Sunday there was on tv in my country about south-africa. Twenty percent of
the population is infected, it was said. Today I read on www.aegis.com that
only ten percent of the infected are tested. The rest are estimates. The
main stream media presents estimates as facts.
Do anyone here know which test was used in the beginning of the epidemic in
the western countries?

linea wrote:

Read these links and weep or laugh satanically, or do both at the same time.

http://www.virusmyth.net/aids/index/africa.htm

Hayek.

"linea" <traviata@hotmail.com> schreef in bericht
news:jaYgb.36192$Hb.564125@news4.e.nsc.no...
Actually, they are _both_ estimates and extrapolations,
but the 11.4% percent is based on better research.

However, neither number is accurate, because they never use
confirmatory tests in these surveys. The real number is likely
to be closer to 2-5%. If that.

Here is the lowdown. Everything is dependent on cash. The
WHO (19% infection, etc.) has very little of it. As a result,
they depended on the collection of blood from women at
antenatal clincs, which was very cheap, because these
clinics already existed and blood was already taken anyway.
However, the result was that they only had data from pregnant
women, who are unrepresentative of general population.
They are younger, they are sexually active (duh) and
they are all female. And pregnant, which brings me
to the second point - the tests used.
The cheapest test out there is called ELISA (actually
a group of tests). The problem with them is, that they're
not very reliable, because they will "spike" (go positive)
when you've been exposed to any of 70 known factors
and pathogens. Malaria, tb, ddt, the common cold, the flu,
herpes, will all make these ELISA tests spike. But the main
reason that these tests never should have been taken
as the final word, is that they "spike" because of...
<drumroll>... pregnancy!!

Then, the South African HSRC came along, with a
little more money, and they took a random population
sample of 10,000 or so people, who they gave
a single, orally administered ELISA. Just for them
being representative and unpregnant, the result was
that 11.4% of the population was now officially HIV
positive.

That's a dive of 40%, just by doing the test
a little right.

The real problem that has remained and still isn't
solved, is that none of the positive testing individuals
were followed up by what are called confirmatory
tests. Usually, that would be another ELISA, just
to see if the first one was right. If this second ELISA
is negative, the positive first test is basically thrown
out of the window (so to speak), and the patient
is considered HIV negative.
If the second is positive, then a Western Blot test
is performed. This is a more specific and also more
expensive test, which is why it is never used in surveys
in Africa. ELISA has an extraordinary false positive
rate, especially in Africa.
However, this combination of tests has previously
resulted in one Italian military transfusion center of
only 6.25% of the original positive testing individuals,
who remain being considered HIV positive.

So you can see. 20% estimated national HIV infection
from blood from pregnant women.
The same test on the general populace - 11.4% infection.
Follow up tests... 0.7% infection rate?

Who knows until the tests are actually done, but it seems
very reasonable to assume. (11.4% times 0.0625 or 1/16th)

Yes they do, all the time.

Probably some form of ELISA.

Alex

" "One in five heterosexuals could be dead from AIDS at the
end of the next three years," the nation's most popular talk
show host Oprah Winfrey warned her audience in 1987. "
http://www.fumento.com/realaids.html

"Alex" <avdeelen.REMOF@wanadoo.nl> wrote in message

Not true according to what we are seeing in the field.

The reasons for using antenatal clinics are varied, but one is that
it is one way of getting blood supplies from people who are
otherwise usually healthy people. The cheapness thereof relates to
the fact that blood samples have to be drawn for sexually
transmitted diseases and blood typing anyway, so there are no extra
*clinical* costs involved in drawing blood.

At most well run clinics pregnant women are counselled to include
formal HIV tests (where they get the result, not random ones) so
that, if they are HIV positive, nevirapine can be administered at
the appropriate time (just before delivery and then, for the baby,
after the birth).

Now, the ELISA test does sometimes give a false positive in the
presence of other pathogens. What you have done is take a fact and
then manipulate it to try and prove what you want to believe.

Now ELISA never tests false negative, so any blood samples testing
negative (and remember that this is the majority of blood samples)
can be discarded. Every blood sample in South Africa that tests
positive to an ELISA test is retested on the more expensive Western
Blot test. This includes bloods taken for antenatal sampling, and
bloods taken for random sampling (I have participated in the latter
and it is amazing to see one's blood sample just being dropped,
unmarked, into a container which will later be tested purely for
statistical purposes).

No, it is not a dive of 40%. Even at antenatal clinics there has
not been a 50% or more test result ever, not even in the most highly
infected areas and groups.

