/(810+5) X 100= 99.40%

This means the test has ability to tell us that if the test is negative by RBS, 99.4.40% will not have  diabetes.

 For a Clearer Picture, this can be applied in all diseases

Sen_spec

Hope, it was helpful!

About the Author

Administrator
Doctor Medchrome writes for and looks overall management of Medical Online Magazine.

Be the first to comment on "How to calculate sensitivity, specificity, positive predictive value and negative predictive value."

Leave a comment

Your email address will not be published.

*