|
Home > Archive > HIV Aids > December 2006 > Pregnancy and False Positive HIV Tests
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
Pregnancy and False Positive HIV Tests
|
|
|
| Here is why pregnant women test false positive on HIV tests.
And why tests that rely on blood from pregnant women are
singularly unreliable. Like all the surveys that 'show' that
30%, 50%, etc. of women in Botswana, Swaziland, etc.
are 'HIV positive'.
Apparently the placenta did it.
" The nature and/or
potential function of these particles/proteins has not yet been fully defined." "
Remember that 'HIV tests' do not really test for HIV, but for
proteins that are presumed to be unique to HIV.
From the NIH:
http://www.ncbi.nlm.nih.gov/entrez/...ve&dopt=abstrac
tplus&list_uids=7473433
"Expression of intact endogenous retroviruses by normal placental
villous trophoblast and immuno-crossreactivity of villous trophoblast
with anti-retroviral antisera have been documented. The nature and/or
potential function of these particles/proteins has not yet been fully defined."
Expression of endogenous HIV-1 crossreactive antigens within
normal human extravillous trophoblast cells.
| |
| John_fr 2006-11-19, 4:22 pm |
|
Alex a =E9crit :
> Here is why pregnant women test false positive on HIV tests.
>
> And why tests that rely on blood from pregnant women are
> singularly unreliable. Like all the surveys that 'show' that
> 30%, 50%, etc. of women in Botswana, Swaziland, etc.
> are 'HIV positive'.
>
> Apparently the placenta did it.
>
> " The nature and/or
> potential function of these particles/proteins has not yet been fully def=
ined." "
>
> Remember that 'HIV tests' do not really test for HIV, but for
> proteins that are presumed to be unique to HIV.
>
> From the NIH:
>
> http://www.ncbi.nlm.nih.gov/entrez/...actplus&db=3Dp=
ubmed&cmd=3DRetrieve&dopt=3Dabstrac
> tplus&list_uids=3D7473433
>
> "Expression of intact endogenous retroviruses by normal placental
> villous trophoblast and immuno-crossreactivity of villous trophoblast
> with anti-retroviral antisera have been documented. The nature and/or
> potential function of these particles/proteins has not yet been fully def=
ined."
>
> Expression of endogenous HIV-1 crossreactive antigens within
> normal human extravillous trophoblast cells.
Alex, it is of use to nothing, they have their agenda. You can
demonstrate them by a+b that they are delirious, they have the power
and take advantage of it.
Thank you in any cases for the link.
| |
| Bennett 2006-11-19, 4:22 pm |
|
Alex wrote:
> Here is why pregnant women test false positive on HIV tests.
>
> And why tests that rely on blood from pregnant women are
> singularly unreliable. Like all the surveys that 'show' that
> 30%, 50%, etc. of women in Botswana, Swaziland, etc.
> are 'HIV positive'.
>
> Apparently the placenta did it.
Alex - I told you about this 6 years ago!!!
http://groups.google.com/group/misc...35f320fc95ac5a6
Cheers
Bennett
| |
| Chris Noble 2006-11-19, 4:23 pm |
|
Bennett wrote:
> Alex wrote:
>
> Alex - I told you about this 6 years ago!!!
>
> http://groups.google.com/group/misc...35f320fc95ac5a6
>
> Cheers
>
> Bennett
And David Canzi specifically refuted Alex in this thread.
http://groups.google.com/group/misc...2015eaa29c86c2/
Testing pregant women with a single ELISA does not give false positives
at anywhere near 30-50%.
You would think that even if Alex has memory problems that he would at
least do a google search to check the last time somebody refuted his
argument.
Chris Noble
| |
| drpsduke@yahoo.com 2006-11-19, 4:23 pm |
|
Chris Noble wrote:
....
> You would think that even if Alex has memory problems that he would at
> least do a google search to check the last time somebody refuted his
> argument.
>
> Chris Noble
Don't you know that a lie, repeated often enough, becomes the truth?
Maybe that is not the way it works in science and other versions of
reality, but that is the way it works in AIDS denial and many forms of
politics.
| |
| Chris Noble 2006-11-19, 4:23 pm |
|
drpsduke@yahoo.com wrote:
> Chris Noble wrote:
> ...
>
> Don't you know that a lie, repeated often enough, becomes the truth?
> Maybe that is not the way it works in science and other versions of
> reality, but that is the way it works in AIDS denial and many forms of
> politics.
Or maybe Google's experience in China has allowed them to develop a
variety of Denialist options that allows evolution "rethinkers", HIV
"rethinkers", germ theory of disease "rethinkers" to surf the internet
without being exposed to any information that might conflict with their
worldview.
Chris Noble
| |
| Chris Noble 2006-11-19, 4:23 pm |
|
drpsduke@yahoo.com wrote:
> Chris Noble wrote:
> ...
>
> Don't you know that a lie, repeated often enough, becomes the truth?
> Maybe that is not the way it works in science and other versions of
> reality, but that is the way it works in AIDS denial and many forms of
> politics.
Or maybe Google's experience in China has allowed them to develop a
variety of Denialist options that allows evolution "rethinkers", HIV
"rethinkers", germ theory of disease "rethinkers" to surf the internet
without being exposed to any information that might conflict with their
worldview.
Chris Noble
| |
|
| "Chris Noble" <ChrisJNoble@hotmail.com> schreef in bericht
news:1163655488.972664.45080@m7g2000cwm.googlegroups.com...
