Talk:Diagnosis of HIV/AIDS

From Wikipedia, the free encyclopedia
Jump to: navigation, search

Negative HIV Antibody Test Results Among People Treated During Acute/Early Infection[edit]

The 9th reference of the article points to a paper of the "The XV International AIDS Conference" that is no longer available. I was able to find a reference to it and further explanations on

I believe it would be important for someone with more insight on the matter to point out in the main article that these negative results do not apply to post exposure prophylaxis.

The reason for this suggestion is that the referenced paper may incline or lead one to believe that after a PEP treatment the hiv test wouldn't be as reliable.

Subjects who discontinued ART in this study did indeed experience viral rebound. Equally important is the need to recognize that all these patients had detectable HIV in their plasma by PCR prior to ART initiation -- these were not cases of post-exposure prophylaxis. 19:59, 14 February 2007 (UTC)

I've fixed the link but made no other changes. Trezatium 20:16, 14 February 2007 (UTC)
"Most false negative results are due to the window period; other factors, such as post-exposure prophylaxis, can rarely produce false negatives."

According to the cited source (C. B. Hare et al.) this is NOT the case as long as the window period (which might or might not be extended by PEP) is taken into consideration. The production of antibodies was only impaired (e.g. false negative ELISA) when viral replication was continuously suppressed using ART for more than 24 weeks. In cases of PEP failure, viral loads would spike after the withdrawal of antiretrovirals leading to the formation of antibodies (e.g. positive ELISA). As the sentence is misleading and might make PEP users believe that testing after PEP is not conclusive, it should be deleted. — Preceding unsigned comment added by (talk) 15:57, 8 October 2011 (UTC)

Article name[edit]

HIV test follows the Wikipedia:Naming conventions (plurals) for an article about the many different tests used to detect HIV. This article is organised around the three main types of test; antibody, antigen and RNA and attempts to present the approved uses and limitations of each type of test in accordance with the neutral point of view policy.

The following links currently redirect to HIV test: AIDS test, AIDS testing, HIV testing, OraQuick and P24 antigen test

Sci guy 14:37, 20 Mar 2005 (UTC)

Merging talk page[edit]

As HIV testing has been merged with HIV test, can this talk page now be refactored?

Sci guy 03:57, 20 Mar 2005 (UTC)

As this page has been listed for peer review it is NOT a good candidate for merging. I suggest we fix the problems with this page first. Specifically this page could be used to agree on the purpose and content of an entry called HIV test (- User:


Wikipedia doesn't need separate articles on HIV test and HIV testing. One article will suffice for both noun and gerund. - Nunh-huh 14:31, 17 Mar 2005 (UTC)

HIV testing is a widely used term as part of a program to encourage people at risk of HIV to see if they are infected. It use different algorithms to blood screening and involves different issues such as confidentiality and counselling.

HIV test is a page listed for peer review. Currently it includes many details and tests of no relevance to HIV testing. Also HIV test has much out of date material.

Would you agree to solve the problems with HIV test with individual listings for each group of tests? Say Antibody, antigen and nucleic acid?(- User:

Exactly what problems do you refer to? In any case, this all should be discussed in a single article rather than fragmenting it all over the place. - Nunh-huh 14:59, 17 Mar 2005 (UTC)

Invalid template?[edit]

the following template is present in the source:

Conference reference | Author=C B Hare, B L Pappalardo, M P Busch, B Phelps, S S Alexander, C Ramstead, J A Levy, F M Hecht | Title=Negative HIV antibody test results among individuals treated with antiretroviral therapy (ART) during acute/early infection | Booktitle= The XV International AIDS Conference | Year=2004 | Pages=Abstract no. MoPeB3107

it gives the following output:

{{Conference reference | Author=C B Hare, B L Pappalardo, M P Busch, B Phelps, S S Alexander, C Ramstead, J A Levy, F M Hecht | Title=[ Negative HIV antibody test results among individuals treated with antiretroviral therapy (ART) during acute/early infection] | Booktitle= The XV International AIDS Conference | Year=2004 | Pages=Abstract no. MoPeB3107}}

anybody up to fix that? (clem 17:32, 8 Apr 2005 (UTC))

It now uses "cite conference". Guy Harris 17:31, 28 April 2006 (UTC)

Interpreting Test Results Section in Dispute[edit]

I think the following sentence from the "Interpreting Test Results" section is clearly biased and violates NPOV:

Such a faulty methodology has had extremely serious consequences, i.e. the world-wide use of HIV-antibody tests, Elisa and Western Blot, which dangerously lack specificity, as demonstrated in 1993 by Papadopulos et al. Papadopulos-Eleopulos, E., Turner, V. F. & Papadimitriou, J. M. (1993)

This methodology is only described as "faulty" by a miniscule number of people (i.e. AIDS dissidents/denialists). See for a discussion on how the Perth Group's (i.e. Papadopulos) conditions for virus isolation are deemed unnecessary by almost all virologists.

How about you await a response here before slapping a {{NPOV}} tag on the article? JFW | T@lk 02:33, 17 October 2005 (UTC)
Oh, I don't know...saying tests with specificity of >98% "dangerously" lack specificity is a bad enough misrepresentation of the facts as to merit an NPOV or a fix. I've done the latter, with a good recent (2005) reference replacing the denialist one from 1993. - Nunh-huh 03:46, 17 October 2005 (UTC)

Anonymous blanking[edit]

These edits didn't include any comments and only removed text. I agree that the article was (is) a bit wordy, but perhaps a shorter version of the removed text would be better than its wholesale removal. Please comment here regarding those edits. (Otherwise, I'll just revert them.) The Rod (☎ Smith) 18:11, 23 April 2006 (UTC)

Hello. I edited out statements that were clearly AIDS denialists' misrepresentation of facts. I am new to Wikipedia and am just learning my way around. If this article needs to include that AIDS denialist point of view, perhaps it could be confined to an "alternate theories" section. Let me know your thoughts. -- 03:18, 24 April 2006 (UTC)

Those were actually reasonable edits, but you made them without edit summaries, which (along with the fact you are using an IP address rather than a nickname) leads to suspicion. It would be good if you set up a name to edit under (not necessary, but it's in fact more anonymous than an IP, and helps to develop a metric of trust), and it's important to use edit summaries when removing information (e.g. "removing misleading argumentation about sensitivity") etc. Remember, people who know nothing about a subject may be checking such edits (to be sure they aren't vandalism) so it's good to explain it to them simply. In cases where it's necessary to remove a lot of material, it may even be a good idea to make a brief explanation (which can cover sever edits) on the talk page (i.e., here.) - Nunh-huh 05:20, 24 April 2006 (UTC)
Thanks, anon and Nunh-huh. (And agreed, I don't see anything compelling removed.) Thanks for the cleanup, anon. Looking forward to welcoming you formally if you register! Also, be prepared for denialist reverts. If that occurs, a short summary of denialist view near the end may become valuable, if for no other reason than to keep such POV contained. The Rod (☎ Smith) 07:33, 24 April 2006 (UTC)
I signed up for an user name is Robbieisfun. I made some small changes to the Interpreting Antibody Results section. The more I read this article, the more I come across denialist pseudoscience. This article may require an somewhat extensive rewrite, rather than just deleting things as I have done. --Robbieisfun 18:40, 24 April 2006 (UTC)
Yes, the article was originally written by a denialist trying to make a point. We tried to get the most misleading points out, but I think you've demonstrated there's more to do. - Nunh-huh 19:02, 24 April 2006 (UTC)

Language improperly says AIDS and HIV are synonymous[edit]

While the HIV article and sister articles are very informative, there are a few places where the terms HIV and AIDS are used interchangeably, which is misleading and ultimately incorrect. It is the virus that causes AIDS. Because of advances in HIV treatment, I believe even the term AIDS virus is misleading because it makes an automatic connection between HIV and AIDS, whereas theorhetically, due to the advances in medicine, a person can live with the HIV virus and not develop into AIDS (which, as you know, is a syndrome that requires many factors to be met, as determined by the CDC). Thanks. Bsheppard 03:04, 9 July 2006 (UTC)

Looks like a virus has never been found. They rely only in antibodies. AIDS also lacks a clear definition... — Preceding unsigned comment added by (talk) 03:43, 3 January 2012 (UTC)


Criticisms have been made by orthodox researchers. In fact, virtually all the Perth group's papers in their references are in the orthodox literature! So, apparently it's not just "dissidents" criticising the tests. 23:28, 15 July 2006 (UTC)

Hello Darin. (Isn't it strange how your user page says you have "retired", yet you keep on making anonymous edits?) Just because some of the Perth Group's papers have been published in (usually obscure) mainstream journals, that doesn't mean that the editors of those journals agree with them. Do you know of any non-dissident scientists who have explicitly criticised modern HIV antibody tests? If not then I think we should revert to my wording. Trezatium 19:02, 16 July 2006 (UTC)
I *HAVE* retired from making positive contributions to the Wikipedia. I feel a moral obligation, however, to respond to the misinformation and claptrap that continues to be peddled at a few particular articles here (specifically AIDS reappraisal and HIV test). Read the 1993 Perth paper. (Obviously you haven't. Or you're too just too flat out stupid to understand it.) BTW, their paper was published in Nature Bio/technology, which is one of the most prestigious journals, about as far from "obscure" as you can get!! Another piece of evidence that tells me you haven't read it or have no clue about the biology world. And its editor at the time was Harvey Bialy, who agrees with their major thrust of their criticisms of the WB test. They [Perth] make direct quotes of ORTHODOX researchers. Almost every reference and quote they make is to the ORTHODOX literature. A few examples:
"In half of the cases in which a subject had a positive p24 test, the subject later had a negative test without taking any medications that would be expected to affect p24 antigen levels...the test is clinically erratic and should be interpreted very cautiously.", Todak, G., Klein, E., Lange, M. et al. 1991. A clinical appraisal of the p24 Antigen test, p326. In: Vol. I, Abstracts VII International Conference on AIDS, Florence.
"On the basis of our positive Western Blot data, it appears that parenteral drug users may have been exposed to HTLV-III or a related virus as early as 1971. An alternative but equally viable explanation is that the HTLV-III seropositivity detected in these specimens represents false positive or non-specific reactions.", Jaffe, J.H., Moore, J.D., Cone, E.J. et al. 1986. HTLV-III Seropositivity in 1971-1972 Parenteral Drug Abusers-A case of false Positives or Evidence of Viral Exposure? NEJM 314:1387-1388.
"To culture is to disturb.", Meyerhans, A., Cheynier, R., Albert, J. et al. 1989. Temporal Fluctuations in HIV quasispecies in vivo are not reflected by sequential HIV isolations. Cell 58:901-910.
This is in addition to many indirect quotes and relaying of results. This all comes from the ORTHODOX literature itself. I would LOVE to see your retort to these people, as these are NOT "denialists". These people are supposedly on "your side". Other examples (not from Perth):
"Problems may be encountered when an HIV Western Blot is done on someone at no identifiable risk of infection. For example, recent studies of blood donors in whom no risk of HIV infection could be ascertained, who were nonreactive on the ELISA, and for whom all other tests for HIV were negative, revealed that 20% to 40% might have an indeterminate Western Blot...", and ""Notable causes of false positive reactions have been antibodies that sometimes occur in multiparous women and in multiply transfused patients. Likewise, antibodies to proteins of other viruses have been reported to cross react with HIV determinants. False positive HIV ELISA's also have been observed recently in persons who received vaccines for influenza and hepatitis B virus.", Proffitt MR & Yen Lieberman B (1993, June). Laboratory diagnosis of HIV infection. Infectious Disease Clinics of North America 7(2).; 203-215.
"Our results document a fourth source of false positive HIV-1 Western Blot results, which is the reproducible but nonspecific reactivity to (proteins from HIV)... Preliminary studies suggest that the basis for this cross reactivity with HIV-1 gp 41 proteins may be infection by paramyxoviruses, carbohydrate antibodies, or autoantibodies against cellular proteins.", # Sayre KR, Dodd RY, Tegtemeier G et al. (1996). False positive HIV-1 Western Bloy tests in noninfected blood donors. Transfusion 36; 45-52.
Montagnier's group concluded that gp 41 "may be due to contamination of the virus by cellular actin which was all the cell extracts", Barre-Sinoussi F, Chermann JC, Rey F, Montagnier L, et al. (1983). Isolation of a T-lymphotrtophic retrovirus from a patient at risk for AIDS. Science 220: 868-871.
"Circulating levels of plasma virus determined by (quantitative) PCR correlated with, but exceeded by an average of 60,000-fold, numbers of infectious HIV-1 that were determined by quantitative culture of identical portions of plasma... Total virions have been reported (in other studies) to exceed culturable infectious units by factors of 1000 to 10,000,000, ratios similar to those we observed in plasma.", Piatak M, Saag MS, Yang LC, et al. (1993). High levels of HIV-1 in plasma during all stages of infection determined by quantitative competitive PCR. Science 259; 1749-1754.
I could go on and on. But it's clear you haven't familiarised yourself with the orthodox literature, let alone "denialist" literature. 06:55, 18 July 2006 (UTC)

