snout & spin - dep

[ArchivalUser]

Can any one plz help me in understanding this concept

Sensitivity detects true positive cases means the ones who really have disease so how does sensitivity rules out disease?

Specificity detects true negative cases means ones who really do not have disease so how does specificity rules in disease?

[ArchivalUser]

With 100% sensitivity, everyone who has the diease will test positive. There will be very low false negatives (meaning it will be very unlikely that a person who has the disease will test negative). Since sensitivity tries to include in everyone as much as it can, when you do a sensitivity test, if it's done right, 100% of people who have the disease will test +, but you will also have a lot of FP (even people who don't have the disease, may test positive). So when it says, sensitivity will RULE OUT, it means.. it will rule out anyone for sure who DOESN'T HAVE THE DISEASE (low FN)..or in other words.. if a pt takes a test for HIV using ELISA and it comes back negative.. u can be pretty sure that the patient doesn't have HIV (meaning low FN with sensitive tests).

With 100% specificity, people who don't have the disease will test negative. So, there will be very low false positives (meaning this is a confirmatory test of sorts.. and the chances of a person without a disease, testing positive is very unlikely). So, when you are RULING IN using specificity, you are making SURE the person has the disease!! Meaning.. again with HIV, if the patient tested positive for it with ELISA.. he may/may not have the disease. So, when you use Western Blot, and if the pastient tests positive.. it can be concluded that the patient in fact is positive for HIV.. that means you are RULING IN.

It's really hard to explain this using words.. I hope I haven't confused u more!

[ArchivalUser]

wow..this scarlett girl...is really good at explaining things...must be brilliant!!!!!!

[ArchivalUser]

HAHA.. no way jose!! that's ALL YOU!!: MR. "WHAT IS CRIMEAN-CONGO HERMORRHAGIC FEVER?..."

[ArchivalUser]

Thanx a lot scarlett0208

u didnt confuse me but clears my concept very well

[ArchivalUser]

I must disagree.... though you all are probably working physicians by now so it probably doesn't matter, but for people who search and need accuracy....

Scarlet - you wrote:
it will rule out anyone for sure who DOESN'T HAVE THE DISEASE (low FN)

FN are those that are in reality diseased, but the test didn't catch it (Negative test, Falsely so). So they do have the disease.

The way I like to think about it:
TP: test is Positive, and that's the Truth
TN: test is Negative, and that's the Truth
FP: test is Positive, falsely so (person is disease-free; test is wrong)
FN: test is Negative, falsely so (person has disease; test is wrong)

Sensitivity = people who have the disease and test caught it, divided by all the people that actually have the disease (regardless of what the test says) = TP/(TP+FN) & then multiply by 100 to get percent

If the test is highly sensitive (TP>>>>>>>FN) the numerator and denominator are almost the same (approaches 100%), and the test result is negative you can be nearly certain that they don’t have disease, because your option is either a TN or FN. We know the number of FN is small (I just explained that - that's the definition of high sensitivity), so the only other option is a TN. That means by performing this test, we got all the positive results, and if you tested negative - it's probably for real - so you're ruled out from having it. Don't be mistaken that we also caught FP here - this is not true because you can't see that in this equation. A test can have high sensitivity and high specificity, or a mix of one and the other.

Hope I cleared things up.

Good luck!

[ArchivalUser]

Disease a
test result + _ sensitivity -------
a + c
true false
+ positive positive specificity d
a b ----
false true d+b
- negative negative
c d

sNout- with a high seNsitivity(low false negative), a Negative test result identifies those withOUT disease. Negative test results are either c or d. If c is low or zero, then a negative test assures a true negative-rules out

sPin- with a high sPecificity(low false positive), a Positive test result identifies those with disease. Positive test results are either b or a. If b is low or zero, then a positive result assures a true positive- rules in

Remember that sNout tells you what a Negative test means when you have a very sensitive test and sPin tells you what a Positive test means when you have a very specific test.

Don't confuse this with what positive test tells you when you have a sensitive test-disease is most likely present or with what a negative test tells you when you have a specific test- disease is most likely not present.

[ArchivalUser]

My first response didn't come across the way I typed it. To understand what I wrote above:

a=true positive
b=false positive
c=false negative
d=true negative

sensitivity= a/a+c
specificity= d/d+b

[ArchivalUser]

So, highly specific test is positive, that rules in disease (SpPIN). Highly sensitive test is negative, that rules out disease (SnNOUT).

Is the opposite also true? A negative result from a specific test rules out disease and a positive result from a sensitive test rules in disease?