I suck at Numbers...

[Future GI Guy]

Have a numbers question for all you epidemioligists...

If you have a test that's 90% sensitive (in this case, the specificity doesn't matter), and you take the test 4 times, and all times the test is negative, what's the new sensitivity? How do you calculate this?

I don't know the PPV, NPV, or any other numbers which may be necessary.

---

Members don't see this ad.

 

[ckent]

Well, can't say that I'm an epidemiologist, but based on the numbers you gave, and what I remember from my epi course, it should just be 1-.1*.1*.1*.1 or 0.9999 . This is assuming that the sensitivity of the tests remains at 90% each time you re-take it, which is almost never the case in real life. If an HIV patient tests negative on a HIV screening test (let's say that the HIV test that is looking for an antibody), then if you looked at all the people who had HIV but tested negative and made them re-take the test, the sensitivity would be markedly lower taking the same test the second time because those people with HIV may just not have the antibody that the test is looking for (so they would be false negatives in the first test, and remain false negative in subsequent tests). You would be better off sending them off to take a test that was looking for something different.

 

[Future GI Guy]

Thanks for the reply. You're absolutely right with your calculation...as soon as you posted how to do it, I realized how stupid I was.

By the way, the test in question was a pregnancy test (taken by my wife).

Unfortunately, it was negative. Four times.