I've nothing against you making anonymous edits during your "retirement" - I just thought it was a bit odd, that's all. I agree that Nature Bio/technology is a very high impact journal. However, it seems not to be typical of the journals in which the Perth Group have published papers. Here's a list from their web site, along with the years of publication and recent ISI impact factors:

  • Bio/technology (1993)

Impact factor 22.4 (among the top 20 scientific journals in the world)

  • Current Medical Research and Opinion (1997, 1998, 1999)

Impact factor 2.945

  • Emergency Medicine (1993)

Impact factor 0.681

  • Genetica (1995, 1995)

Impact factor 2.085

  • Medical Hypotheses (1988, 1992, 2004)

Impact factor 0.725

  • Research in Immunology (1992)

Impact factor 1.321 (later merged to form "Microbes and Infection")

  • World Journal of Microbiology & Biotechnology (1995)

Impact factor 0.478

So eleven out of twelve articles appeared in journals with an impact factor below 3, and six of them were in journals with impact factors below 1.4. Perhaps the term "usually obscure" was a bit of an exaggeration, but still these are quite specialised journals, and none of them is dedicated to AIDS. I'd be much more impressed if they got something published in, for example, AIDS (impact factor 5.893) or JAIDS (3.681). Trezatium 19:40, 18 July 2006 (UTC)

Typical. I give you a string of quotes from actual articles in the literature to respond to, and you start yammering about impact factors. 09:03, 20 July 2006 (UTC)

My challenge was this: can you find any non-dissident scientist who has explicitly criticised modern (say post-1990) HIV antibody tests?

I gave you references. Apparently, you can't read. That's not my fault. 09:02, 20 July 2006 (UTC)

Has any non-dissident scientist expressed serious concern that current HIV testing protocols are unacceptably inaccurate and therefore need to be reappraised? As far as I can see, all of the papers cited by the Perth Group are limited to advising caution when interpreting the results of tests in unusual circumstances - the same kind of thing you'll find associated with all clinical tests.

"Advising caution??" If those statements aren't equivalent to "serious concern", there's no point in talking to you. I think anyone who reads the papers can see the implications. Again, you can live in your denial-fantasy land. It's amazing how you can just deny reality. The quotes (and literature) speak for themselves. 09:02, 20 July 2006 (UTC)

(Note that the p24 test detects antigen, not antibody, and is not generally used to make a diagnosis.

Hey! The p24 is one of the protein bands on the WB ANTIBODY test. Or do you even know how that test works?? Your ignorance is astounding. 09:02, 20 July 2006 (UTC)
You should immediately run to the phone and inform the LA County Coroner that "p24 antigen is not generally used to make a [HIV/AIDS] diagnosis." 09:11, 20 July 2006 (UTC)

Also note that an intermediate indeterminate (typo) WB test is definitely not the same thing as a positive WB.

Come on, they got up to 40% indeterminate on ELISA-neg serum. That's outrageous. Any other test with that rate of indeterminate result would not be the basis of life-or-death treatment decisions. Wake up. 09:02, 20 July 2006 (UTC)

And be careful to check whether the Perth Group are honestly representing the studies they cite - for example, they say, "Recipients of negative blood seroconvert and develop AIDS while the donors remain healthy and seronegative", yet the article they cite says, "On evaluation 8 to 20 months after transfusion ... All seven donors were found to be infected with HIV. On interview, six reported a risk factor for HIV infection, and five had engaged in high-risk activities or had had an illness suggestive of acute retroviral syndrome within the four months preceding their HIV-seronegative donation. Thus, these donors had apparently been infected only recently, and so were negative at the time of blood donation according to available antibody tests." In other words, the donors were tested during the window period between infection and seroconversion. The Perth Group representation of this study is a flat out lie.) Trezatium 20:46, 18 July 2006 (UTC)

Where do I start? The author's observations don't even make sense. First of all, if they had "illness suggestive of 'acute retroviral syndrome' (whatever that means) within the four months preceding their HIV-seronegative donation", then they would be in late-stage AIDS, well past the "window period". So there would be no way they could originally test negative. Unless you're suggesting they contracted HIV and developed AIDS-illnesses BEFORE seroconverting. That would actually be more in line with how retroviruses might work in practice!! The only conclusion is that their original 'acute retrovial syndrome' illnesses had nothing to do with HIV, or the tests were false-negative. But, "Antibody tests give false negative results during the window period of between three weeks and six months from the time of HIV infection until the immune system produces detectable amounts of antibodies. The vast majority of people have detectable antibodies after three months. A six month window is extremely rare with modern antibody testing." In other words, the only way that the "window period" would make sense is if all these activities supposedly leading to "HIV infection" had taken place within about a month or 2 before the original testing, in which case, why did the authors say 4 months?? If they really were "infected" 3 or 4 months before testing, they should have tested positive in the first place. And what the hell is an "illness suggestive of acute retroviral syndrome"? What on earth does that mean?? Almost everything in the quote you've given is either nonsensical or paradoxical. 09:02, 20 July 2006 (UTC)
I have to apologize for the slight of missing the word "acute". Late night blunder. Obviously, this is not the same as late-stage AIDS. That was an oversight. The point still stands regarding the 4-month period, though. If these (unnamed?) illnesses were indicative of recent infection, and they occured more than a month or 2 before infection, even up to 4 months before infection, then one would expect the "window period" to have passed and they shouldn't have tested negative. The "window period" of false-negatives and acute illness is usually (vast majority of time) within days, weeks or at most a couple months. 11:48, 20 July 2006 (UTC)
I have found the "study" you are apparently referring to. Not in the stacks, but at NEJM. It's listed under "correspondence", which basically means (a) it's not peer-reviewed, (b) there is no requirement for the authors to submit actual detailed data so we can examine whether their statements actually jive with their data. In the words of Richard Strohman, such correspondence is little more than "a forum for opening debate". If they really had something more striking to say about these results, I wonder why they chose the method of "correspondence"? In any case, although I haven't seen the letter itself, I'm not too sanguine that there will be sufficient actual data to know what on earth they mean by "illness suggestive of acute retroviral syndrome"...or what on earth they mean by "found to be infected with HIV". What you have quoted does not explicitly contradict what Perth said...I am not saying unequivocally that Perth were 100% accurate, I would need to read the entire "correspondence". However, the letter you quote as saying "HIV infected" while Perth says "seroconvert". The letter was in 1992, at a time when PCR was beginning to be used to justify "HIV infection". I wonder, they talk about "screened negative for antibody" with reference to the original "test results", yet later they use the phrase "found to be infected with HIV", they do not say "seroconverted", at least not in the portion of the quote you give, which would be the usual jargon if they had actually run additional antibody tests. Again, without the results published as a real study, not "correspondence", any of the authors personal conclusions are just non-peer-reviewed "heresay". The only thing that can prob be said for certain is that donors tested antibody negative originally, and the recipients later tested positive. Given the fact this was simply "correspondence", this is all I would have referred to if I were referencing the correspondence. 09:52, 20 July 2006 (UTC)

"there is broad scientific consensus that HIV is the cause of AIDS". I don't think that there is broad consensus. There are many papers, interviews, and videos that discount the simple equation presented here. There is much controversy that HIV alone causes AIDS symptoms. Why would you write this? Few evidence exists (broadly accepted) that HIV infection will result in AIDS symptoms. Dr. Luc Montagnier video interview for one.

Pure circular reasoning[edit]

"Almost all HIV-infected persons with indeterminate Western-Blot results will develop a positive result when tested in one month; persistently indeterminate results over a period of six months suggests the results are not due to HIV infection."

More to the point, if "persistent indeterminate results suggests [sic] the results are not due to HIV infection", why are they called indeterminate?? THAT'S WHAT INDETERMINATE MEANS!! "Indeterminate" means "we can't tell if it's due to HIV infection or not". You're basically saying "repeated indeterminate" = "negative", which is nonsense. There's no way you can take several indeterminate results, each of which individually are indeterminate, and claim altogether they mean it's negative. That's asinine. 10:15, 20 July 2006 (UTC)

It appears that the sentence in question may be based on this article, or something similar. A quick literature review using Pubmed (only articles with abstracts) found twelve studies supporting the idea that most people with indeterminate WB results are not infected (according to various types of assay), and that those with persistent indeterminate WB results are highly unlikely to be infected. One further study contradicts the first of these assertions, but not the second. For references see my talk page. Trezatium 23:48, 21 July 2006 (UTC)

What? these comments are horrible! The original quote makes perfect sense. It's not just that indeterminate tests are indeterminate, it's that in general they ARE NOT related to HIV. Furthermore, the fact that they will develop a positive result in one month is extremely relevant. Stop flipping out. —Preceding unsigned comment added by (talk) 17:01, 4 February 2010 (UTC)

The turmoils of Trezatium[edit]

Hey, Trezatium. YOU GOT THE WRONG PAPER. The actual paper is

  • Conley, L.J. and Holmerg, S.D. 1992. Transmission of AIDS from blood screened negative for antibody to the human immunodeficiency virus. NEJM 326:1499.


  • JW Ward, SD Holmberg, JR Allen, DL Cohn, SE Critchley, SH Kleinman, BA Lenes, O Ravenholt, JR Davis, MG Quinn, and et al. Transmission of human immunodeficiency virus (HIV) by blood transfusions screened as negative for HIV antibody. Volume 318:473-478 February 25, 1988 Number 8

Nice try. Either

(a) You're too stupid to find the article Perth ACTUALLY referenced,


(b) You're too naive to think I'd actually take the time to check to see if you wouldn't try to pull a fast one by me.

Take your pick. I don't care which is the case. Either way, it's YOU with egg on your face. Darin 14:32, 20 July 2006 (UTC)

Reply to Darin[edit]

I guess I'm just stupid. The articles have similar titles, appeared in the same journal and share an author, so I must have clicked on the wrong result in Google or Pubmed. I should have been more careful. I'm not interested in scoring points off you or anyone else, I'm just trying to ensure that the information in Wikipedia is accurate and unbiased, so I'm grateful to you for pointing out my mistake.

As you say, the Perth Group's reference was in fact a letter sent to NEJM. I agree that they would do better to cite a peer-reviewed article rather than a piece of correspondence. They could, for example, have referenced the article that I quoted. I wonder why they chose not to.

