How to Calculate Positive (PPV) and Negative Predictive Values (NPV) - YouTube

In this video, I'm going to explain what positive predictive values and negative predictive values are and how to calculate them.

Hi, and Welcome back to physiotutors.

Before you watch this video on how to calculate positive predictive values and negative predictive values,

you should know what sensitivity and specificity are and how they are calculated.

If you want to learn more about that make sure to click in the top right corner to watch our videos on those topics.

Now remember that in the clinical setting you do not know if your patient has the disease or not.

So the positive predictive value of a test tells you how likely it is that the patient has a disease

after he tested positive

and the negative predictive value tells you how likely it is that your patient does not have the disease if he tested negative.

As the predicted values are always dependent on the prevalence of the disease

it is wise to first calculate the prevalence

and we do that if we take all the people who have the disease

so in this case we have 220 plus 30 so 250 people who have the disease

and we have to divide them through all the people in our study

and those are 250 plus the 650 people who do not have the disease so 250 through 1,000

and we're going to end up at a prevalence of 25%

So the positive predictive value is the proportion of patients who have the disease amongst all the patients who test positive.

So we have to divide all the true positives

through all the people that test positive

in this case it's 220 people who are truly positive

divided through all the people that test positive so 220

plus the false positives as well

plus 75

So in this case we end up with a positive predicted value of around 0.75 which is 75%

Now on the other hand the negative predictive value is the proportion of patients who do not have the disease

amongst all the patients who test negative

So in this case we have to take all the true negatives, 675

and divide them through all the patients that test negative,

so these two values through negative

675

Plus all false negatives which are 30 in this case

Then we're going to end up with an NPV of 675

through 705 which is about 0.96 so 96% negative predictive value.

it's good to remember that you calculate the positive predictive value from the first row

and the negative predictive value is always calculated

by the use of values from the second row

So if we go back to our prevalence of 25% that we calculated earlier

We can now say that the chances that a patient actually has the disease

increased from 25 percent to 75%

with a positive test outcome

on the other hand if we had 25% of sick people

we also had 75% of healthy people

So the chances that a patient does not have the disease increased

from 75% to actually 96% with a negative test outcome

The good thing is

that we can obtain the prevalence of a certain disease or injury that we see in our respective settings

from literature

The problem is that we can only use the predictive values of a test

if the prevalence is identical to the one reported in a study

So for example the prevalence of ACL tears will be much lower in a generalised PT practice

in comparison with a sports clinic that is specialised in knee injuries

Therefore the predictive values are not really useful in most cases

and the best and most effective tool that we can use are likelihood ratios

If you want to know what they are and how you can calculate that

make sure to click on a video right next to me

All right,

this was our video on predicted values

I hope this video made a lot of things clearer

if it did give it a thumbs up or comment in the section down below

and make sure to subscribe to our youtube channel

and check us out on Facebook Instagram or physiotutors.com

We'll see you in the next video.

Bye