PCR testing

[vs7829]

Hi Doctor,

I know you don

[H. Hunter Handsfield, M.D.]

There are two basic problems with most of the questions about symptoms of HIV/AIDS.  First, most symptoms of primary HIV infection (acute retroviral syndrome, ARS) are nonspecific, i.e. occur in many conditions other than ARS.  Among someone whose exposure is low risk, such as vaginal sex or a single episode of oral sex, the vast majority of people with such things as sore throat, swollen glands, or fever have garden-variety infections, not ARS.  Second, many people confuse symptoms of AIDS with those or ARS.  For example, a recent questioner asked about perianal abscess, which indeed is more common in people with AIDS, but is not a sign of ARS.

I don't have time to list all the symptoms of ARS or AIDS, but they are available on lots of websites.  But as I said above, most occur with many other conditions, which generally are more common than HIV.

Hope that helps--   HHH, MD

[HHH,MD]

Paragraph 2, sentence 2:  Should say that isn't the reason that "I almost always recommend AGAINST the test..."

HHH, MD

[vs7829]

Doctor,

Thanks for the quick response. I have been following your board and notice a lot of the times you do not agree that most symptoms are indicative of HIV or STD's. For our educational purpose and assuming someone does show symptons of HIV, what would they be and when would they usually occur (in a typical case). Thanks.

[H. Hunter Handsfield, M.D.]

It is true that PCR for HIV DNA is positive during the 'window period' after infection, before antibody develops; and therefore can detect infection before the standard antibody tests (EIA, ELISA, Western blot).  The timing of test positivity you summarize in paragraph (1) sounds about right, although I don't vouch for the details.

Regarding paragraph (2), I am not aware that the test has "a high rate of false positive" results.  That is not the reason that I almost always recommend the test for most people who ask about it on this forum.  If a person's risk of having HIV is only 1 in 1000 to 1 in a million to start with (which is the usual range in most of the threads on this issue), is it really worth $350 to drop that risk to one tenth the starting value?  If your risk is 1 in 10,000, will you really sleep better at 1 in 100,000?  Or is it smarter to save the $350, wait a few days, then have the antibody test?  And I think it is particularly dumb for someone who has a negative antibody test at 3-4 weeks, which is  90-98% reassuring all by itself, to seek PCR testing in addition.  If it is true that false positives are a frequent problem, that simply adds to my reasons not to be tested.

HIV PCR testing is an important diagnostic test in some settings.  But in general, those settings require a pretty high prior probability, i.e. a signficant risk that HIV infection actually has occurred; or at least a particularly high pay-off in prevention despite the rarity of the outcome (e.g., routine PCR of HIV negative blood donors, because every person who receives such a person's blood products would otherwise get HIV).

I suspect many (most?) people seeking (or being advised to have) PCR testing don't completely understand this stuff.  But if someone is aware of these facts and still chooses to have PCR testing for his/her peace of mind, and can afford it, fine with me.

Regards--  HHH, MD