Anyway, the abstract for the letter in question can be found here. It appears to be discussing the same issue as the article I quoted, namely the risk of HIV transmission from blood taken from donors during the "window period", when viral load is very high but antibodies are not yet present. I haven't seen the full letter, but I very much doubt that it supports the assertion that, "Recipients of negative blood seroconvert and develop AIDS while the donors remain healthy and seronegative". Since you work at a university, perhaps you could look it up and report back?

I only cited this article as an example of how the Perth Group misrepresent other people's studies. And I think that my point still stands. If you follow up the references from this web page then you'll get an idea of how much data they are choosing to ignore.

To address your points in order:

  1. I listed the impact factors to answer your rebuttal of my "usually obscure" statement.
  2. I don't think that you've answered my challenge. Who but dissidents have explicitly called for the reappraisal of modern HIV antibody testing?
  3. To repeat, the p24 test detects antigen, while the WB test detects antibodies, including antibodies to the p24 antigen. Your mistake not mine.
  4. To repeat, the p24 test is not generally used to make a diagnosis (I never said never). Here is a reference.
  5. Indeterminate=indeterminate. Nobody is diagnosed positive until they get a positive result. Under current guidelines, they won't start treatment until they have also developed an illness or severe immunodeficiency.
  6. The CDC says, "Ninety seven percent will develop antibodies in the first 3 months following the time of their infection. In very rare cases, it can take up to 6 months to develop antibodies to HIV." The purpose of the NEJM article was to discuss seven such "very rare cases". That's why there were only seven of them, in a country where around 40,000 people become infected with HIV each year. Another NEJM article estimates that the risk of HIV transmission from screened blood was 1 in 450,000-660,000 in 1995, though it was previously higher when tests were less sensitive.

Trezatium 19:48, 20 July 2006 (UTC)

Also note that the authors said "within the four months", so in some cases the period may have been just a few weeks or even days. How about checking the full text? Trezatium 19:59, 20 July 2006 (UTC)

Since no one has presented evidence that non-dissident scientists have criticised modern HIV antibody tests, I've changed back the first sentence of the Criticism section. According to the most extensive studies, the tests are more than 98% accurate. The papers presented here by either investigate what happens in the other less than 2% of cases, or look at unusual situations in which accuracy may be lower. This does not constitute criticism. As far as I am aware, only dissidents are calling for the tests to be reappraised. Trezatium 18:27, 7 August 2006 (UTC)

The p24 test[edit]

The article states that, "The p24 antigen test is not useful for general diagnostics." I think this is slightly misleading, since the p24 test can be useful for testing people with very recent exposure (after a window period of a few days, and within about three weeks of exposure), and for testing newborn babies. The article also says that the p24 test "is no longer used routinely in the US or the EU to screen blood donations." Although the US reference cited recommends switching to nucleic acid testing, it doesn't say that p24 testing has ceased. The Eurpopean reference only discusses a collection of Western European countries, not the whole EU. It might be worth doing a bit more research on this topic. Trezatium 19:38, 1 September 2006 (UTC)

- New User

I agree here. What is also known as the PCR test is commonly used in Africa to determine recent exposure. The fact that other nations refuse to make this test, with a 2 - 10 day window period on average, available to the public shocks me.

A due note is that each PCR test has to be tailored for each strand of HIV. An African test is different to the common European / Americas strand and minor strands have developed over time.

Although the ELIZA and WESTERN BLOT tests test for all HIV antibodies, PCR tests for viral load of a specific strain only. In the past these were labeled PCR I and PCR II, now there are several more divisions.

Knowing the strand of HIV from the party infecting the tested party allows quick testing for viral load with PCR and whether there was a contracted infection. —Preceding unsigned comment added by Cecilpickardbrown (talkcontribs) 18:54, 29 December 2007 (UTC)


I've made mostly minor updates to the style, writing, and organization, as well as adding some citations and a few more significant changes. I'm not clear offhand on the status of p24 for diagnosis but can look into this. As far as User:, quite a few anonymous IP's from Albequerque NM are used to make edits from an AIDS-denialist POV. These edits and talk page comments all share a common tone (generally abusive, lots of personal attacks, occasionally threatening to vandalize a page when s/he doesn't get his/her way, etc); either there are quite a few uncouth AIDS denialists in Albequerque, or these are all the work of the same, anonymous editor ("Darin"). The latter seems more likely. MastCell 02:00, 8 October 2006 (UTC)

Darin is Darin Brown, creator of the dissident "AIDS Wiki". Trezatium 18:18, 9 October 2006 (UTC)
Ah so. MastCell 18:26, 9 October 2006 (UTC)

Sourcing For Test Accuracy[edit]

The sources for the accuracy of HIV tests are more than a decade old. How reliable are the rates statad in this wiki (.003% and .0006%)? Is this information current? —The preceding unsigned comment was added by DiggyG (talkcontribs) .

Actually, the most current evaluation is probably that quoted in the article from the United States Preventive Task Force review, published 2005 in Annals of Internal Medicine. The article is available online here. They found the chance of a false-positive in the general population as 1 in 250,000 (0.0004%). The USPTF in turn cited a 1998 study by Kleinman et al. (JAMA. 1998;280:1080-5 PMID 9757856) as their source for the false-positive number. So the numbers from this Wikipedia article are up to date and accurate. Since testing methodologies haven't changed much (ELISA plus WB), and the original studies were robust, they probably won't be updated too often, but the Kleinman (1998) study suggests that things haven't changed much in terms of accuracy from 1989-1998. MastCell 19:46, 20 October 2006 (UTC)
Should it be mentioned that the accuracy of the "HIV test" (since there is no gold standard for HIV) was obtained by repeating the test and looking for the same result? --Loundry 19:13, 11 December 2006 (UTC)
If you look at the above USPTF article as well as the sources cited in the Wikipedia article (PMID 2046708, PMID 2648922, among others), you'll see that the ELISA/WB strategy was in fact verified by culture of HIV. Culture of the pathogen in question is a widely accepted gold standard in microbiology. So I think inserting your suggestion would be inappropriate. MastCell 01:16, 14 December 2006 (UTC)

The article says: "With confirmatory Western blot, the chance of a false-positive identification in a low-prevalence setting is about 1 in 250 000". This overoptimistic estimation is based on Bayesian probability theory assuming strict independency between the variables tested by Western Blot and those tested by EIA. The assumption of independency is wrong:

"...Care must be taken, however, when interpreting the

results from a sequence of tests. Assays are generally not strictly independent, since one source of bias may simultaneously affect multiple laboratory techniques. For example, EIA, W B and IFA are all techniques that detect antibodies to HIV, as opposed to techniques that detect viral antigens of viral RNA directly (see below). Pure Bayesian analysis, which assumes strictly independent tests, will typically therefore lead to overestimation of predictive values with most supplementary tests (see Table 8.1):"

Source: 2004, Gary P. Wormser, AIDS and other manifestations of HIV infection, 4th edition, Elsevier Academic, ISBN 0127640517. , Chapter 8, page 156, "Persons at Low Risk"

The authors issue a warning against the current medical malpractice of diagnosing "infection" solely on a serological test while ignoring the absence of a clinical picture in the patient:

"...The context within which any test is used

is of critical importance to its interpretation. No test, per se, should be the basis for diagnosis on its own, but rather a test is merely an aid in correct diagnosis. The practitioner must use test results in the context of a clinical picture to reach an accurate diagnosis."

Relative to the Positive Predictive Value of the usual testing sequence (2 x ELISA + confirmation Western blot) Gerd Gigerenzer, Ulrich Hoff rage, and Axel Ebert in page 4 of "AIDS Counselling for Low-Risk Clients" make the following remark:

"...What is the predictive value of a positive test for a 20- to 30-year-old heterosexual German man who does not engage in risky behaviour? Inserting the previous estimates — a prevalence of 0.01%, a sensitivity of 99.8%, and a specificity of 99.99% (repeated ELISA and

Western blot) — into Bayes’ rule, the PPV results in 0.50, or 50%. An estimated PPV of about 50% for heterosexual men who do not engage in risky behaviour is consistent with the report of the Enquete Committee of the German Bundestag, which estimated the PPV for low-risk people as “less than 50%” (Deutscher Bundestag, 1990, p. 121)."

The USPTF quote in the article claims, referring to a 1998 source: "the chance of a false-positive identification in a low-prevalence setting is about 1 in 250 000 (95% CI, 1 in 173 000 to 1 in 379 000)."

But the source says literally:

"Of 421 donors who were positive for HIV-1 by Western blot, 39 (9.3%) met the criteria of possible false positivity because they lacked reactivity to p31. Of these, 20 (51.3%) were proven by PCR not to be infected with HIV-1. The false-positive prevalence was 4.8% of Western blot–positive donors and 0.0004% (1 in 251000) of all donors (95% confidence interval, 1 in 173000 to 1 in 379000 donors)."

So if you're low risk (a donor) and come out HIV-positive by ELISA + WB, chances of being misdiagnosed is 4.8%. So the PPV value of ELISA + WB is 95.2%. Instead, the USPTF gives absolute figures, creating the false impression that a positive diagnosis by ELISA + WB is extremely accurate. But it's a whopping 5% wrong!-- (talk) 16:33, 29 November 2012 (UTC)

Positive / negative predictive value[edit]

I am somewhat surprised that neither of the above is mentioned. Many people confuse the two (which they are actually interested in) with sensitivity and specificity. Even though he latter are correctly described it is not explained that the two are not the same as predictive values. Without knowledge and understanding of this difference, the section on accuracy of tests cannot be readily understood. Also I get the feeling the authors of "Screening for HIV: A Review of the Evidence for the U.S. Preventive Services Task Force" are not too sure on that difference either. They mention a false-positive rate of 1 in 251.000 in a "low-prevalance setting" where a false-positive rate is only really important for a predictive value. The source of these data reported that of 421 Western-blot-positive donors, 20 were found to be negative by RT-PCR, which means that roughly 5% of these 421 received a false-positive result. The individual meaning of a positive test cannot be interpreted without knowledge of the baseline risk. --Docvalium 13:04, 12 November 2006 (UTC)

Okay, let me hold your hand and walk you through it.
I'm going to assume 0.6% of the population is HIV+, that the canonical false positive rate of 0.0004% is accurate, and the canonical false negative rate of 0.003% is accurate. If you wish to suggest a conspiracy that is suppressing the "real data" then by all means do so.
600000 per hundred million people in the world have HIV. Let's divide these groups. On the one side we have 600000 people who are HIV+, and on the other side we have 99400000 who are HIV-. A false negative rate of 0.003% is a test sensitivity of 99.997%. So of those 600000 people who are HIV+, 599982 will test positive. And a false positive rate of 0.0004% means that of those 99400000 people who are HIV-, 398 of them will test positive.
This gives us a PPV of 99.93%. Happy now? -- (talk) 13:25, 15 February 2008 (UTC)
No conspiracy needed to prove that actual canonical rates are nothing near the wishful figures you suggest. Western Blot is touted as the most specific antibody test. Let's take a close look at a model that's widely use in confirmation testing: Genelabs/MP HIV BLOT 2.2 Western Blot assay
No figures on specificity are given, only the test outcomes of a total of 258 subjects (Table 2 on page 7): 208 blood donors and 50 people with various non-HIV viral conditions. No positives reported. It's like hey! you do the math!
Because the number of subjects tested was insufficient to produce some false positives, the specificity of this test is unknown. Optimistically, it should be around 99.5% (1/208) for healthy subjects and 98.2% (1/58) for subjects with a non-HIV viral condition. Using this test means happily extrapolating the results of these statistically insignificant 258 subjects onto the population at large.
Applying these figures to the general UK population, 99.5% specificity (for the healthy) means 5 positives per 1,000 people tested. 98.2% specificity (for the ill) means 18 positives per 1,000 people tested. The prevalence of HIV in the UK is (estimated) 1.5 per 1,000 (see page 6 of HIV in the United Kingdom: 2011 Report). If your test comes out positive and you're healthy, the PPV is just 30% (1.5/5). If you're sick it's a lousy 8% (1.5/18). That's asuming the prevalence is real, but being lower than the specificity of the test means the prevalence itself may be just a testing artefact (testing noise higher than the epidemic signal).
To make bad matters worse, the test still requires an interpretation. One would expect an "extremely accurate" test to have an extremely consistent interpretation, right? But alas! On page 6 the manufacturer warns that "..guidelines for interpretation may differ depending on the local policies.". Now the scientific consensus is impressive, is it? "Local regulations" is a good laugh. Laisser faire would seem a more apt description. Next, it gives a table with 7! different interpretations in use of what a positive result should be... 7! If this chaos wasn't bad enough, the manufacturer adds yet one interpretation criteria, its own (also in page 6)... that makes a total of 8!
Insufficient number of subjects testsed + 8 different interpretations of positiviy = extremely accurate test??? Seriously, whoever uncritically accepts the pitch about the extreme accuracy of HIV antibody tests needs a reality check. — Preceding unsigned comment added by (talk) 01:04, 28 November 2012 (UTC)