> And David Canzi specifically refuted Alex in this thread.
>
> http://groups.google.com/group/misc...2015eaa29c86c2/
>
> Testing pregant women with a single ELISA does not give false positives
> at anywhere near 30-50%.
>
> You would think that even if Alex has memory problems that he would at
> least do a google search to check the last time somebody refuted his
> argument.
>
> Chris Noble
Actually, the explanation is very simple. You are confusing positives
and false positives.
Among pregnant women in the west, Positive ELISA tests are extremely
rare. However, if you look at how many test positive, and then STAY
positive after confirmation tests are carried out, False Positives are THE NORM.
Then, you must have a clear explanation of why these 25%, 33% infection
rates disappear as soon as testing involves non-pregnant girls in Africa.
Or to be specific, why infection rates go from 32.5% to 6% in Swaziland.
If there are any real infections at all.
Also, don't count out the effect of environmental pollution, but that is
for another thread.
Alex
UNICEF DHS IN SWAZILAND
MBABANE, 27 August (PLUSNEWS) - A dramatically lower number of Swazi
teenage girls are being infected by HIV than was previously estimated, suggesting
a turning point in the battle against HIV/AIDS in a country with the world's highest
HIV infection rates.
The findings in the report, 'A Baseline Study on HIV Risk Factors', commissioned
by the UN Childrens' Fund (UNICEF) are derived from interviews and blood tests
of over 1,000 Swazis in two rural areas and revealed that only six percent of girls
aged from 15 to 19 were found to be HIV-positive, with most of the HIV infections
occurring among older girls.
"This is the first time we have had data from a scientifically accurate survey of
randomly selected households. It confirms some trends we had suspected, but
which were belied by previous HIV estimates," said Dr Alan Brody, country
representative for UNICEF.
"This is different from anything that has been seen before. The conventional
wisdom is that many more girls were infected," he told PlusNews.
The study was prompted by the results of the government's 2002 sero-surveillance
study, which estimated that 32.5 percent of teenage girls between the ages of 15
and 19 were HIV-positive.
http://www.plusnews.org/AIDSReport....ectCountry=SWAZ
ILAND
[Wow, now we're getting to the meat of things. Of 26 positive Rapid
ELISAs, only 26 REMAINED positive after Western Blot testing. So
much for false positives being 'extremely rare'. You see the key is, that
you compare the number of false positives to the total number of positives,
not the total number of people who were tested, most of whom in the West
test negative. This would put the number of False Positives among positive
testing pregnant women at a whopping 62.3% and the percentage of True
Positives 37.7%. It doesn't need to be repeated that confirmation tests in
surveys in Swaziland are NEVER carried out - Alex]
http://cat.inist.fr/?aModele=afficheN&cpsidt=16551024
RESULTS: A total of 69 patients had a positive rapid HIV-ELISA out of
9,781 deliveries. Of those, 26 were confirmed as HIV infected by Western
blot (overall HIV prevalence: 0.27%, ELISA-positive predictive value: 37.7%).
http://womenshealth.about.com/cs/ai...shivwomen_2.htm
A reactive test may give a false positive reading to anyone with kidney
or renal failure, to a woman that has had multiple pregnancies, anyone
receiving the influenza vaccine, or to anyone that has received gamma
globulin.
http://www.dmt123.com/qa/pregnancy_...851-dmt123.html
Can pregnancy cause a false-positive HIV test result?
Question
I am pregnant, and I just had an HIV test done. To my horror, it came
back positive. As I was commiserating with my girlfriend, she said that
she heard that being pregnant could cause a false positive result. Is this true?
Answer
Yes, pregnancy can occasionally cause a false positive HIV antibody test.
Thank you for your question.
[Comment - notice that this says nothing about ELISA only tests, which
is what we are talking about in surveys, including in Swaziland; it most likely
refers to results from multiple tests, including Western Blot, or some
calculation that takes into account all tests in a huge survey, most of
which are negative in the West - Alex]
http://www.sogc.org/health/pregnancy-hiv_e.asp
Although it is extremely rare (about 1 out of every 20,000 tests,) a test
may give a false-positive result.. A false-positive test says you are infected
when you are not. If there are any doubts, you can be retested.
[So much for false positives being 'extremely rare', and this is in
a non-tropical, well to do environment - Alex]
http://findarticles.com/p/articles/...20/ai_107894045
One out of 14 previously pregnant women given the OraQuick
Rapid HIV-1 Antibody Test, which the CDC is promoting,
showed false positive results. (6, 7)
http://www.thebody.com/Forums/AIDS/...ed/Q141298.html
HIV False + & Pregnancy
Sep 2, 2002
I was recently tested for HIV in January 2002 with the result being negative.
I hadn't slept with anyone in over 9 months. I then became intimate with
someone else 6 months later and almost immediately became pregnant
which led to a routine test for the HIV virus. After just two months of being
with this individual I received a HIV positive test. The Westeren Blot has
not been administered yet. I am to have it done in a few more months.
However, my question is, would a test come up positive in just 2 months
and what causes a pregnant womans test to come back as a false positive.
The antigens or antibodies that can possible become active in pregnant
able to reproduce at a faster rate than in someone recently infected.
Please, please respond as I am depressed beyond belief.
Response from Dr. Aberg
Prgenancy may cause a false positive screening test called an Elisa. The
western blot should have been done on the same specimen and the results
back relatively soon unless there was not enough blood sample to be tested.
