False positives on HSV tests?
January 25, 2006 8:58 AM Subscribe
False positives on HSV tests?
My girlfriend tested positive a few years ago to genital herpes, though she's never had any symptoms. After switching to a new ObGyn clinic, her new doctor mentioned in passing that she probably got a false positive if she's never had any symptoms, as the HSV test is notorious for spitting out positives if you've ever had chicken pox as a kid. Is this the case? What's the history of this test in terms of accuracy, and are there new tests that are actually accurate?
My girlfriend tested positive a few years ago to genital herpes, though she's never had any symptoms. After switching to a new ObGyn clinic, her new doctor mentioned in passing that she probably got a false positive if she's never had any symptoms, as the HSV test is notorious for spitting out positives if you've ever had chicken pox as a kid. Is this the case? What's the history of this test in terms of accuracy, and are there new tests that are actually accurate?
The answer to any question like this is going to be complicated.
If the accuracy of a given test is 95%, that (normally) means that:
(1) IF you actually have HSV, THEN it will come out positive 95% of the time (or for 95% of people). Call this prob(+|H), or positive given H.
(2) IF you actually do NOT have HSV, THEN it will give a false positive only 5% of the time. Call this prob(+|C), or positive given clear.
But that's backwards. What most people care about is:
IF the test comes out positive, THEN what is the probability that I actually have HSV? That is, people care about prob(H|+), or the probability of H given a positive result.
The answer to this depends on the prevalence of HSV in the population -- if everyone gets a test for a rare condition, than 5% of a whole damn bunch can be more than 95% of a relative few, so you end up with more false positives than true positives. It also depends on why your girlfriend got a test a few years ago, since the relevant "population" for her might be higher- or lower-risk than the population as a whole. Broadly, prob(H|+) will be lower if the test was a routine screen done without prompting by any particular event.
I dunno if this is helpful to you. My point is that you have to interpret accuracy of tests, especially routine tests, rather cautiously.
posted by ROU_Xenophobe at 11:25 AM on January 25, 2006
If the accuracy of a given test is 95%, that (normally) means that:
(1) IF you actually have HSV, THEN it will come out positive 95% of the time (or for 95% of people). Call this prob(+|H), or positive given H.
(2) IF you actually do NOT have HSV, THEN it will give a false positive only 5% of the time. Call this prob(+|C), or positive given clear.
But that's backwards. What most people care about is:
IF the test comes out positive, THEN what is the probability that I actually have HSV? That is, people care about prob(H|+), or the probability of H given a positive result.
The answer to this depends on the prevalence of HSV in the population -- if everyone gets a test for a rare condition, than 5% of a whole damn bunch can be more than 95% of a relative few, so you end up with more false positives than true positives. It also depends on why your girlfriend got a test a few years ago, since the relevant "population" for her might be higher- or lower-risk than the population as a whole. Broadly, prob(H|+) will be lower if the test was a routine screen done without prompting by any particular event.
I dunno if this is helpful to you. My point is that you have to interpret accuracy of tests, especially routine tests, rather cautiously.
posted by ROU_Xenophobe at 11:25 AM on January 25, 2006
I'm surprised that this question hasn't had more answers by now. If you don't get many more, you might consider re-posting, but with the word "herpes" visible somewhere from the front page (i.e. not only in the "more inside" part).
posted by unknowncommand at 1:46 PM on January 25, 2006
posted by unknowncommand at 1:46 PM on January 25, 2006
Anecdotal: had childhood chicken pox. Consistently negative on repeated blood tests for both HSV1 and HSV2.
posted by orthogonality at 2:09 PM on January 25, 2006
posted by orthogonality at 2:09 PM on January 25, 2006
Here's an answer from Go Ask Alice to just about the same question. They say it's about 5% false positive, and old tests can't differentiate between HSV1 and 2 (genital herpes vs. coldsores). New tests can differentiate between the two.
Chickenpox is actually a member of the herpes virus family, so yes, that could have triggered a false positive result, particularly if the test was a Tzanck test.
Still, if you're worried, it doesn't hurt to get checked out again!
posted by orangskye at 4:40 PM on January 25, 2006
Chickenpox is actually a member of the herpes virus family, so yes, that could have triggered a false positive result, particularly if the test was a Tzanck test.
Still, if you're worried, it doesn't hurt to get checked out again!
posted by orangskye at 4:40 PM on January 25, 2006
This thread is closed to new comments.
I wonder if the doctor specified HSV 1 or 2? If it's 1, it could just be the regular old coldsore on the lip variety.
However, realistically: A lot of people do have HSV2. Some of them are silent carriers.
This doesn't answer all your questions, but I hope it's helpful.
posted by veronica sawyer at 9:35 AM on January 25, 2006