Criticisms of HIV Tests is POV[edit]

The following section:

HIV tests have been criticized by a number of so-called "AIDS dissidents" (people who reject the scientific consensus that HIV causes AIDS). For example, Eleni Papadopulos-Eleopulos and a group of AIDS dissidents wrote an article in 1993 entitled "Is a Western Blot Proof of HIV Infection?"[17] Their arguments rest on issues of specificity, standardisation, reproducibility, and validation.

However, the accuracy of serologic testing has in fact been verified by isolation and culture of HIV and by detection of HIV RNA by PCR; these are widely accepted "gold standards" in microbiology.[18][19] While the AIDS dissidents focused on individual components of HIV testing, the combination of ELISA and Western Blot used for the diagnosis of HIV is in reality remarkably accurate, with very low false-positive and -negative rates as described above. The vast majority of scientists believe that the view of AIDS dissidents are based on highly selective analysis of mostly outdated scientific papers; there is broad scientific consensus that HIV is the cause of AIDS.[20][21][22]

Under "Criticisms of HIV Tests" violates WP:NPOV; specifically: "None of the views should be given undue weight or asserted as being the truth, and all significant published points of view are to be presented, not just the most popular one."

I suggest removing the scare quotes around "AIDS dissidents", the snide "so-called", the spin "in reality", and by citing the Perth Group by name as well as providing a link to the writings of their organization.

"Isloation" of HIV has never been demonstrated (unless the definition of "isloation" has changed, as is implied by its usage by some scientists), so the "verified by isolation and culture of HIV" is misleading.

I'll make these changes in a few days if there are no objections. I'm still new to Wikipedia at this point. --Loundry 19:29, 11 December 2006 (UTC)

The policy goes on to give "an important qualification", which is that, "Articles that compare views need not give minority views as much or as detailed a description as more popular views, and may not include tiny-minority views at all (by example, the article on the Earth only very briefly refers to the Flat Earth theory, a view of a distinct minority). We should not attempt to represent a dispute as if a view held by a small minority deserved as much attention as a majority view, and views that are held by a tiny minority should not be represented except in articles devoted to those views." The Perth Group are a tiny minority. They have no relevant qualifications or practical experience in HIV testing. Virtually all scientists believe that their views are crazy. Therefore it would be not unreasonable to omit mentioning them altogether in this article. I personally think that the section should remain, but I also think that minority views should be presented as such, as per the policy.
Having said that, I agree that the section was biased, and I've edited it accordingly. Trezatium 20:11, 11 December 2006 (UTC)
Thank you for the edit, it looks much better now. That said: 1) If the Perth Group is a "tiny minority" and thus their information can be omitted, then why can't that same rationale be used to omit ALL information regarding AIDS reappraisal? Certianly AIDS dissidents are also a "tiny minority"? 2) "Virtually all scientists believe that their views are crazy" is an appeal to authority. 3) It is not the "views" of mainstream / dissedents that concerns me, but the evidence (or lack thereof). --Loundry 20:45, 11 December 2006 (UTC)
This isn't about the AIDS reappraisal movement as a whole. As far as I'm aware, Duesberg and other reasonably well-qualified scientists haven't alleged that HIV tests are grossly inaccurate, or that HIV hasn't been properly isolated. Apart from the Perth Group, few others have pursued these lines of argument. Trezatium 20:54, 11 December 2006 (UTC)
To be honest, I haven't heard any AIDS dissident claim that the "HIV tests" are "grossly inaccurate". Rather, what I've read is that they claim that the accuracy of the "HIV tests" is impossible to verify (because of the admitted lack of a gold standard) and that, therefore, the "HIV status" is meaningless. For an AIDS dissident to claim that the "HIV tests" are grossly inaccurate would be for them to imply that there is such a thing as an accurate "HIV test", and that implies that HIV exists and is the cause of AIDS, and those notions are the very ones most frequently in dispute. That said, I still don't understand why you can't apply the "tiny minority" rationale of omission to the whole lot of AIDS dissidents. Do you think that they are large enough to merit inclusion whereas the Perth group is so small that they should rightly be omitted? If so, then how do you gague the size/percentage of each? I'm trying to understand your policy of applying the "tiny minority" rationale of omission vis-a-vis the ideas of AIDS dissidents because it seems that you apply it arbitrarily. --Loundry 21:24, 11 December 2006 (UTC)
The Wikipedia policy talks about "tiny-minority views" but doesn't define exactly what it means. Nevertheless, the verification argument is, as far as I'm aware, only made by the Perth Group, who surely meet any definition of "tiny minority" as they number fewer than a dozen. The entire AIDS reappraisal movement is definitely a minority, but probably not a "tiny minority". Trezatium 21:30, 11 December 2006 (UTC)
You have given a subjective rationale for the omission of information. --Loundry 21:58, 11 December 2006 (UTC)
I'm just trying to follow the Wikipedia policy, which forces us to be subjective about what constitutes a "tiny minority". Trezatium 22:07, 11 December 2006 (UTC)
Another point. When you look at references 18 and 19 attached to the claim: "testing has in fact been verified by isolation and culture of HIV and by detection of HIV RNA by PCR; these are widely accepted "gold standards" in microbiology.[18][19]" you find out that the term "isolation" acquires a new meaning, namely the application of immunocytochemical tecniques to a co-culture in order to detect proteins p18 and p24 which are deemed exclusive of HIV. This detection is then used as indirect "proof" of the presence of HIV. Therefore there's no HIV isolation but another surrogate test! In [19] they go even further and re-test the culture having p18 and p24 using PCR. In the end the tests (antigen, antibodies and PCR) are being verified against a surrogate "gold standard" that is looking for the same thing. There's no independence between the "gold standard" and the test, so any correlation cannot be attributed to "specificity" in the tests, but to the fact that the gold standard is not independent. Without isolation it's impossible to know what the accuracy of the tests is. — Preceding unsigned comment added by (talk) 14:25, 13 March 2012 (UTC)
While that's an interesting viewpoint, it's not one that's shared by experts or people knowledgeable in the field. Since Wikipedia is intended to be a serious, respectable reference work, we tend to try to convey expert scholarly opinion in a field rather than idiosyncratic editorial beliefs. MastCell Talk 15:59, 14 March 2012 (UTC)
It depends of who you prefer to listen to. I've already cited experts knowledgable in the field that call the lack of indepenency between EIA and Western Blot and therefore challenge the validity of the second as a "confirmation" test. When others use non-independent "gold standards" of the same surrogate nature as the tests they pretend to validate, the same objection applies. It's so obvious that's a sore to the eye. — Preceding unsigned comment added by (talk) 17:08, 14 March 2012 (UTC)
Relating to your contempt towards "old papers" having anything to do with "modern" tests, let me give you an example to illustrate the stagnation of HIV testing science. The following is the bibliography referenced by the 2009 edition of the "Core HIVI & 2 Rapid test for the detection of Antibodies to Human Immunodeficiency Virus (HIV) in Serum, Plasma and Whole Blood" (see

1. Popovic, M., Detection Isolation and continuous production of Cytopathic Retroviruses (HTLV-lll) from patientswithAlDS and pre-AIDS. Science 1984; 224:497.
2. Carlson, J. R., AIDS serologytesting in low and high riskgroups. JAMA 1985; 253:3405.
3. Centers for Disease Control, Update on Acquired Immune Deficiency Syndrome (AIDS) MMWR 1982; 31:507.
4. Gallo, R. C., et. al. Frequent detection and isolation of Cytopathic Retroviruses (HTLV-lll) from patients with AlDS and a risk for AlDS. Science. 1984; 224:500."

I find it worrying that the most recent results test manufacturers were relying upon in 2009 were from papers published in 1985!
I see no citation to any reliable source for medical claims that says anything wrong with the current testing regimen. This talk page is for the improvement the article based on reliable sources, not your own original research on the topic or your feelings about them. Yobol (talk) 17:26, 14 March 2012 (UTC)
2004, Gary P. Wormser, AIDS and other manifestations of HIV infection, 4th edition, Elsevier Academic, ISBN 0127640517. , Chapter 8, page 156, "Persons at Low Risk"

"...Care must be taken, however, when interpreting the results from a sequence of tests. Assays are generally not strictly independent, since one source of bias may simultaneously affect multiple laboratory techniques. For example, EIA, W B and IFA are all techniques that detect antibodies to HIV, as opposed to techniques that detect viral antigens of viral RNA directly (see below). Pure Bayesian analysis, which assumes strictly independent tests, will typically therefore lead to overestimation of predictive values with most supplementary tests (see Table 8.1):"

Source: Indian Journal of Medical Sciences, 2005,Volume 59, Issue 10, Pages 443-450; Iqbal H Syed1 et al. "HIV-1 western blot assay: What determines an indeterminate status?" (

"...Different regulatory bodies have proposed criteria for the interpretation of the band profiles. The differences in the sensitivity and specificity associated with each criteria may lead to differences in results for the same specimens,[1],[5],[15] and has sometimes even resulted in false positive reports.[16] Herein lies the importance in of the interpretation criteria for accurate diagnosis based on the WB band profiles.[15] ...Of the 23 specimens that were run on both the LAV BLOT I and the genetic systems kit, we found a lot of discrepancies in the results. Based on the criteria suggested by the manufacturer the LAV BLOT I kit identified 3 specimens as positive, 4 specimens as negative and 16 specimens as indeterminate. However, the genetic systems identified only 2 specimens as positive, 7 specimens as indeterminate but 14 of these 23 specimens as negative. This implies that the LAV BLOT I could be wrongly reporting 10 negative HIV-1 western blot assay specimens as indeterminate. On further analysis 1 out of this 10 specimens could be read positive if CDC & CRSS criteria were used. This again proves the variations of results with respect to different criteria."