Has your partner been tested? If he is the only one you have had sex with
since your last negative test and he is negative, then that may help ease your
mind. If your doctor and you are concerned that you just acquired HIV,
then you should have a HIV viral load measured and antibody testing in
follow up. You can have false positive HIV viral loads but the amount of
virus is usually much lower than one would expect in primary infection.
Best to do repeat Elisa and if positive, western blot testing.
http://www.ncbi.nlm.nih.gov/entrez/...s&list_uids=962
9106&itool=pubmed_AbstractPlus
New, third-generation HIV tests have reduced false-positive rates and
thus diminished the harm of screening.
| |
| Chris Noble 2006-12-06, 9:26 pm |
|
Alex wrote:
> "Chris Noble" <ChrisJNoble@hotmail.com> schreef in bericht
> news:1163655488.972664.45080@m7g2000cwm.googlegroups.com...
>
>
> Actually, the explanation is very simple. You are confusing positives
> and false positives.
> Among pregnant women in the west, Positive ELISA tests are extremely
> rare. However, if you look at how many test positive, and then STAY
> positive after confirmation tests are carried out, False Positives are THE NORM.
Now you are confusing specificity with positive predictive value.
> Then, you must have a clear explanation of why these 25%, 33% infection
> rates disappear as soon as testing involves non-pregnant girls in Africa.
There are many reasons why seroprevalence statistics from pregant women
may overestimate seroprevalence in non-pregnant girls.
What I find interesting is that seroprevalence based on antenatal
scrrening has increased over time. The first national antenatal survey
in South Africa found a prevalence of only 0.8%. This has increased
dramatically over time. Why?
> Or to be specific, why infection rates go from 32.5% to 6% in Swaziland.
> If there are any real infections at all.
>
> Also, don't count out the effect of environmental pollution, but that is
> for another thread.
>
Environmental pollution causes false positives on HIV tests?
Chris Noble
| |
| Bennett 2006-12-06, 9:26 pm |
|
Alex wrote:
> Actually, the explanation is very simple. You are confusing positives
> and false positives.
> Among pregnant women in the west, Positive ELISA tests are extremely
> rare. However, if you look at how many test positive, and then STAY
> positive after confirmation tests are carried out, False Positives are THE NORM.
Err, but by definition if the confirmation tests are negative, then the
test is negative.
A false positive would be a positive ELISA AND confirmation test, when
in fact there was no HIV present.
>
> Then, you must have a clear explanation of why these 25%, 33% infection
> rates disappear as soon as testing involves non-pregnant girls in Africa.
>
> Or to be specific, why infection rates go from 32.5% to 6% in Swaziland.
> If there are any real infections at all.
>
The article you quote from gives you that answer. It's simply that the
32.5% of girls who were tested in the early years GREW OLDER.
"Nevertheless, the new baseline survey found a 40 percent
HIV-prevalence rate among older girls, from 19 years upwards, which
matched the sero-surveillance study results."
You don't help your case much Alex when you either fail to read the
articles properly or (worse) are misrepresenting them on purpose. The
seropositive rates dropped because kids stopped having unprotected sex
(which would only work, of course, if they were preventing transmission
of a sexually transmitted infection!)
http://www.plusnews.org/AIDSReport....untry=SWAZILAND
>
> http://womenshealth.about.com/cs/ai...shivwomen_2.htm
> A reactive test may give a false positive reading to anyone with kidney
> or renal failure, to a woman that has had multiple pregnancies, anyone
> receiving the influenza vaccine, or to anyone that has received gamma
> globulin.
Sadly the bit about the flu vaccine is probably wrong, and based upon
one batch of one years vaccine from one company reacting with
polystyrene beads in one type of HIV ELISA (no false positives occured
as the WB's were all negative). One of the perils of believing
everything you read from the web.
The rest we already know - hell we had a case of a positive
ELISA/negative WB from a pregnant woman in our ID clinic just a couple
of weeks ago. No-one batted an eyelid - she was HIV-negative, simple
as that. You're preaching to the choir.
Cheers
Bennett
| |
|
| "Chris Noble" <ChrisJNoble@hotmail.com> schreef in bericht
news:1164074737.749203.232550@m7g2000cwm.googlegroups.com...
>
> Alex wrote:
>
> Now you are confusing specificity with positive predictive value.
No I'm not. Please check again, I said 'if you look at how many test positive'.
Most of those who test positive on ELISA are false positive (they don't stay
positive after confirmation testing with Western blot.
In Africa, you just have a lot more people who test positive. I never understood
why it is presumed why there are only huge numbers of false positives ELISAs
when very few people out of the total survey test positive on ELISA, than when
many people in a survey test positive on ELISA.
What I would like to see, is a survey that uses both ELISA for screening
purposes, and Western Blot to confirm positive ELISAs.
What is happening right now, is that DHS surveys only use Western Blot
on ELISAs that seem neither positive nor negative.
So if you know of a DHS type survey in Africa, that uses Western Blot
as a confirmation test, that would be great.
>
> There are many reasons why seroprevalence statistics from pregant women
> may overestimate seroprevalence in non-pregnant girls.
Not according to UNAIDS, who for years have used
ELISA tests of pregnant women as a direct proxy for
national infection rates.
Which is why you ended up with the 33.4% of (adult) Swazis are HIV positive yarns.
(Source: Adult HIV prevalence (%), 2003 33.4)
http://www.hivinsite.org/global?page=cr09-wz-00
So, the HIV Insite, which still shills the UNAIDS (political) line,
still claims that 33.4% of adults in Swaziland (15-49 years of age)
are HIV positive.