Now when test results are released they indicate that they are drawn
from antenatal clinics (our official rate) or from other sources,
for example, "In a test of mineworkers conducted by ...", or "In a
test of 36 matric pupils from the same class at one school in Soweto
...."

ELISA tests don't give false negatives. Now what you have indicated
here is simply not true. The same sample that tests positive on an
ELISA is run on a Western Blot test. These are highly accurate.

Not true. They are used as a standard in South Africa.

From what I hae been advised is that the ELISA test has a slightly
higher rate of positive results than are confirmed by the Western
Blot. I forget what the differential is, but it is hardly an
"extraordinary false positive rate".

No, it doesn't seem very reasonable to assume anything of the sort.
Those of us who live and work in South Africa are starting to have
friends, family and colleagues who have died of HIV related
illnesses. This evidence is a clear indication that the problem is
bigger than the 1 -2% you are claiming. We would like to have a low
rate of HIV infection, but self-delusion is not the answer to that
desire.

Moira, the Faerie Godmother

| However, this combination of tests has previously
| resulted in one Italian military transfusion center of
| only 6.25% of the original positive testing individuals,
| who remain being considered HIV positive.
|
| So you can see. 20% estimated national HIV infection
| from blood from pregnant women.
| The same test on the general populace - 11.4% infection.
| Follow up tests... 0.7% infection rate?
|
| Who knows until the tests are actually done, but it seems
| very reasonable to assume. (11.4% times 0.0625 or 1/16th)


Stated another way. IF there was ONE person in a million that had
AIDS. Say Moira then somehow came up with a list of TEN possible
people out of this million, you'd say that's it's B/S as she only
had a 10% success rate. It's a little cheaper, and smarter
testing 10 people, than testing all the million people, however.

Now if the population was a billion, Moria still came up with a
list of ten people and only the same one guy had aids. Then the
success rate of the MOIRA test according to Alex would be 10%
still.

The way I see it it that ELISA can be used to do the rough work,
then fine tuning done by WB? How do you think the Italian
military you refer to came up with the 6.25% figure? ELISA and
then a WB of course. ELISA says who almost surely doesn't have
AIDS, not who surely does.

Another point - those tests were done in 1985. Aids wasn't so big
then. Also the ELISA tests have improved drastically since then.
So have cellphones, computers etc. You're still running an 8088
or a 286 today?

If there really was an infection rate of 30% that 6,25% ppv would
change to >80%. even for those old ELISA tests.

HIV+ HIV- total
ELISA SAYS + 2850 70 2920
ELISA SAYS - 150 6930 7080
3000 7000 10000

10000 is a sample population.
10000 * 0.30 = 3000 ( assuming 30% really have AIDS)
7000 * 0.94 = 6930 ( assuming the OLD elisa assays got 94%
accuracy in their negatives)
3000 * 0.99 = 2850 ( assuming this ELISA 99% specific )

2850/2920 = 88% ([really HIV+ overlaps with ELISA+ result] /
[total ELISA+ result])

AND NO I HAVE VERY LITTLE MEDICAL KNOWLEDGE. I don't even have
Level 1 1st aid. I got this method off a stats site, which was
probably demonstrating P(A given B) not the same as P(B given A)

|
| > The main stream media presents estimates as facts.
|
| Yes they do, all the time.
|

As opposed to taking a census. Random samples are used in stats
for this very reason.

| > Do anyone here know which test was used in the beginning
| > of the epidemic in the western countries?
|
| Probably some form of ELISA.
|

"Probably some form of ELISA"
Which means you're taking a guess, and this isn't your field
after all...

Rob

Nevirapine can be very dangerous for the liver. During oneandhalf a year
there were reported to WHO 62 cases of death - related to nevriapine.
Knowing that the frequence of reporting on side-effects is very low all
over - and that the virus can be blamed for every side-effect, I find it
strange to go for nevirapine to newborn babies. Besides, the baby kick on
to its own immunesystem when it is about eighteen months.
Another thing; are the positiv tested mothers told not to breastfeed their
children? What do they give them then? Do anyone know if the virus is found
to exist in breastmilk?
Some of us remember the seventies - when companies sold
breastmilk-substitute in Africa - telling mothers that if they used this,
their children would be healthy. Many babies died because first of all, the
mothers were not told to boil the water.
A friend of mine visited her homecountry in Africa. She brought her meds
with her. But there was no fridge where she stayed, so the meds could not be
used.