What do you propose to change the wording in our article? (As has been pointed out, this is not a WP:FORUM for general discussion of the topic.) I do not see anything in the cited material above which would require us to change what is currently worded in the article, or frankly, which supports the original assertion made that started this thread. Please suggest specific wording or changes here and the specific citation which you would use to support those changes. Yobol (talk) 18:09, 14 March 2012 (UTC)
I would suggest to change the following wordings:
Where it says: "The combination of these two methods is highly accurate", I would add: ", however, because both Western Blot and ELISA detect antibodies to HIV, one source of bias may simultaneously affect the other. This lack of statistical independence leads to overestimation of the predictive value of this testing sequence when estimated by Bayesian theory, which assumes independence." Source: 2004, Gary P. Wormser, AIDS and other manifestations of HIV infection, 4th edition, Elsevier Academic, ISBN 0127640517. , Chapter 8, page 156, "Persons at Low Risk"
Where it says: "these tests are inexpensive and extremely accurate.", the misconception is that accuracy (Positive and Negative Predictive Values) depends on an external variable, the prevalence of the infection among the population tested. Tests by themsleves only have specificity and sensitivity ratings. Base on reference Decision 2002/364/EC on common technical specifications for in vitro-diagnostic medicaldevices, pages L 39/40 to L 39/42), I would ammend: "The tests are rated for sensitiviy and specificity using test panels with sera from subjects classified by clinical criteria (AIDS-defining diseases, risk levels) and other antibody tests as indirect markers of actual viral presence. In spite of calibration against indirect markers of HIV infection, they are considered to have good sensitivity and specificity."
Where it says: "Rare false positive results...", it begs the question: false compared to what standard of true positivity? Hard to find the answer, and since the final sentence is delivered after Western Blot confirmation, which has its own non-negligible rate of false positives, I'm afraid the answer might be "there's none". — Preceding unsigned comment added by (talk) 02:07, 28 November 2012 (UTC)

ELISA, two ELISA, ELISA + WB, two ELISA + WB[edit]

The article does not mention that different countries use different combinations of antibody tests to determine HIV status and is thus misleading about how HIV status is diagnosed. If no one sees fit to add this section then I will do so. --Loundry 01:23, 12 December 2006 (UTC)

I don't know about "misleading" (the article clearly states "In the U.S..."), but it is US-centric, which is a shortcoming. If you'd like to add information about how tests are utilized and interpreted elsewhere in the world, that would be useful, provided such information meets the Wikipedia guidelines for reliable sourcing, neutral point-of-view, and verifiability. MastCell 19:36, 14 December 2006 (UTC)
It is misleading because "HIV positive" in the USA is assumed to be the same thing as "HIV positive" in another country. If, in another country, such a diagnosis is reached by two positive-result ELISAs, whereas in the USA such results would need to be "confirmed" by a positive-result WB (which might be a negative result and thus lead to a status of "HIV negative") then the only conclusion that can be drawn is that not every "HIV positive" status is equally meaningful (or meaningful at all). Hence, I think this article is misleading (by omission, not by commission) about how a "HIV positive" status is diagnosed. That said, I am having a hard time coming up with the locale-by-locale lists of how an "HIV positive" status is reached. I am aware of the CDC guidelines and the Bangui definition (no testing required if you're in Africa), and I have also read that Scotland does not use WB. I will add to this article once I can get some definitive data. --Loundry 16:02, 19 December 2006 (UTC)
You don't have to go elswhere. Within the USA there are different criteria for a positive Western Blot. The CDC criterium is ((gp160 OR gp120) AND gp41) OR (p24 AND (gp160 OR gp120 OR gp41)). It is lax because it doesn't requiere any POL components. Instead, the American Red Cross (ARC) is more strict and requires at least one component of all three genes ENV, GAG and POL. This difference is very significant because in low risk-settings the probability of a Western Blot coming out positive varies from 1 in 1300 for the CDC criterium to 1 in 8900 for the ARC. This is readily verifiable using the band frequency probabilities in low-risk settings documented in and , then adding up the joint probabilities of each positive band combination for each criterium (assuming independency of the bands). Depending on who interprets your test you have a 10-fold variation in your chaces to be "infected". This is anything but a scientific testing procedure. — Preceding unsigned comment added by (talk) 14:54, 13 March 2012 (UTC)
I'm not sure that your personal interpretation of 2 20-year-old papers is an appropriate basis to change the article content; see Wikipedia:No original research. The modern clinical testing algorithm is well-studied and its accuracy is cited in this article with reference to reliable sources. MastCell Talk 15:53, 14 March 2012 (UTC)
The prevalence of the different Western Blot bands among low-risk population 20 years ago and today shouldn't have changed substantially. The likelihood of a "positive" conbination appearing by chance given an interpretation criterium are computable. It's straighforward to do the math, and the likelihood is of the same order as the observed "prevalence of hiv infection", which is suspicious. The band pattern interpretation criteria today for the CDC and the ARC are as I said, plus other criteria being used for example in Australia. Therefore the "modern" clinical testing algorithm still depends on variable, subjective criteria which largely influence the rate of positives. These tests were initially conceived for the purpose of screening blood - it doesn't hurt anyone to throw suspicious blood away - not for diagnosing asyntomatic people and write them a death sentence and chemo for life. — Preceding unsigned comment added by (talk) 17:02, 14 March 2012 (UTC)
There are lots of reports (complains) in the literature of Western Blot interpretation standards being fuzzy, shifting and and inaccurate:
"...Using CDC and WHO criteria, 6 (19.4%) and 2 (6.5%), respectively, of these WB assays would have been considered falsely positive. In addition, 17 indeterminate samples were negative when assessed by a nucleic acid-based amplification assay for HIV-1 viremia. In general, there was 97.8% concordance between the ARC and WHO criteria and 85.7% concordance between the ARC and CDC criteria for an indeterminate WB result. The ARC criteria best met the specified objectives for diagnosis in our setting..."
"...We observed that criteria laid down by ASTPHLD, CDC and CRSS scored all the specimens as positive while ARC and WHO criteria scored 13 (2.8%) and 18 (3.8%) of specimens, respectively, as negatives which were detected as positives by other criteria (discordant specimens)...CONCLUSIONS: It is felt that before selecting any criterion for Western blot positivity, it should be evaluated in the local population at risk for HIV-1 infection with additional or follow-up tests..."
"...Our diagnostic criteria gave far better results in our population than the existing published criteria. This suggests that new criteria could be developed to improve WB interpretation in African settings..."
2007 - English Public Health Laboratory Service discontinued the use of Western Blot.
"...Yet this case did publicly raise the issue of how HIV is identified in UK laboratories. It pointed to dissimilarities in global diagnosis of HIV. For example, in the USA, where physicians always diagnose HIV infection using confirmatory Western blots - an HIV-test not supported by our Health Protection Agency (12) - there are differing definitions for a 'positive' Western blot (13)..."
1991 - Here's why the English Public Health Laboratory Service discontinued the use of Western Blot.
"..Western Blot detection of HIV antibodies began as, and should have remaines, a research tool.."
"...Different regulatory bodies have proposed criteria for the interpretation of the band profiles. The differences in the sensitivity and specificity associated with each criteria may lead to differences in results for the same specimens, [1,5,15] and has sometimes even resulted in false positive reports.[16] Herein lies the importance in of the interpretation criteria for accurate diagnosis based on the WB band profiles.[15]...On further analysis 1 out of this 10 specimens could be read positive if CDC & CRSS criteria were used. This again proves the variations of results with respect to different criteria."
"neither the WHO nor the CDC takes into account the reactivity to the pol gene product in defining the patient’s infection status....since the adoption of the revised CDC criteria, which dropped the requirement for p31 reactivity, an increase of false-positive HIV-1 test results has been reported [6]..." — Preceding unsigned comment added by (talk) 03:54, 28 November 2012 (UTC)

First Paragraph[edit]

I'm about to change the wording of the last sentence, but I'm wondering if it should be there at all. I mean limmiting an international thing down to one country is kind of strange, not to mention the fact that we should be using United States (with the link) since it's the first time we mentioned it. But what about Britan? What about India? What about _____? I would support putting that in a section of some kind, but not the first paragraph.Daniel()Folsom T|C|U 22:17, 16 February 2007 (UTC)

My bad - the U.S/United States thing is fine ... I missed the second sentence. But now I think both those sentences should be taken out - an intro is supposed to give an overview - you don't go into specifics in an overview. Barring objections I'll delete them by tommorow.Daniel()Folsom T|C|U 06:02, 17 February 2007 (UTC)

Opening Sentence, current: "HIV tests are used to detect the presence of the human immunodeficiency virus (HIV), the virus that causes acquired immunodeficiency syndrome (AIDS), in serum, saliva, or urine. Such tests may detect antibodies, antigens, or RNA." This is an oxymoron between the "detect the presence of" and "such tests may detect antibodies, antigens, or RNA." Why? They aren't testing the presence of the virus, but rather the presence of the antibodies/antigens/RNA. Well, in my book, it isn't actively seeking out nor isolating any HIV, and thus, it's not detecting HIV at all. If someone has the wrong cold or flu during an hiv test, they might be producing enough of the antigens/antibodies to show a positive result without even harboring hiv in their bodies. The body produces same antibodies for a variety of different viruses, which could mean false positives for some cold/flu carriers, etc. The point is, this opening sentence needs to change "detect" to "determine." After fixing, it should read as follows: HIV tests are used to determine the presence of the human immunodeficiency virus (HIV), the virus that causes acquired immunodeficiency syndrome (AIDS), in serum, saliva, or urine. Such tests may detect antibodies, antigens, or RNA. — Preceding unsigned comment added by (talk) 15:14, 27 June 2011 (UTC)

Why not use the manufacturer's disclaimer directly? Alere Determine™ HIV-1/2 Ag/Ab Combo. It explains: "The test is intended as an aid to detect HIV-1 p24 antigen and antibodies to HIV-1/HIV-2 from infected individuals.". It is not even a detection of antibodies, but an "aid" in the detection. This implies additional procedures just for the antibodies. No claims to the virus are made. -- (talk) 15:03, 29 November 2012 (UTC)


Saying something is accurate might be crossing the npov line, and frankly what bad does adding "considered" do. I mean according to the sources, they are considered to be very accurate. But it's a theory that they are accurate - not a fact. I mean take the entire population that says "HIV doesn't exist" - by saying they are accurate, we're saying HIV does exist, which is NPOV - I'm not sure what the issue is here - it's a no-brainer.Daniel()Folsom T|C|U 07:05, 17 February 2007 (UTC)

Just so everyone knows, I'm moving that section of the text as soon as this conversation gets through with - it has no place in the first ip as mentioned above.Daniel()Folsom T|C|U 07:07, 17 February 2007 (UTC)
HIV does exist. - Nunh-huh 07:11, 17 February 2007 (UTC)
You might believe that - and frankly so do I, but some people don't. Therefore you saying it is NPOV - and if that's your only argument I'm reverting it back - we must follow Wikipedia's guidelines.Daniel()Folsom T|C|U 07:16, 17 February 2007 (UTC)
NPOV doesn't mean unanimous consent. It means views are attributed and represented proportionally. By adding considered you are giving the HIV denialist view undue weight. That some people refuse to accept well-established fact is irrelevant. We reference their dissent, but that dissent need not color our description of facts. Thus our article on the Earth notes that some people believed in a flat earth, but that doesn't color its description of the actual facts. - Nunh-huh 07:21, 17 February 2007 (UTC)
For you to compare this to the flat earth theory is rediculous. I mean I'm trying to think of a metephor but frankly it's so unbelievably strange that I can't. It's such an exageration. As far as undue weight goes - the majority does not define wp's opinion. And it's not just people who don't think HIV exist - there are plenty of articles suggesting that HIV test may not pick up the disease in the early stages. In fact - if I can find enough articles for either side, I'm going to say the accuracy has been disputed and then have both links - unless of course again you come up with another argument. Also - I assume you agree with moving the last 2 sentences out of the first IP (as stated above)Daniel()Folsom T|C|U 07:27, 17 February 2007 (UTC)
It's actually rather a good analogy. To say "accuracy has been disputed" as though both sides of that "dispute" were equally well supported would be misleading and inappropriate. - Nunh-huh 07:48, 17 February 2007 (UTC)
HOLY CRAP - [1][2] I found the motherload of artilces - this first one says that the HIV test is accurate in testing for antibodies which may or may not be HIV. Given that it's called the "HIV Test" - then accuracy must depend on how many times HIV shows up - which obviuosly varies. this second (in a q+a format) in the first answer at one point says it could take a few weeks until the antibodies even show up.Daniel()Folsom T|C|U 07:35, 17 February 2007 (UTC)
Flat earthers are not geologists. In contrast, voices critical to the HIV/AIDS theory, including doubts on the proper demonstration and characterization of HIV, rise from within the scientific community. There are virologists (Stefan Lanka, Duesberg), biochemists (Nobel prize Kary Mullis^, Perth Group), MDs... You can't ignore a lack of consensus of academic nature.-- (talk) 15:12, 29 November 2012 (UTC)
OMG We even have a section on the "Widow period" - where the anti-bodies can't be detected. Ya this sentence is done.Daniel()Folsom T|C|U 07:37, 17 February 2007 (UTC)