This of course coincides approximately with the belief that
32.5% of pregant women in Swaziland were HIV positive.
(Source:
http://www.plusnews.org/AIDSReport.ASP?ReportID=3819&
SelectRegion=Southern_Africa&SelectCountry=SWAZILAND
)
What you see here, and this is the criminal part, is the direct
translation of results from pregant women, to the entire
(adult) population.
And that was always unscientific, and anti-scientific, even
if you believe that the tests are accurate, that HIV causes
AIDS, etc. You can believe all of that, but still, just as someone
with a vague background in surveys and statistics, know
that you cannot simply extrapolate the results from
pregnant women at antenatal clinics, to everyone in the population.
And that is where I think UNAIDS is criminally negligent.
> What I find interesting is that seroprevalence based on antenatal
> scrrening has increased over time. The first national antenatal survey
> in South Africa found a prevalence of only 0.8%. This has increased
> dramatically over time. Why?
What year was that? If it was during apartheid, did they only screen
white pregnant women?
Please include the source.
>
> Environmental pollution causes false positives on HIV tests?
Depends on what makes the already extremely sensitive ELISA
HIV test make positive. Whether that is exposure to DDT,
copper, mercury, etc. I don't know.
What I do know, is that environmental exposure to banned
pesticides, pollution, industrial spills etc. is much greater in
Africa than in Oxford, just to mention one place.
Alex
| |
|
| "Alex" <avdeelen.REMOFETHIS1@wanadoo.nl> schreef in bericht
news:4561d3c3$0$29576$dbd43001@news.wanadoo.nl...
And just another thing:
> http://womenshealth.about.com/cs/ai...shivwomen_2.htm
> A reactive test may give a false positive reading to anyone with kidney
> or renal failure, to a woman that has had multiple pregnancies, anyone
> receiving the influenza vaccine, or to anyone that has received gamma
> globulin.
So what is it exactly, that is making an ELISA false positve
under these circumstances, and how does it do that?
Alex
| |
|
|
"Bennett" <njb35@cantab.net> schreef in bericht
news:1164079319.752336.172690@f16g2000cwb.googlegroups.com...
>
> Alex wrote:
>
>
> Err, but by definition if the confirmation tests are negative, then the
> test is negative.
You mean when the confirmation test is negative, the result is negative.
> A false positive would be a positive ELISA AND confirmation test, when
> in fact there was no HIV present.
In the context of this discussion, a false positive is a positive screening test,
which is then refuted by a negative confirmation test.
>
> The article you quote from gives you that answer. It's simply that the
> 32.5% of girls who were tested in the early years GREW OLDER.
And this cured them of HIV?
What happened, is that a different survey type was used, nothing more.
The 32.5% is based on the Antenatal Clinic Survey, which tests about 200
or so pregnant women at antenatal clinics, which are usually in urban settings.
On the other hand, the survey that showed that only 6% of girls tested
positive, was part of a DHS survey, which routinely tests the blood of
10,000 statistically representative people. In other words, the test subjects
are selected on their representativeness of the general population.
And obviously, most of these girls who were tested weren't pregnant.
That is the difference.
Now if only HIV Insite would get clued up to this fact.
Alex
| |
| Bennett 2006-12-06, 9:26 pm |
|
Alex wrote:
> "Alex" <avdeelen.REMOFETHIS1@wanadoo.nl> schreef in bericht
> news:4561d3c3$0$29576$dbd43001@news.wanadoo.nl...
>
> And just another thing:
>
>
> So what is it exactly, that is making an ELISA false positve
> under these circumstances, and how does it do that?
Renal failure - not sure about this one but some reports put it at ~5%
prevalence in screening of dialysis patients. The only primary
reference provided in the bible of all "false positive results"
(Christine Johnson's article) was withdrawn by the authors about 6
months later. A literal handful of case reports and series exist in
the literature. In those cases where the ELISA is positive but the WB
is negative and the antigens are investigated, it appears to be
cross-reactions to antigen-presenting proteins on white blood cells
probably due to multiple blood transfusions. May not be as important
today with the widespread usage of leukodepleted blood tranfusions but
I don't think anyone has checked,.
Pregnancy - under 1% with many modern tests but very race-dependant.
Likely due to activation of endogenous RV's and antibody formation to
them (we know this happens, I just don't know if anyone has proven that
these particular antibodies cross-react with HIV ELISAs).
Flu vaccine - one batch of one year's dose from one manufacturer
reacted with some older ELISAs (in fact one component of the test kit,
not any of the viral antigens in the test kit). It was probably due to
weirdness in how that batch of vaccine was made. Not likely to be
relevant today.
IVIG (intravenous immune globulin) - simply put, massively high doses
of non-specific antibodies will cause false positive results on many
antibody tests, HIV included. The large amount of poorly-binding
antibody looks similar to small amounts of highly-specific antibody.
If you wait for the IVIG to leave the system the test results return to
normal. Not really relevant.
Any other questions?
Cheers
Bennett
| |
| Bennett 2006-12-06, 9:26 pm |
|
Alex wrote:
> "Bennett" <njb35@cantab.net> schreef in bericht
> news:1164079319.752336.172690@f16g2000cwb.googlegroups.com...
>
> You mean when the confirmation test is negative, the result is negative.
Correct. By 'test is negative' I meant 'the result of the test is
negative'.
>
>
> In the context of this discussion, a false positive is a positive screening test,
> which is then refuted by a negative confirmation test.