.."Moira de Swardt" <moira.deswardt@wol.co.za> skrev i melding
news:hSqdndipH_doExqiXTWJig@is.co.za...

"Moira de Swardt" <moira.deswardt@wol.co.za> schreef in bericht
news:hSqdndipH_doExqiXTWJig@is.co.za...
Actually, it is the most likely range for HIV infection,
and probably a little too high.

Yes, it is a dive of 40%. You went from an estimated national
prevalence of 19% to 11.4%

11.4 / 19 = 0.6 which is 0.4 or 40% lower.

http://www.independent.co.za/index.p...3345189C308684



What I have indicated above is the simple, standard procedure
of how tests are supposed to be done. The fact is that the
standards of testing in surveys is a lot lower than when
HIV infection is diagnosed in an individual. And that shouldn't be.

As to false negatives, they don't occur according to the
WHO. However, it is false positives that are the issue.

Also, these SURVEYS have no confirmatory tests
performed. Certainly not Western Blot, whatever
may happen at _some_ South African private clinics.

For the WHO data I know this for a fact, because
I e-mailed them personally.

Again, not in these large surveys. And not in the
HSRC study, which is the most accurate of the lot.

As is clear from the data below, Moira, in these Italian data,
there are 31 people who tested positive, only *2* remained
positive after further testing.
That _is_ an extraordinary false positive rate.

Table II
HIV Screening in Military Blood Transfusion Centers

Number of Blood Donations: 25,562
Number of blood donations ELISA positive: 31
Number of blood donations after confirmation test: 2

http://www.certi.org/CMA/newsletter/v03n01.pdf

Alex

"linea" <traviata@hotmail.com> wrote in message


And in the same eighteen months how many babies who would otherwise
have been HIV positive are now HIV negative?
All meds have side effects, some of which can be fatal. The pay off
often makes the risk worthwhile.

And the baby may sero-convert or may not. Babies who have been
given a single dose of Nevirapine after birth when the mother was
given a single dose of Nevirapine just before the birth have a much
higher rate of sero-conversion than babies untreated by Nevirapine.

HIV positive mothers are told not to breastfeed their children if
possible. Clinics now provide formula free of charge where this is
a prophylactic treatment. This assumes that the mother can provide
proper artificial feeding. Otherwise exclusive breast feeding is
recommended. The mother is then counselled *not* to supplement the
breast feeding with anything else. Combination of breast and bottle
is the most dangerous because the formula increases the risk of the
virus which *is* found in breast milk being introduced through the
stomach lining. I am not 100% how this works, but for HIV positive
mothers no breast is best (assuming sterility, proper mixing etc.)
and exclusive breast feeding is nearly as good. Mixed breast and
formula is the most risky.

That is true. But before AIDS there was no reason for substituting
formula for breast except for the financial gain of the companies
concerned. For HIV negative women there is no reason to use formula
and many reasons not to.

Surely she knew, if it was her home (country), that there would be
no fridge? An HIV positive friend of mine travels to Uganda,
Kenya, Zimbabwe, Zambia, Senekal, Nigeria, Mocambique, Swaziland and
Lesotho (amongst other African countries) doesn't have any problems
with his meds whatsoever. He leaves home with this meds in a cooler
bag with ice. When he arrives at his destination the meds are
simply put into the fridge. Where there is no fridge for a short
space of time (like an overnight visit to a remote village) he
simply buys ice and keeps his meds in the cooler bag. He would not
go to a place where there was no fridge or electricity, although
such remoteness doesn't tend to get many Western visitors (I realize
your friend was visiting her own home) simply because his life
depends on the correct taking of his meds, and part of that is the
correct storage of his meds. I have travelled with him, and have
watched him monitor the storage of his meds.

Moira, the Faerie Godmother

Again - when I read the newspaper in South Africa - to which Alex gave the
URL; they take about estimates as facts.
The rate of HIV is 11,5 pst, they write - and they continue with saying that
this means that 4,5 mill africans are hivpos. But this is not true, it's
estimates. If less than ten percent of the estimated infected population are
tested, we know then that 450 000 persons are tested hivpos.
I think we all should be more sceptic than we seem to be to statistics and
its estimates. Remember what was estimated in Uganda som years ago - and
what was said in the beginning of the epidemic in US and Europa(when they
used ELISA)
A person I know thought for two weeks that he was hivpos. He was told so
when he should donate blood. Then the WB was taken and came out negative.He
was told that a flue alone could make a ELISA pos.
So - what about Africa?
And - is this WB to trust so much?
"Alex" <avdeelen.REMOF@wanadoo.nl> skrev i melding
news:3f88403a$0$64180$1b62eedf@news.euronet.nl...
6063345189C308684