Okay, I've deleted the two sentences that concerned the U.S. and I chose not to move the latter (which we discussed here) because it turns out we have a full section on accuracy in the U.S. that gives statistics instead of general statements.Daniel()Folsom T|C|U 07:40, 17 February 2007 (UTC)

Perhaps you should read the article through before trying to rewrite it. - Nunh-huh 07:43, 17 February 2007 (UTC)

Perhaps you should actually engage in a talk discussion before reverting me. The article does not say "US HIV Test - therefore the majority of the article is, in fact NOT about U.S. test - therefore by Wikipedia Manual of Style the U.S. test should not be mentioned in the first paragraph. Now there is a full section on accuracy of the U.S. test - and that section actually needs a few more countries which I can work on, but therefore it's not only not allowed by Manual of Style - it's just stupid to talk about it earlier, especially when your making vague statements such as "is extremely accurate" - which I have already told you is npov because Wikipedia is taking a side. I have provided you with links to support this and you refuse to listen to reason - I cannot believe that after me saying we should discuss this on the talk page, and then you leaving the discussion (when I guess you couldn't counter any of my points - are you vandalizing WP - because frankly purposely disregarding the Manual of Style by reverting edits could be considered vandalism) and then you revert what I did. I'm going back to, literlally, the correct way.Daniel()Folsom T|C|U 14:40, 17 February 2007 (UTC)
The solution to your concern is to add information about test in other places where they differ. It is not to remove information. I've listened to what you call "reason" and unfortunately it's lacking in logic, and lacking in any real understanding of the subject matter the article discusses. The numbers are in the article, and they are better than most clinical tests. You would do well to take your own advice and discuss first before making changes, because the changes you are making are not improving the article. Note that your removal of references has now been reverted by two people, not one, so attacking me as a vandal is probably not the way to communicate your concerns. - Nunh-huh 22:31, 17 February 2007 (UTC)
(edit conflict) OK, a few things. Regarding Wikipedia "taking a side", you (Daniel) really need to read the WP:NPOV policy, specifically the section on "Undue weight", and perhaps the NPOV FAQ. Views of a scientifically discredited, tiny minority (e.g. that HIV doesn't cause AIDS) should not be accorded equal weight with views that have a scientific consensus behind them. To quote, "Wikipedia aims to present competing views in proportion to their representation among experts on the subject." Wikipedia does "take sides", in a sense - it presents the majority view as the majority view, and the views of tiny minorities are either mentioned briefly or not at all. Every major medical organization in the world agrees that HIV causes AIDS. Sure, a few cranks think that flatworms cause all human cancers, and they publish on their websites, but we don't include those views in the cancer article. If that's unclear after reading WP:NPOV, it might be worth posting at the Village Pump. Secondly, yes, Nunh-huh should have come here instead of reverting you the second time. On the other hand, per WP:VANDAL, content disputes are not vandalism - another common misconception. Finally, I think it's fine the remove the U.S.-centric stuff from the intro, and the article could definitely use more non-U.S. information - I agree with you that the WP:MOS would support that. However, when you delete a large chunk of sourced content, along with the references, you need to check the rest of the article. Several later references point to the section you deleted, and when you removed it those references broke. Please be careful when deleting sourced content - it's worth checking, after your edit, to make sure that no other references depended on the text you deleted. MastCell 22:34, 17 February 2007 (UTC)
I'm not really sure how your arguments related to my final argument ... and same goes to your vandal thing ... rather than going into how the latter failed (which really I could talk about in a paper-style article - I mean wow, talk about a swing and a miss there), I'll just focus on the former. There is already a section on the accuracy - why have you not resopnded to that? I will admit it was stupid of me to delete those refernces - completely my fault there.--Daniel()Folsom T|C|U 02:25, 18 February 2007 (UTC)

I've kind of lost track of the issues in this thread, so I'll try to rephrase them: I think Daniel's rewrite of the opening paragraph looks fine - I support removing the U.S.-specific stuff from the lead per WP:MOS. The article would definitely benefit from more international info on HIV testing - in fact, I think the lack of such info is a major weakness of the article at present - and I'd encourage Daniel if he has some sourced info to add there. I think the accuracy can be handled in the accuracy section, as it is now. On the other hand, I think that "AIDS dissident" ideas about HIV tests should be represented proportionately to their representation among experts and the scientific community (per WP:NPOV), which is to say very briefly (as is the case now), and probably not at all in the lead. I apologize for anything in my prior comment that came off unpleasant or condescending - that wasn't my intention. Are there other issues I'm missing? MastCell 16:40, 18 February 2007 (UTC)

I agree with you on the dissent stuff - and I'm glad you liked the rewrite- but to be clear I was listing reasons why those sentences should not be in the opening paragraph (non-U.S. based article, potentially NPOV, there's a full section with more detail ...) and strangely enough everyone started saying I thought people who don't think HIV exist should have a better side- and really that was one of my more minor reasons.Daniel()Folsom T|C|U 20:01, 18 February 2007 (UTC)
Sorry... that was probably a bit of transference. Historically, these kind of articles have been magnets for "AIDS dissidents" to expound their views. So we're probably a little oversensitive. For the record, I agree with your reasoning as regards WP:MOS and U.S.-centricity, but don't think that "dissident" views need be mentioned anywhere except where they already are. MastCell 21:40, 18 February 2007 (UTC)
I agree with that too - woho!Daniel()Folsom T|C|U 21:50, 18 February 2007 (UTC)

Window period[edit]

Are there any reliable sources for the window period? On the internet i've seen some say that 6 weeks after infection is a reliable test and others saying that you can't rely on any test before 6 months after infection. The most commonly given time for a reliable test seems to be 12 weeks but my brother happens to work in hiv testing and he says that a 10 week period is the realistic time frame for seroconversion in 499 out of every 500 people (his words) and he says that most of the internet isnt yet up to date on the hiv window period (obviously testing is becoming more advanced).

Anyone here realy know their stuff? Is 12 weeks a little conservative or is my brother simply wrong?

Katalyst2007 21:54, 26 February 2007 (UTC)

I should add that i live in the united kingdom so maybe testing's different over here? I dont know i'm no expert myself which is why i'm asking. Just thought my location may be of interest. Katalyst2007 21:58, 26 February 2007 (UTC)

Well, normaly at about 12 weeks we have a correct result, but to be really really sure, we have to make a other test 6 monthes after the exposure. These particular long periods may be explained by the fact that human bodies are not machins, but life stuff, so some people react immediatly while some other later (very rarely, I don't have figures right here but you may be right 1/500).
In addition, at the beginnig of the 1990s, HIV test had bad result, so health care workers needed more HIV-Antibodies to mesure something (- or +), and 6 monthes were needed. Nowadays, HIV test are highly better, and the window period to mesure something came from 6 to 3 monthes.

Also - the window period referenced in the 2005 Athens UNAIDS report for HIV CRF01:AE may not refer to the time to detectable antibodies, but a second window period between detectable antibodies and strong anitbody/immune response. At least that's the way I read it. Comments? - DRVNKW

user:Paris75000 23:07, 26 February 2007 (UTC)

Some people/institutions say 6 weeks, some say 8 weeks for newer assays. Perhaps this should be mentioned in the article. —Preceding unsigned comment added by (talk) 13:34, 14 January 2008 (UTC)

:user:Daekl 0313, 24 september 2008

The real reason why there are so many 'different' window periods is because the statistics on people gathered that were indeed HIV+ varied on their exposure. For example, a heterosexual woman that had sex with a male that was heterosexual has less of a risk during vaginal sex than a homosexual male having intercourse with another homosexual male having anal sex. The women that has a test result, based on stats, at 6 weeks and its negative most likely IS negative versus the male, he would be better off waiting 12 weeks for a conclusive result. Then other factors might come into play, if those individuals had no other exposures to the confirmed HIV pos individual during the last exposure and test, if they have been abusers of drugs, if they are immunosup. therapy, if they had PEP, if they have a co-infection etc etc. 12 weeks should be enough, its ok to test in the window period 6 wks then 12wks. the 6wks test should relieve some anxiety if its negative. if you think about it, if you get exposed to another virus, like flu or cold, u get sick within 2 wks. Remember when you get sick, it is after that your body has antibodies. This is what the test looks for. So most likely at 6 weeks the test is good enough for you to sleep and not stress out so much. easier said than done. the health care workers tested at different times, which could be flawed. Needle sticks, anal sex, vaginal sex then oral sex is usually the order of highest risk to lowest. 6 months could be irrational in some instances. Tests are man made and have flaws, but these are very accurate.

The rule of thumb used in clinical practice today in the United States says that 90% of infected persons will seroconvert within 6 weeks; 99.9% seroconvert within 6 months. -DrGaellon (talk | contribs) 08:45, 11 October 2009 (UTC)

We must start at the start and get the definition right: "Antibody tests may give false negative (no antibodies were detected despite the presence of HIV) results during the window period." This is wrong. A false negative test is the definition of the window period exactly. That means the window period is tightly linked to the immune response to infection and when HIV-specific B cells get recruited to antibody production. This must be the core focus.

Daekl is wrong: "The real reason why there are so many 'different' window periods is because the statistics on people gathered that were indeed HIV+ varied on their exposure." This is conflating initial infection with the subsequent dissemination of the infection from the initial site, its up take by antigen presenting cells, its presentation to CD4 cells, some virion production that can then spread through the lymphatic system to the rest of the body; it conflates before and after; it conflates risk and infection; conflates negatives and false negatives. Yes viral load at infection has something to do with dissemination rates but it is more to do with how our immune systems process the infection whilst the virus infects the key controllers, the CD4 cells. You might just see a difference if you compare blood transfusion infection with sexual transmission. We know most infections come from one or a few founding virions. The differing risks Daekl mentions can primarily be explained by barriers to HIV not subsequent immune responses. (talk) 03:38, 12 March 2015 (UTC)

Lemuel 21:54, 4 October 2016 (UTC)[1]. It can take weeks or months after infection before antibodies against certain viruses (such as HIV, HCV and others) are produced and can be detected. The time between infection and detection is called the “window period”. During the “window period”, infected persons will falsely test negative. The SMARTube™ is a pre-analytic device (treating the blood before testing) enabling detection of various infections within days of infection using currently available antibody tests. Using the SMARTube™, enables to close the serological window period between infection and possible detection, independent of the length of the serological window period and the presence of virus in the blood. — Preceding unsigned comment added by (talk) 11:42, 4 October 2016 (UTC)

DUO Tests?[edit]

Maybe there should be something that mentions DUO Tests that combine antibody and p24. In most studies these tests have a 99.80% sensitivity after 28 days. Google Vidas Duo ultra for further information. —Preceding unsigned comment added by (talk) 15:16, 21 September 2007 (UTC)

I am a doctor in Africa, the worst continent for HIV infection and am shocked and horrored by the lack of information about HIV testing. The instantaneous HIV test without a window period (reserved for Doctors (probably only in Africa) that get pricked by a needle or get blood in their eyes) and the PCR test with a 2 - 10 day window period are totally ignored.