>
You can redefine concepts all you like to support your argument, but by
definition a positive screen/negative confirmation is a negative
result, not a false positive. The HIV test is a multi-step process.
>
> And this cured them of HIV?
No, read the bloody article. The 32.5% of girls UNDER 19 turned into
the 35-40% of women OLDER than 19 in the more recent study.
I quote:
****
The study was prompted by the results of the government's 2002
sero-surveillance study, which estimated that 32.5 percent of teenage
girls between the ages of 15 and 19 were HIV-positive.
.......
Nevertheless, the new baseline survey found a 40 percent HIV-prevalence
rate among older girls, from 19 years upwards, which matched the
sero-surveillance study results.
****
Tada!
In summary -
In 2002 32.5% of girls aged 15-19 were thought to be HIV+ (based on
antenatal screening)
In 2004 6% of girls aged 15-19 tested HIV+, with a big difference
between 18 and 19 (i.e. mostly older kids).
In 2002 AND 2004 women aged 19+ (whether antenatally screened or
systematically studied) were just under 40% HIV+
So it doesn't take a degree in epidemiology to guess that most of the
HIV burden in 2002 was in women who are now older than 19 years... In
fact the current study put the 12-18 year old prevalence at only 4%.
If the methods used in each study were so fundamentally flawed then why
does the rate in women older than 19 stay roughly the same?
You're comparing apples and oranges. The 32.5% rate in the 2002
teenagers was an estimate anyway, not a direct measurement, but also
the paper clearly states that there appears to have been a change
around 2001 in sexual attitudes and behaviours, with a reduction in the
rate of first-time sex prior to age 18. This quite possibly explains a
change in seroprevalence, as you would expect dealing with an
infectious (and not a ubiquitous environmental) cause of positive test
results.
Cheers
Bennett
| |
| Chris Noble 2006-12-06, 9:26 pm |
| Alex wrote:
> "Chris Noble" <ChrisJNoble@hotmail.com> schreef in bericht
> news:1164074737.749203.232550@m7g2000cwm.googlegroups.com...
>
>
> No I'm not. Please check again, I said 'if you look at how many test positive'.
> Most of those who test positive on ELISA are false positive (they don't stay
> positive after confirmation testing with Western blot.
This is true of any test applied to a low prevalence population. In a
low prevalence population even highly specific tests can have a low
positive predictive value. All positives resulting from testing a group
of men with a pregnancy test will be false positives.
> In Africa, you just have a lot more people who test positive. I never understood
> why it is presumed why there are only huge numbers of false positives ELISAs
> when very few people out of the total survey test positive on ELISA, than when
> many people in a survey test positive on ELISA.
If a test is 99.9% specific and you test 1000 people then you would
expect 1 false positive. If you get 300 testing positive it is very
unlikely that a significant proportion would be false positives.
> What I would like to see, is a survey that uses both ELISA for screening
> purposes, and Western Blot to confirm positive ELISAs.
>
> What is happening right now, is that DHS surveys only use Western Blot
> on ELISAs that seem neither positive nor negative.
>
> So if you know of a DHS type survey in Africa, that uses Western Blot
> as a confirmation test, that would be great.
Some surveys have used multiple ELISAs of different contructions. These
protocols are comparable to ELISA + WB.
>
> Not according to UNAIDS, who for years have used
> ELISA tests of pregnant women as a direct proxy for
> national infection rates.
Nobody simply equates the HIV prevalence from antenatal clinics to the
general population. Statistical models are used. The final figure is an
*estimate* with a confidence interval. The true figure may be
substantially lower but still way too big to deny.
> Which is why you ended up with the 33.4% of (adult) Swazis are HIV positive yarns.
>
> (Source: Adult HIV prevalence (%), 2003 33.4)
> http://www.hivinsite.org/global?page=cr09-wz-00
>
> So, the HIV Insite, which still shills the UNAIDS (political) line,
> still claims that 33.4% of adults in Swaziland (15-49 years of age)
> are HIV positive.
>
> This of course coincides approximately with the belief that
> 32.5% of pregant women in Swaziland were HIV positive.
>
> (Source:
> http://www.plusnews.org/AIDSReport.ASP?ReportID=3819&
> SelectRegion=Southern_Africa&SelectCountry=SWAZILAND
> )
>
> What you see here, and this is the criminal part, is the direct
> translation of results from pregant women, to the entire
> (adult) population.
>
> And that was always unscientific, and anti-scientific, even
> if you believe that the tests are accurate, that HIV causes
> AIDS, etc. You can believe all of that, but still, just as someone
> with a vague background in surveys and statistics, know
> that you cannot simply extrapolate the results from
> pregnant women at antenatal clinics, to everyone in the population.
>
> And that is where I think UNAIDS is criminally negligent.
>
>
> What year was that? If it was during apartheid, did they only screen
> white pregnant women?
>
> Please include the source.
http://www.doh.gov.za/docs/reports/...iv-syphilis.pdf
The first survey was in 1990.
The data is from women attending antenatal clinics and if anything
undersample white women.
>
>
> Depends on what makes the already extremely sensitive ELISA
> HIV test make positive. Whether that is exposure to DDT,
> copper, mercury, etc. I don't know.
Well, the last three words are true.
> What I do know, is that environmental exposure to banned
> pesticides, pollution, industrial spills etc. is much greater in
> Africa than in Oxford, just to mention one place.
These pollutants are definitely major health risks but it is hard to
see how they could possibly have an effect on HIV tests.