PS,

About the WHO - they even stipulate on their site, that
in populations where HIV infection is _assumed_ to
be greater than 10%, they use "strategy 1", or
a single ELISA with no confirmatory testing, in
their surveys. (See Table A at page 7)

"Strategy I
All serum/plasma is tested with one ELISA or
simple/rapid assay. Serum that is reactive is considered
HIV antibody positive. Serum that is non-reactive
is considered HIV antibody negative. "

http://www.who.int/bct/Main_areas_of...V_Diagnostics/
Evaluation_reports/Operational%20Characteristics_HIV%20Report9_10.pdf

The HSRC is doing the same.

So Moira, whether Western Blot is available in
clinics in South Africa, it is _not recommended
or required_ by the WHO when they create
their antenatal data based estimates.
The WHO makes a very clear distinction between
testing for surveillance purposes, and the diagnostic
testing of individual patients.

And it's all money related. To quote:

"UNAIDS and WHO recommend three testing strategies,
which have been recently updated, to maximize accuracy
while minimizing cost. "

And what's more, in countries where they _assume_ infection
rates of greater than 30%, only one single ELISA will do
in diagnosis.

Alex

"Rob MacMillan" <greencalx@remove-me.mighty.co.za> schreef in bericht
news:bm8lmd$pnu$1@ctb-nnrp2.saix.net...
She did? Just kidding.

Not for finding out the overall infection rate. For that, you need
a large sample size, that is representative (age, gender, geography,
income, etc.) of the entire population.
Pregnant women are not representative of the entire population.
Which is why when using one single (orally adminstered) ELISA,
the HSRC ended up with a much lower estimate than had previously
been arrived at.

Robbo, no one is disputing the use of sampling or using estimates.
However, to be relevant instead of misleading, they have to be
done correctly.



Of course. The point being that a single, positive ELISA
should never be the last word.
It _was_ the last word in the WHO data, AND in
the HSRC study.

That's not what they said in 1985...

And at 11.4%? (Single ELISA?)

But the operative word being *if*. And how do we know?
Because of the tests.

Here something to mull over. If these tests were so great,
wouldn't they be accurate *no matter what* the level of
infection in the general population is?

Also, the assumption is that there is either a fixed number
of fixed percentage of false positives to the overall number
of tests. I think false positives are much more variable, and
reliant on the presence of factors that _cause_ false positives.

It makes all the sense in the world to assume that false
positives have something to do with the factors that
cause them - African strains of malaria, ddt use, etc.

Things that aren't occuring in Italy, for instance.

(You mean HIV.)

Which is where things go wrong. The _assumption_ of
HIV infection in the general populace. It was 19%. Now
it's 11.4%. Based on...? (1.5% less, and the WHO would
have to do follow-up tests in it's antenatal data.)

Where they go wrong is the a priori presumption of a high infection
rate. Rationally speaking, there is no reason to _ever_ assume a
high HIV infection rate. Not before the tests say so.

That's not what the previous poster was alluding to.
When it's estimated that one woman is raped every
thirty seconds, it is stated in the media that one
woman *is* raped (not "estimated to be raped")
every thirty seconds.

That's what he meant, he wasn't criticizing the use
of estimations.

And by the way, my field of specialty is political science,
so statistics is within my area of expertise.

Alex

I don't believe a lot of stuff you write considering it (HIV
stuff) comes from sites such as virusmyth etc. Maybe to argue in
your favour, websites directly disputing what virusmyth, perth
group etc write just don't seem to exist... If there was an
advisory on the WHO site explaining where 'dissident' sites are
going wrong, less misinformation would be passed around.
Personally I don't really care that much about the whole thing,
as I've got a lot of other things to worry about. And I'd be
arguing about something I have little clue about anyway.

If you were arguing against GM foods, yes, I'd agree. Regarding
AIDS - Treating people with AIDS is big money, but that hits the
medical aids and private companies AFAIK. Prevention, where
possible, would be a better solution and means the problem could
be controlled.

You've got some points about the surveillance, and I'm neutral on
that. BUT I'm assuming they've already taken what you've written
into account when they make the estimates.

I agree that things should be done properly - IMO those with HIV
should get proper medication and councelling, not just a single
ELISA and sent away. THAT would be big money, but life should be
more important than paper. But then poor countries usually
resulted from governments who couldn't care less in the first
place.