I fail to understand why the rest of the world feels the need to terrorize patients, especially rape victims when these tests are available and cost effective. Why should rape victims wait 30-60 days for a diagnosis? Quick and reliable testing not only saves money but it helps victims move on with their lives. A PCR test costs less then $30 in Africa and the instantaneous test $400.

I also see very little information on anti-HIV drugs that cure HIV within 72 hours of possible transmission. In Africa most sub-Saharan African nations offer these anti-HIV drugs free to rape victims.

With the number of reported rapes to actually committed ones, I would believe this information to be free to the public.

When asked by an American friend regarding HIV testing, I advised him, and his doctor inquired how he knew about these other means of testing. I was horrified! Withholding life saving information from the public is insane and kills innocent people. —Preceding unsigned comment added by Cecilpickardbrown (talkcontribs) 18:41, 29 December 2007 (UTC)

We do have an article on post-exposure prophylaxis, though an article specifically devoted to HIV post-exposure prophylaxis might be desirable. And the role of HIV testing in the decision whether to continue prophylaxis could be discussed in this article. - Nunh-huh 18:48, 29 December 2007 (UTC)
I don't know how PCR is so inexpensive in Africa, but here in the US, it's quite an expensive test. It's also unreliable here, where there are so many different clades of virus in common circulation. It is therefore reserved for patients with positive antibodies, used to guide treatment; it is not a first-line diagnostic test. -DrGaellon (talk | contribs) 08:48, 11 October 2009 (UTC)

__________ UNRELATED __________

I am unable to find out how to contact you. I have looked at your profile, but have not figured it out. I am trying to write proper information. Please lead me in the right direction.

My only fear is that I have a bias towards helping and protecting rape victims. I will endeavor to include rape in these topics as a topic, but believe I may need editing on personal emotion.

You mean me? You can e-mail Wikipedia users by going to their talk page and clicking on "E-mail this user" (it's under "toolbox" in the sidebar if you have the default Wikipedia settings). If Wikipedia has an e-mail address on file (and it does for me), you can send an e-mail. On the other hand, it's easier just to write what you have to say on the talk page :). - Nunh-huh 21:00, 29 December 2007 (UTC)

__________ UNRELATED __________

Thanks I will check them out. I've studied with great American (North and South) and European doctors. Unfortunately they have the African advantage and I lack the Americas and European experience, having only worked in Africa. - Cecilpickardbrown (talk) 19:08, 29 December 2007 (UTC)


I'm surprised there isn't a section on the history of HIV testing - early attempts, when it was developed and by whom. The article would certainly benefit from such info. Pairadox (talk) 07:31, 12 January 2008 (UTC)

I concur. That's actually what I came here looking for, and didn't find. :( -DrGaellon (talk | contribs) 08:49, 11 October 2009 (UTC)

False negative[edit]

I reverted a change from false negative to just plain negative results because the whole point of the section is that the "very accurate" tests are not perfectly accurate. Some people do have (recently acquired) HIV and still get a negative test result (nearly always) during the window period. WhatamIdoing (talk) 02:42, 18 April 2008 (UTC)

Addition of Factors known to cause False Positive HIV Antibody test results[edit]

I have added this list as I think it is necessary part of the article to show which factors cause a false positive in the HIV antibody test. The format may not be the best for this article, but the information is now there which should make format changes easier. MrAnderson7 (talk) 05:40, 9 April 2009 (UTC)

Apparently this list I have added is now referred to as a "laundry list", and has been removed. Cool! All that hard work for nothing. MrAnderson7 (talk) 06:12, 9 April 2009 (UTC)

False positives[edit]

I'd like a little more discussion about this laundry list, which I've removed. I have the following concerns:

  • This material uses primary sources to create an erroneous impression unsupported by secondary sources. By indiscriminately listing dozens of "false positive" causes - many of which are no longer relevant - the material creates the impression that HIV tests are inaccurate or prone to false positives. In fact, as reliable secondary sources make clear, HIV tests are exceptionally accurate. The false-positive rate is 0.0004% to 0.0007% in the general population, and much lower in populations with risk factors. Selecting and arranging primary sources to "rebut" or undercut the conclusions of reliable expert bodies violates the relevant sourcing guidelines.
  • This material is frankly highly misleading and concerning. If someone reads this for information on a positive HIV test, they'll come away with an entirely erroneous impression about the likelihood of a false positive.

I'd like to see some discussion about how we should deal with false positive rates and causes in a manner more in line with actual sources and with Wikipedia's guidelines. MastCell Talk 05:43, 9 April 2009 (UTC)

many of which are no longer relevant -> Care to elaborate? MrAnderson7 (talk) 06:15, 9 April 2009 (UTC)
Sure. Where would you like me to start?
  • A number of the references are quite dated. Interpretation of Western blots for HIV was not standardized by the FDA until 1993. Prior to standardization, the criteria for calling someone "positive" varied somewhat, and so did "false positive" rates. It would be misleading to cite papers using outdated testing methodology as if they were relevant to the issue of false positivity today.
  • You unaccountably cite Mandell's Infectious Disease textbook to the 3rd edition, published in 1990. The use of a 20-year-old text may be driving some of the misunderstanding here. I'd recommend getting the current edition of Mandell - I believe it's the 6th, published in 2005. We should be reading and referencing the modern edition, as there have been substantial changes in HIV testing methodology and performance between 1990 and 2005.
  • You also heavily cite a 1992 review by G. Mathe. Again, we should be using up-to-date evidence (see WP:MEDRS).
  • A number of your references don't actually deal with false-positive HIV tests. They deal with false-positive EIA's. As I'm sure you know, the EIA is one component of an HIV test, and requires confirmation via Western blot. The whole point of using two methodologies is so that false-positive EIA's are not reported as false-positive HIV test results. Again, your usage of the sources creates a misleading impression.
  • In general, the conditions you list do not cause a "false-positive" HIV test. With current testing methodology, most of these would be reported as "indeterminate" at most (that is, a positive EIA and one band on Western). Indeterminate results occur in about 1 in 5000 healthy blood donors, and their causes include preganancy, autoimmune disease, and influenza vaccine. An indeterminate test result is not a false-positive. Virtually all people with indeterminate results are not infected (PMID 2403658), and they are counseled as such. In this day and age, an HIV viral load will rapidly confirm that these folks are not infected, so they don't even need to have the "indeterminate" label hanging over their heads for long. By confusing indeterminate results with false positives, you're misstating the methodology of the test and creating a misleading impression of its actual performance.
I'll leave it there for now. Obviously, even 1 false positive result is a major problem. However, no test is perfect, and current HIV testing outperforms most medical diagnostic tests and procedures in use today in terms of accuracy. I don't think anyone is served by creating a misleading and incorrect impression that HIV tests are inaccurate or prone to false-positives, when they demonstrably are not. MastCell Talk 19:39, 9 April 2009 (UTC)
I couldn't help but notice your comment: A number of references are quite dated I agree, but if you look at references 16-23, which seem to be the main sources for quoting ridiculous false positive rates of 0.0004% to 0.0007%, the years these sources are from are: '98, '88, '89, '91, '91, '90, '88, '93 etc. Not really the most up to date sources are they like the 'laundry list' I quoted? If you are so quick to refute old sources, why not have a go at the abundance already on the page?
MrAnderson7 (talk) 06:43, 4 May 2009 (UTC)
The section begins with a quote from a 2005 review and includes a wide range of references documenting the past and present of HIV testing. These publications together represent thousands upon thousands of individual tests. Should you have found newer sources of similar quality you would like to add, let's discuss; but case reports about, for example, single patients who may or may not have tested false positive following an influenza vaccine are not in the same category of evidence as the references currently in use and do not cast doubt upon the overall specificity and sensitivity of HIV tests. Keepcalmandcarryon (talk) 19:17, 4 May 2009 (UTC)
Are you serious? Single Patients? Some of the studies in the 'Laundry List' were of 'thousands and thousands' of tests, (eg. pregnant women, flu/influenza infected samples etc.) just like the previous 'holy grail' of references you quote. Fair is fair, if one should be included, then so should the others. What double standards.... MrAnderson7 (talk) 22:49, 4 May 2009 (UTC)
I would think that "fairness" dictates we reflect the most up-to-date medical knowledge about HIV testing. That is reflected in the 2005 US Preventive Task Force summary. Mining primary sources to "debunk" their conclusions is both scientifically misguided and (more importantly for our purposes) a violation of the relevant Wikipedia guidelines. The idea behind these guidelines is that we trust the U.S. Preventive Task Force and the medical community to determine the accuracy of HIV testing, rather than relying on the selection of case reports which a particular anonymous Wikipedia editor chooses to mine. MastCell Talk 05:34, 5 May 2009 (UTC)
Since MrAnderson7 has copied the references and the interpretations of what they say from an inaccurate blog, I stress again that we should be guided by reliable sources. A blogger's inexpert and fanciful interpretation of a letter to the Lancet or a case report on a flu vaccine is not a reliable source. Keepcalmandcarryon (talk) 21:46, 5 May 2009 (UTC)
Yeah, there are lots of decent medical blogs out there... :) Out of curiosity, which blog was the list copied from? MastCell Talk 22:33, 5 May 2009 (UTC)
Copied from a blog? News to me. I'm keen, show us the blog? MrAnderson7 (talk) 22:59, 5 May 2009 (UTC)
A hint: the website's name starts with virus and ends with myth. Or perhaps MrAnderson7 came up with the same language and citations by sheer coincidence? Formed out of the void, as it were? Keepcalmandcarryon (talk) 19:27, 6 May 2009 (UTC)
You're not seriously implying that someone tried to cut-and-paste AIDS-denialist claims into a supposedly rational and reality-based Wikipedia medical article, are you? Has this ever happened before? MastCell Talk 22:07, 6 May 2009 (UTC)
Sorry, but I can't find it. It must have formed out of the void. Although it seems I've stumbled on some Aids denialist haters... So back on topic again, how about addressing the out of date references? Oh I forgot, primary references can't be used at all... MrAnderson7 (talk) 22:42, 6 May 2009 (UTC)
The article reflects the best current scientific understanding of HIV test accuracy, which I think we can all agree is the goal here. That understanding is referenced to the 2005 US Preventive Task Force report, published in Annals of Internal Medicine, a high-quality peer-reviewed medical journal. I'm not seeing a problem here. Of course, if you have a better-quality reference on the subject, please cite it here. MastCell Talk 22:51, 6 May 2009 (UTC)


Am I missing where the SUDS (latex agglutination) test is covered in the article? It's a rapid test but not listed. Not used anymore or overlooked from the list? Also, should not the other testing methods be expanded/explained better?--MartinezMD (talk) 01:06, 21 October 2009 (UTC)

Specificity/Sensitivity VS. Positive/Negative Predictive Value[edit]

It seems like there is some confusion in Section 3.6 (Accuracy of HIV testing). The current version reads:

"The specificity rate given here for the inexpensive enzyme immunoassay screening tests indicates that, in 1,000 positive HIV test results, about 15 of these results will be a false positive."