Chris Noble
| |
|
| "Chris Noble" <ChrisJNoble@hotmail.com> schreef in bericht
news:1164176805.392870.28460@m7g2000cwm.googlegroups.com...
> Alex wrote:
>
> This is true of any test applied to a low prevalence population. In a
> low prevalence population even highly specific tests can have a low
> positive predictive value.
That is just a restatement, not an explanation.
> All positives resulting from testing a group
> of men with a pregnancy test will be false positives.
What exactly is the low frequency of pregnant men in that
survey population? Could it be zero?
So perhaps you are saying that the ELISA HIV test is irrelevant
to the testing of HIV, like a pregnancy test is irrelevant to men?
Seriously though, if a test is truly 99% specific and sensitive, then
it shouldn't matter what the prevalence in the test population is.
The ppv should be exactly the same.
Unless there are factors that influence false positive results,
and that differ from population to population.
>
> If a test is 99.9% specific and you test 1000 people then you would
> expect 1 false positive.
I'm not talking about specificity (false positives out of all uninfected people)
or sensitivity, I am talking about positive predictive value, which looks at
who tests false positive out of all positive tests.
> If you get 300 testing positive it is very unlikely that a significant
> proportion would be false positives.
And if out 10,000 people, and 69 test positive?
Because that is the example we're dealing with. And of course,
the numer of false positives wasn't 1, or 10, but 40.
Also, you are presuming that specificity is a constant. However, shouldn't
specificity be directly related to the very factor that makes false positives
false positives?
If you test 10,000 ordinary citizens, shouldn't the specificity of a test
be different than when you test 10,000 pregnant women?
If you say that pregnancy is a factor in causing false positive results,
then the number of false positives should change with whatever subpopulation
you test?
And if that is true for pregnant women, what happens when you shift
your testing population to an entirely different continent?
>
> Some surveys have used multiple ELISAs of different contructions. These
> protocols are comparable to ELISA + WB.
They're not the same. And the FDA doesn't approve anything of the
sort for diagnosis if HIV/AIDS in the US.
The CDC only used Western Blot in cases where ELISA results
are 'anomalous' (their word). What they do not do, is use Western
Blot as a confirmation test for positive ELISAs, which is required
for diagnosis of individual patients, in the US.
>
> Nobody simply equates the HIV prevalence from antenatal clinics to the
> general population. Statistical models are used. The final figure is an
> *estimate* with a confidence interval. The true figure may be
> substantially lower but still way too big to deny.
Yes they did and do. Notice again the presumed 32.5% of HIV prevalence
among pregnant Swazi women at antenatal clinics, and the 33.4%
national prevalence among adults (15-49 years).
>
> http://www.doh.gov.za/docs/reports/...iv-syphilis.pdf
> The first survey was in 1990.
A lot of changes happened between 1990 and 2002. There
was a lot of urbanisation (meaning a lot of people from the
countryside were able to move to the cities).
If you look at the chart (figure 1) much closer, you will
see that the significant changes were between 1992 and
1998. From 1998 to 2002, the numbers pretty much
stayed the same.
So what happened between 1992 and 1998? Well,
the so-called independent homelands were dismanteled
and became officially part of South Africa.
We had this discussion before, when William Makgoba
used very similar data to claim that South Africa was
being decimated by an AIDS epidemic. That was back
in 2000, when the population was under 44.8 million (2001),
while it is 47.4 million today (mid-year estimate 2006).
See much of the arguments recounted in the article:
NOT-SO-YOUNG, NOT-SO-GIFTED AND DEAD WRONG!
http://www.virusmyth.net/aids/news/noseweek.htm
>
> Well, the last three words are true.
And neither do you know.
>
> These pollutants are definitely major health risks but it is hard to
> see how they could possibly have an effect on HIV tests.
So what does have an effect on HIV tests, and how?
What exactly is it that makes a false positive ELISA from a pregnant
woman false positive? What protein? What antibody?
Alex
Just to recap:
Specificity = (True Negatives)/(True Negatives + False Positives)
Sensitivity = (True Positives)/(True Positives + False Negatives)
Positive Predictive Value = (True Positives)/(True Positives + False Positives)
Negative Predictive Value = (True Negatives)/(True Negatives + False Negatives)
| |
| Moira de Swardt 2006-12-06, 9:26 pm |
|
"Chris Noble" <ChrisJNoble@hotmail.com> wrote in message
> These pollutants are definitely major health risks but it is hard
to
> see how they could possibly have an effect on HIV tests.
That's because you use your brain. Others, like Alex, don't use
theirs so they come up with all sorts of drivel.
--
Moira de Swardt posting from Johannesburg, South Africa
Remove the dot in my address to find me at home.
| |
| Chris Noble 2006-12-06, 9:26 pm |
|
Alex wrote:
> "Chris Noble" <ChrisJNoble@hotmail.com> schreef in bericht
> news:1164176805.392870.28460@m7g2000cwm.googlegroups.com...
>
> That is just a restatement, not an explanation.
>
>
> What exactly is the low frequency of pregnant men in that
> survey population? Could it be zero?
That is exactly my point.
>
> So perhaps you are saying that the ELISA HIV test is irrelevant
> to the testing of HIV, like a pregnancy test is irrelevant to men?
No. We can safely say that men are not pregant without giving them a
pregnancy test. The same is not true for women even in low prevalence
populations.
> Seriously though, if a test is truly 99% specific and sensitive, then
> it shouldn't matter what the prevalence in the test population is.