Rob

"linea" <traviata@hotmail.com> schreef in bericht
news:S4Zhb.29906$os2.436285@news2.e.nsc.no...
That's very common. (See the Italian data below.)

No such luck. Just the single ELISA, as according to the WHO.

Western Blot has problems of it's own, but at least it is more specific
than the ELISA. One problem is that it isn't standardized, but at least
it is accepted that in Africa, two of the three ENV proteins being
positive equals a positive test (two of p41, p120, p160). GAG (p18,
p24, p39 and p55) and POL (p32, p53 and p68) positive bands are
discarded and ignored.
This is different in different parts of the world, however. In fact
the standards for a positive test in Africa are lower, as only
2 (ENV) positive bands are required - at least 3 in the rest
of the world.

Alex

"Rob MacMillan" <greencalx@remove-me.mighty.co.za> schreef in bericht
news:bma4i3$60c$1@ctb-nnrp2.saix.net...
Well that's simply not true, because I've only been quoting from WHO.int,
certi.org, etc. websites.

Besides, "it's from Virusmyth.net so it isn't true" simply isn't an argument.

Alex

On Sat, 11 Oct 2003 19:54:13 +0200, "Moira de Swardt"
<moira.deswardt@wol.co.za> wrote:

Indeed, it is unlikely that these deaths are attributable to the
single dosages taken by pregnant mothers but rather from cumulative
toxicities of standard use in an ARV combination. Pre-existing
conditions like HCV or chronic HBV or alcohol consumption may
exacerbate this toxicity.

George M. Carter

"Alex" in news:3f88403a$0$64180$1b62eedf@news.euronet.nl :

If HIV is only a 10%-killer virus, then a 20-50% infection
rate leads to the same (death-)outcome as a 2-5% infection
rate in the case of a 100% killer virus.

Alex, I wonder why you are so convinced that HIV is a
100%-killer-virus, and why you provide AIDS orthodoxy
with a simple and convincing excuse for the troubling but
obvious fact that death rates southern Africa (where HIV
always has been endemic) are relatively low.

In the beginning of 2000 (when holding the presidency of the
UN Security Council) the US declared AIDS in Africa a
national security thread. Look for instance at the effect
of this absurd proclamation on the exchange rate of the
South African currency. AIDS remains a thread to the nation
reponsible for the AIDS hysteria, but the Rand has already
recovered.


Cheers, Wolfgang
http://groups.google.com/groups?selm=bg9jj4$hbs$1@rex.ip-plus.net

"Alex" <avdeelen.REMOF@wanadoo.nl> wrote in message
news:3f88bf30$0$64200$1b62eedf@news.euronet.nl...

| > I don't believe a LOT OF STUFF you write considering it (HIV
| > stuff) comes from sites such as virusmyth etc.
|
| Well that's simply not true, because I've only been quoting
from WHO.int,
| certi.org, etc. websites.
|
| Besides, "it's from Virusmyth.net so it isn't true" simply
isn't an argument.
|
|
I said a "LOT OF STUFF" not ALL THE STUFF. Learned something
new regarding the surveillance. But quite honestly just because
you're right about one thing, doesn't make you right about
everything. Skeptical would have been a better word to have used.

I didn't say "it's from Virusmyth.net so it isn't true" either
although that is my opinion quite honestly. I found that site
almost 2 years ago and was shocked. However at the time, I had
friends who were doing masters in microbiology, and I'm taking
their word over these sites.
I've found a lot of misinformation over the internet, so if I
think it looks wrong, then I'm skeptical.

Again, I don't really care about the issue. I'm not going to
pretend to know more than I do.

Some 'dissidents' deny the link between HIV and AIDS, some demand
ALL research be banned immediately. Personally, I believe HIV
causes AIDS. More importantly I also believe there is no cure, so
I'm going to try avoid getting it in the first place. I don't
believe deodorants, amyl/alkyl nitrates, poverty etc cause AIDS.
Vitamin C supplements won't cure AIDS. Neither will raping
anybody.

You often quote Rian Malan's Rolling Stone article and link to
virusmyth. The substance of this thread is the same as the others
you've written that do, only the "WHAT IS AIDS" is not linked in
this one.

Rob

"Wolfgang G. Gasser" <siehe@homepage.li> schreef in bericht
news:bmbtdj$5c5$1@rex.ip-plus.net...
Hi Wolfgang,

Why do you assume that the "tests" are accurate?
There are lots of reasons why an inaccurate test like
ELISA would go off, known reasons, like pregnancy,
malaria, ddt, etc.