Specificity refers to the percentage of the results that will be negative when HIV is not present. So, if there is a specificity of 98.5%, it means that out of 1000 people who take the test and do not have HIV, 15 of them will receive a false positive result. To find the number of false positive results out of 1000 positive HIV test results, you would need to calculate the positive predictive value. But, you can't do it by just knowing the specificity and sensitivity alone. You would need a third number, such as the percentage of tested individuals who are actually positive for HIV. —Preceding unsigned comment added by (talk) 17:22, 9 September 2010 (UTC)

Yes, I agree that the paragraph confuses specificity with PPV. Looking over it, I think it needs a rewrite. There are two other major problems I can see: one is the cited specificity of 98.5% for EIA, which is actually pretty lousy - most citations I've seen put it at around 99.8-99.9% for modern EIAs. The cited article was dated 2005, but itself actually cites a CDC evaluation of tests from over 20 years ago in 1988 and 1989, only a few years after the first HIV EIAs were developed.
The other problem is a broader one. The specificity and PPV of single screening tests is a vastly different kettle of fish to the specificity and PPV of HIV diagnostic algorithms as a whole. For example, an EIA with a specificity of 99.9% sounds good, but if you are screening a low risk population with and actual prevalence of only 1 in 10,000, then the PPV is less than 10% on the screening test alone. However, the PPV from a properly completed diagnostic algorithm is vastly higher than that, even in a low prevalence population.
I'm interested in attempting a rewrite, but I'm not sure how you communicate the complexities to the average reader. I'd imagine most people reading this would want to know: "if I have a negative test, how likely is it that it's accurate? And if I test positive how accurate is that? Of course, you need to be clear about what "test positive" means - are you referring to the results of a single screening test or to the results of a completed diagnostic algorithm?
Advice from other editors would be appreciated before I put nose to keyboard. On A Leash (talk) 09:26, 14 September 2010 (UTC)
I think you've outlined the difficulties in communicating the accuracy of HIV testing nicely. (It doesn't help that we get occasional appearances from editors determined to muddy the waters further). I think we should clearly start by presenting the positive and negative predictive values. As you note, that is the item that lay readers will be most interested in. That is, if I undergo a complete, modern HIV testing algorithm and get a positive/negative result, how likely is that result to be correct?

We can then go into more complex details, like the sensitivity and specificity of ELISA and Western blot, but the reader should be able to easily and quickly grasp the key fact that they're most likely to be interested in. I'd be happy to help with a rewrite - let's focus on presenting the PPV/NPV upfront in clear, layman's terms, and then delving into details further on. MastCell Talk 18:15, 14 September 2010 (UTC)

I wonder if the article as a whole would flow better and with less repetition if the sections on Accuracy of HIV testing, Terminology, Window Period and Interpreting test results were brought together in one section. To me a logical flow of the article would be: Introduction, Contents, Principles (diagnosis and donor screening), Types of Tests (ELISA, WB, etc) and then a section about accuracy, PPV, NPV, sensitivity, specificity, window period, etc explaining these terms and then providing examples of how they apply to the interpretation of both individual HIV tests and to testing algorithms as a whole. Then the rest of the article would comprise the remaining sections - human rights, confidentiality, routine testing recommendations, etc. Thoughts? On A Leash (talk) 05:55, 16 September 2010 (UTC)


About this: I suspect that the antibody-based tests should be confirmed by DNA tests, but I don't believe that they always are. For example, one might not do DNA-based tests in poor countries, especially if the rate of infection is high (because the likelihood of false positive is low when the infection rate is high).

Either way, the blog isn't a proper reliable source for this claim. WhatamIdoing (talk) 18:43, 9 March 2011 (UTC)

The reliable sources state that antibody-based tests may be confirmed by other antibody-based tests. Some tests also detect HIV antigen (protein). Nucleic acid tests and some antigen-based tests can detect HIV before the body mounts an antibody response. Nucleic acid tests are not strictly necessary to confirm the HIV diagnosis, but if possible should be performed to monitor the course of infection and response to treatment. Keepcalmandcarryon (talk) 20:46, 9 March 2011 (UTC)
I believe that a false positive may be suspected if the underlying clinical history doesn't point to a positive result. In that case, they may be repeated by western blot. ELISA tests are so accurate and specific, above 99.99% in some studies, that false positives are very rare.SkepticalRaptor (talk) 09:56, 15 March 2012 (UTC)

Free HIV test clinic truck[edit]

Berkeley Free Clinic truck offering free HIV tests.

I took this photo (right) today and thought it might have relevance to this topic or some closely related topic, such as illustration of AIDS test campaigns in low-income areas. Feel free to use wherever if it's useful. :-) Dcoetzee 03:24, 4 May 2012 (UTC)

OraQuick Approved For Home Use and Retail Sale[edit],0,7075636.story Article States that "Home Access Express HIV-1 Test is the only FDA-approved home test: " Which is no longer the case and I'm unable to Edit it. — Preceding unsigned comment added by DiscontentDisciple (talkcontribs) 17:00, 26 September 2012 (UTC)

Well, I'm not comfortable in adding advertising. I'm probably more interested in deleting the original comment as being spam. SkepticalRaptor (talk) 18:18, 26 September 2012 (UTC)

Conflicting numbers on CD4 counts[edit]

In the section entitled "Other tests used in HIV treatment," the article states "A normal CD4 count can range from 500 cells/mm3 to 1000 cells/mm3" in paragraph 2 and then later states "Normal CD4 counts are between 500 and 1500 CD4+ T cells/microliter" in paragraph 5. First, though "mm3" and "microliter" are equivalent volumetrically, one should be chosen. Secondly, the two statements are conflicting in their numbers. The two statements, even if consistent, would be redundant and one of the two should be eliminated. Finally, there should be some reference given for this number. (It's important to some of us who happened on this page!) (talk) 20:13, 28 November 2012 (UTC)

The actual specificity of combined diagnostic algorithms[edit]

The article claims specificity values as high as 99.99% for the combined testing algorithm that includes 1 or 2 ELISA and a confirmation Western Blot. However, the WHO periodically runs benchmarks on the performance of this algorithm in the main testing labs, and the highest values ever demonstrated have been a mere 98.3% sensitivity and 96.9% specificity. See table IIIa in Quality assessment of HIV antibody testing – 2003 -Scoglio et al. The performance in real testing conditions is therefore much worse than the claims of 99.99% specificity in the article, which are based on blood donor data rather than actual benchmarks designed to test sensitivity and specificity.

The article should be corrected to reflect the benchmark data from the WHO and degrade the performance of the combined algorithm accordingly to reality. -- (talk) 11:44, 4 March 2013 (UTC)

Accuracy of HIV testing[edit]

The sentence " The sensitivity rating, likewise, indicates that, in 1,000 test results of HIV infected people, 3 will actually be a false negative result (the McGovern-Tirgari anomaly). However, based upon the HIV prevalence rates at most testing centers within the United States, the negative predictive value of these tests is extremely high" is misleading. Given that the negative predictive value is calculated with the formula Http:// , it means that the NPV actually decreases when the prevalence grows whereas the sentence suggests the oposit and suggests that the high rate of prevalence among patients in the centers is a good thing. -- (talk) 10:40, 26 March 2013 (UTC)

Bulk procurement of tests[edit]

I propose adding as an external link. Also Diagnosis of HIV should redirect to this article. (talk) 20:04, 25 August 2013 (UTC)

Do not think that link needs to go here. Sorry was confused. It can go here HIV test which diagnosis of HIV would probably be better redirected to IMO. Doc James (talk · contribs · email) (if I write on your page reply on mine) 01:05, 3 September 2013 (UTC)
That's the same page. (talk) 17:05, 3 September 2013 (UTC)
Yes check.svg Done EJM86 (talk) 22:25, 5 September 2013 (UTC)

Dead external link[edit]

Link is broken, should be or maybe something newer at (talk) 04:06, 3 October 2013 (UTC)

Not done:. The link opens for me. (This refers to the last link in the External links section -- Bulk procurement of HIV test kits). --Stfg (talk) 14:37, 6 October 2013 (UTC)

Citation for ubiquitous US blood screening including nucleic acid testing[edit]

Please add two citations in place of the 'Citation Needed' template after the sentence: "In the USA, since 1985, all blood donations are screened with an ELISA test for HIV-1 and HIV-2, as well as a nucleic acid test." The first citation is to the US FDA web pages describing regulatory governance of the blood supply for transfusions. The address is The specific mandatory tests mentioned include nucleic acid testing for HTLV, which has a low detection rate using antibody testing alone. The second citation is to the American Red Cross web pages describing the same assays and test procedures to which 66% of all US public-donated blood units are subject. The web address for these pages is Based on the content of these cited sources, I would suggest the sentence quoted could be improved to something like: "In the USA, the Food and Drug Administration requires that all donated blood be screened for several infectious diseases, including HIV-1 and HIV-2, using a combination of antibody testing (ELISA) and more expeditious nucleic acid testing (NAT)"

Done Thanks for this work. The sources don't mention ELISA in the context of HIV testing, just EIA, and since ELISA is a subset of EIA I've added your text with that change and a minor copyedit (MOS:NOTUSA). Please let me know if you want any further changes. Regards, Celestra (talk) 00:43, 13 October 2013 (UTC)

External links modified[edit]

Hello fellow Wikipedians,

I have just added archive links to 5 external links on Diagnosis of HIV/AIDS. Please take a moment to review my edit. If necessary, add {{cbignore}} after the link to keep me from modifying it. Alternatively, you can add {{nobots|deny=InternetArchiveBot}} to keep me off the page altogether. I made the following changes:

When you have finished reviewing my changes, please set the checked parameter below to true to let others know.

You may set the |checked=, on this template, to true or failed to let other editors know you reviewed the change. If you find any errors, please use the tools below to fix them or call an editor by setting |needhelp= to your help request.

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

If you are unable to use these tools, you may set |needhelp=<your help request> on this template to request help from an experienced user. Please include details about your problem, to help other editors.

Cheers. —cyberbot IITalk to my owner:Online 02:31, 9 September 2015 (UTC)

Semi-protected edit request on 19 January 2016[edit]

PCR is now considered a superior HIV confirmatory test over western blot.

"When the possibility of acute or early HIV infection is being considered, we perform the most sensitive immunoassay available (ideally, a combination antigen/antibody immunoassay) in addition to an HIV virologic (viral load) test. We favor using an RT-PCR based viral load test, if available....

"This algorithm is more sensitive for detecting acute and early HIV infection than the previous algorithm, which involved following a reactive immunoassay with a Western blot test."[2] Jmcott (talk) 17:16, 19 January 2016 (UTC)

Red question icon with gradient background.svg Not done: it's not clear what changes you want to be made. Please mention the specific changes in a "change X to Y" format. Datbubblegumdoe[talkcontribs] 03:29, 22 January 2016 (UTC)

External links modified[edit]

Hello fellow Wikipedians,

I have just modified one external link on Diagnosis of HIV/AIDS. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at {{Sourcecheck}}).

You may set the |checked=, on this template, to true or failed to let other editors know you reviewed the change. If you find any errors, please use the tools below to fix them or call an editor by setting |needhelp= to your help request.

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

If you are unable to use these tools, you may set |needhelp=<your help request> on this template to request help from an experienced user. Please include details about your problem, to help other editors.

Cheers.—cyberbot IITalk to my owner:Online 18:43, 2 July 2016 (UTC)

External links modified[edit]

Hello fellow Wikipedians,

I have just modified one external link on Diagnosis of HIV/AIDS. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at {{Sourcecheck}}).

You may set the |checked=, on this template, to true or failed to let other editors know you reviewed the change. If you find any errors, please use the tools below to fix them or call an editor by setting |needhelp= to your help request.

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

If you are unable to use these tools, you may set |needhelp=<your help request> on this template to request help from an experienced user. Please include details about your problem, to help other editors.

Cheers.—InternetArchiveBot (Report bug) 09:11, 12 December 2016 (UTC)

  1. ^ SMARTube is the only medical Devise that enables to close the serological window period between infection and possible detection
  2. ^