> The ppv should be exactly the same.
No. You obviously do not understand what the term means.
http://en.wikipedia.org/wiki/Positive_predictive_value
PPV is a function of prevalence!
> Unless there are factors that influence false positive results,
> and that differ from population to population.
>
>
>
> I'm not talking about specificity (false positives out of all uninfected people)
> or sensitivity, I am talking about positive predictive value, which looks at
> who tests false positive out of all positive tests.
>
>
> And if out 10,000 people, and 69 test positive?
>
> Because that is the example we're dealing with. And of course,
> the numer of false positives wasn't 1, or 10, but 40.
>
> Also, you are presuming that specificity is a constant. However, shouldn't
> specificity be directly related to the very factor that makes false positives
> false positives?
> If you test 10,000 ordinary citizens, shouldn't the specificity of a test
> be different than when you test 10,000 pregnant women?
> If you say that pregnancy is a factor in causing false positive results,
> then the number of false positives should change with whatever subpopulation
> you test?
> And if that is true for pregnant women, what happens when you shift
> your testing population to an entirely different continent?
>
As we saw from the famous "Voevodin letter" the false positive rate
froma a single ELISA is only marginally higher in pregant women.
It is within the realms of possibility that the false positive rate in
pregnant women is higher in Africa but the assertion needs evidence
rather than ad hoc reasoning.
Personally, the reason why I find this scenario extremely unlikely is
that the seroprevalence of women attending antenatal clinics in South
Africa in 1990 was less than 1%.
You need to provide a rational explanation for why the false positive
rate in preganant wome should increase dramatically over time.
Ad hoc explanations that ultimately stem from your initial assertion
that they are false positives do not count.
>
>
> They're not the same. And the FDA doesn't approve anything of the
> sort for diagnosis if HIV/AIDS in the US.
>
> The CDC only used Western Blot in cases where ELISA results
> are 'anomalous' (their word). What they do not do, is use Western
> Blot as a confirmation test for positive ELISAs, which is required
> for diagnosis of individual patients, in the US.
Can you provide some evidence for this statement.
>
> Yes they did and do. Notice again the presumed 32.5% of HIV prevalence
> among pregnant Swazi women at antenatal clinics, and the 33.4%
> national prevalence among adults (15-49 years).
Can you provide the references to the actual study and the methods
used. You have simply pulled out two figures from different sources and
implied a connection.
>
>
> A lot of changes happened between 1990 and 2002. There
> was a lot of urbanisation (meaning a lot of people from the
> countryside were able to move to the cities).
>
> If you look at the chart (figure 1) much closer, you will
> see that the significant changes were between 1992 and
> 1998. From 1998 to 2002, the numbers pretty much
> stayed the same.
> So what happened between 1992 and 1998? Well,
> the so-called independent homelands were dismanteled
> and became officially part of South Africa.
> We had this discussion before, when William Makgoba
> used very similar data to claim that South Africa was
> being decimated by an AIDS epidemic. That was back
> in 2000, when the population was under 44.8 million (2001),
> while it is 47.4 million today (mid-year estimate 2006).
>
> See much of the arguments recounted in the article:
> NOT-SO-YOUNG, NOT-SO-GIFTED AND DEAD WRONG!
> http://www.virusmyth.net/aids/news/noseweek.htm
>
>
>
> And neither do you know.
>
>
> So what does have an effect on HIV tests, and how?
It depends on the test. Antibodies to other retroviruses.
>
> What exactly is it that makes a false positive ELISA from a pregnant
> woman false positive? What protein? What antibody?
>
Probably the expression of a HERV. I don't know which proteins
generate antibodies that cross react with HIV antigens.
Chris Noble
| |
|
|
"Moira de Swardt" <moira.ds@wol.co.za> schreef in bericht
news:Cs6dndTqipuWBPnYnZ2dnUVZ_qWdnZ2d@is.co.za...
>
> "Chris Noble" <ChrisJNoble@hotmail.com> wrote in message
>
> to
>
> That's because you use your brain. Others, like Alex, don't use
> theirs so they come up with all sorts of drivel.
Oh pray tell, Moira, how is Chris 'using his brain' when he
states it is 'hard to see how they could possibly have an effect
on HIV tests'?
Are the facts know what, and especially HOW, unknown
factors are responsible for false positives when using the
HIV ELISA test?
Time to use YOUR brain, Moira.
Alex
| |
|
| "Bennett" <njb35@cantab.net> schreef in bericht
news:1164153493.156970.196650@e3g2000cwe.googlegroups.com...
>
> Alex wrote:
>
> Renal failure - not sure about this one but some reports put it at ~5%
> prevalence in screening of dialysis patients. The only primary
> reference provided in the bible of all "false positive results"
> (Christine Johnson's article) was withdrawn by the authors about 6
> months later. A literal handful of case reports and series exist in
> the literature. In those cases where the ELISA is positive but the WB
> is negative and the antigens are investigated, it appears to be
> cross-reactions to antigen-presenting proteins on white blood cells
> probably due to multiple blood transfusions. May not be as important
> today with the widespread usage of leukodepleted blood tranfusions but
> I don't think anyone has checked,.
>
> Pregnancy - under 1% with many modern tests but very race-dependant.
> Likely due to activation of endogenous RV's and antibody formation to
> them (we know this happens, I just don't know if anyone has proven that
> these particular antibodies cross-react with HIV ELISAs).