So as far as I'm concerned, there is no basis for
assuming that HIV is endemic in Southern Africa.
No rational, sociological or demographic reason
either. HIV doesn't spread more easily between
men and women in Africa than in Europe, America,
etc. as serodiscordant couple studies have shown again
and again.
No greater sexual activity, as sociological studies
have shown again and again.

Well it are signs like that, which show that the disbelief
of the apocalyptic paradigm of the AIDS industry isn't
limited to the government.

Alex

"Alex" <avdeelen.REMOF@wanadoo.nl> wrote in message
I have worked in the field of emergency medicine and/or HIV/AIDS as
either a volunteer or in full-time paid employment since 1984. Back
in the early 90s I was "seeing" about 2% of patients that I
*thought* were HIV positive (already had AIDS) based on clinical
observations. In 2003 I am seeing about 50% of patients that I
*think* are HIV positive based on clinical observations. Granted
this is related to people who are already in trouble so the estimate
is not that 50% of people are HIV positive, but that 50% of sick
people who are in contact with my particular type of work are HIV
positive and already evidencing signs of AIDS.

I certainly think that more than 5% of the population is HIV
positive with the likely range being closer to that of the official
statistics.

OK, the percentage figure refers to the change, not to the total.
That's somewhat more plausible.


I am not sure that it *is* lower for the diagnosis. ELISA followed
by a Western Blot is pretty conclusive. What happens with
individuals who can afford private medical care is that viral load
tests are usually also run immediately so as to determine what
treatment, if any, to initiate.

It is my understanding that Western Blot tests are run on all
positive samples. At *all* private clinics there would be either a
Western Blot and/or a viral load test. Remember the clinic gets
paid for the onerous task of ordering the test.

And this has been discredited by another poster. See that post for
the discreditation of your "research".

Moira, the Faerie Godmother

"Moira de Swardt" <moira.deswardt@wol.co.za> schreef in bericht
news:Ye2cnZr0JaCWPxSiU-KYvg@is.co.za...
Wow, that's amazing. Do you clinically/visually diagnose cancer too?

But seriously, is it possible that what you're seeing, and a relatively
new influx of poor people from the countryside? A change in
demographics after the lifting of restrictions?

But they're not tested...

Which keep changing.

No, it's a dive of 40%. ;-)


Yes but, there is a big difference diagnosing individuals and
testing these large surveys.

In these large surveys **confirmatory tests aren't done**.



" Strategy I
All serum/plasma is tested with one ELISA or simple/rapid assay.
Serum that is reactive is considered HIV antibody positive. Serum
that is non-reactive is considered HIV antibody negative. "


Page 7

Table A UNAIDS and WHO recommendations for HIV
testing strategies according to test objective and prevalence
of infection in the sample population

Surveillance >10% I
<10% II

Diagnosis
clinical symptoms/
signs of HIV infection >30% I
<30% II


The WHO doesn't depend on posh private clinics, but on (leftover) blood
collected at antenatal clinics.

And they are very specific. Countries where HIV infection is assumed
to be greater than 10% (like South Africa), "strategy 1" (page 7) is
carried out, which is a single ELISA test, nothing more. If it's positive,
the sample is assumed to be HIV positive.
In countries where HIV is assumed to be greater than 30%, this
strategy 1 is even used for diagnosis.

http://www.who.int/bct/Main_areas_of...V_Diagnostics/
Evaluation_reports/Operational%20Characteristics_HIV%20Report9_10.pdf

This document is from 1998.

Really? What did he say? That in places with low prevalence,
the huge majority of positive ELISAs are false? I already knew
that, and it's even spelled out in the WHO report and in their
own words (page 7):

" When a single screening assay is used for testing in a population
with a very low prevalence of HN infection, the probability that
a person is infected when a positive test result is obtained (i.e.,
the positive predictive value) is very low, since the majority of
people with positive results are not infected. "

The problem with using this as an excuse of not assuming the same
for so-called high infected regions, is:

1) you have to assume a high infection rate before testing starts and
2) it doesn't take into account a local high prevalence of factors
that will cause false positives and that are unique to that local area.
(Malaria, tb, ddt [outlawed in most places], leprocy, all kinds of
bacteria, and of course the exclusive use of blood samples from
pregnant women.)

In fact, it assumes that there is a fixed, limited number of
false positives out there. When in fact the number of people
who have been exposed to factors that will make them
test false positive will vary hugely, and increase hugely in Africa
relative to Europe or North America.

Alex