>
> Flu vaccine - one batch of one year's dose from one manufacturer
> reacted with some older ELISAs (in fact one component of the test kit,
> not any of the viral antigens in the test kit). It was probably due to
> weirdness in how that batch of vaccine was made. Not likely to be
> relevant today.
>
> IVIG (intravenous immune globulin) - simply put, massively high doses
> of non-specific antibodies will cause false positive results on many
> antibody tests, HIV included. The large amount of poorly-binding
> antibody looks similar to small amounts of highly-specific antibody.
> If you wait for the IVIG to leave the system the test results return to
> normal. Not really relevant.
>
> Any other questions?
>
> Cheers
>
> Bennett
Sure, and thanks for the substantive answers.
I was thinking of what makes these tests false positive
on a basic level.
http://www.biology.arizona.edu/immu...technique.html?
According to this website, ELISA HIV antibody testing works as follows:
1) Partially purified, inactivated HIV antigens pre-coated onto an ELISA plate
Patient serum which contains antibodies.
2) If the patient is HIV+, then this serum will contain antibodies to HIV, and those
antibodies will bind to the HIV antigens on the plate.
3) Anti-human immunoglobulin coupled to an enzyme. This is the second antibody,
and it binds to human antibodies.
4) Chromogen or substrate which changes color when cleaved by the enzyme
attached to the second antibody.
So my question is, what exactly goes wrong during this procedure,
that a woman's pregancy can cause a false positive test result? What
(if anything) is attaching to the HIV antigens to cause a false positive?
Alex
| |
| Moira de Swardt 2006-12-06, 9:26 pm |
|
"Alex" <avdeelen.REMOFETHIS1@wanadoo.nl> wrote in message
> Oh pray tell, Moira, how is Chris 'using his brain' when he
> states it is 'hard to see how they could possibly have an effect
> on HIV tests'?
Because it is hard to see how they could possibly have an effect on
HIV tests unless one is daft or stupid.
> Are the facts know what, and especially HOW, unknown
> factors are responsible for false positives when using the
> HIV ELISA test?
While false positives to the ELISA certainly exist, there are very
few false positives even to the ELISA. No-one in South Africa is
*diagnosed* with an ELISA, a situation which has existed since
almost the beginning although ELISA tests were used for statistical
purposes.
> Time to use YOUR brain, Moira.
I've been using it since the beginning, but I'm not a denialist.
--
Moira de Swardt posting from Johannesburg, South Africa
Remove the dot in my address to find me at home.
| |
| DavidT 2006-12-06, 9:26 pm |
|
Alex wrote:
> So my question is, what exactly goes wrong during this procedure,
> that a woman's pregancy can cause a false positive test result? What
> (if anything) is attaching to the HIV antigens to cause a false positive?
Sigh!
You can lead a horse to water and all that...
Read Bennett's post (#3) again......
Go to the explanation he gave you 6 years ago......
(http://groups.google.com/group/misc...35f320fc95ac5a6)
If there are circulating antibodies to these retroviral fragments in
maternal serum, then when an HIV ELISA is performed they may bind to
one or more of the HIV antigens used in the assay, causing a biological
false positive reaction.
| |
| DavidT 2006-12-06, 9:26 pm |
|
Alex wrote:
[vbcol=seagreen]
> Depends on what makes the already extremely sensitive ELISA
> HIV test make positive. Whether that is exposure to DDT,
> copper, mercury, etc. I don't know.
>
> What I do know, is that environmental exposure to banned
> pesticides, pollution, industrial spills etc. is much greater in
> Africa than in Oxford, just to mention one place.
Your hypothesis seems to be that something which you claim is commoner
in Africa (environmental pollution) is the cause of something else you
claim is commoner in Africa (false positive HIV tests).
Usually when someone proposes a hypothesis there is an underlying
plausible explanation. What is yours? Do you have any reason to believe
that pollution might lead to false positive serological tests? Please
share the evidence with us, and we can decide for ourselves.
Why do you think there is more pollution in Africa? It is a vast
continent, largely unsullied by the industrialisation we have in the
West. Sure, some localised spots are very contaminated because of local
industries and toxic waste dumping. But in the 3rd world one is more
likely to find widespread industrial pollution in countries like Brazil
with their unusual mining methods. And if you really want to see bad
pollution, go to countries like Estonia. Do they have high rates of
false positive HIV tests?
And why single out pollution as a cause anyway? Why not pick on
something else common in Africa, like sunshine?
| |
| DavidT 2006-12-06, 9:26 pm |
| Also give this paper a read - it goes through many of the issues under
discussion here.
http://archfami.ama-assn.org/cgi/reprint/9/9/924.pdf
You must appreciate that false positive results are rare, and not the
norm, as denialists would have us believe.
In this paper, you will see the false positive prevalence in 5 million
blood donors was 0.0004% (1 in a quarter of a million).
http://jama.ama-assn.org/cgi/reprin...24d856c70a9f380
This also has good explanations about the PPV questions you have.
| |
| Bennett 2006-12-06, 9:26 pm |
|
DavidT wrote:
> And why single out pollution as a cause anyway? Why not pick on
> something else common in Africa, like sunshine?
Heck, that might explain why the HIV prevalence rate in the UK is
substantially lower than that in the US :o)
Cheers
Bennett
| |
| DavidT 2006-12-06, 9:26 pm |
| Perhaps he subscribes to the theory:
"Don't blame it on the sunshine,
Don't blame it on the moonlight,
Don't blame it on the good times,
Blame it on the boogie."
|
| |
|
|