Do HIV Tests Detect HIV?

March 5, 2012

SENSITIVITY & SPECIFICITY

In fact, the manufacturers of HIV antibody tests do report measures of accuracy known as sensitivity and specificity and, in many cases, these reported measures of accuracy are 99.9 percent, or even greater.

However, if one reads the package inserts for these test kits carefully, they will discover that these measures of accuracy have nothing to do with whether or not HIV is present.  In fact, the vast majority of these reported measures of accuracy refer to how well one test performs as compared to another test already on the market.  In other words, these measures of accuracy refer to the concordance between two tests; namely, how often a new test says a sample is positive when one already on the market says it is positive (sensitivity), and how often does this same test says a sample is negative when the one already on the market says it is negative (specificity).  Given the fact that all HIV antibody tests are based on the same molecular principles and built from the same portfolio of basic molecular building blocks, it is no wonder that there is a remarkable concordance between their outputs.  However, even if two different tests demonstrate 100 percent concordance, that does not justify the use of either for purposes other than what they have been validated and approved for.

By way of analogy, if two different manufactures were to construct two different timepieces using the same specifications for the gears and wheels, only to put them in different casings, it is very likely that these two clocks would perform quite similarly.  However, even if there were greater than 99.9 percent concordance between these two timepieces, one cannot conclude that either clock accurately measures time.

Since the first ELISA and WB tests approved by the FDA had no other approved test for comparison, the manufacturers of these tests choose to validate their products according to their ability to distinguish between persons with confirmed clinical AIDS, and healthy blood donors; the former of which have experienced a plethora of infections and illnesses, and the latter of which represent the healthiest of the healthy and are pre-screened to be free of any risk factors or infections.  And in fact, these two tests performed well in distinguishing between these two select populations.  However, this tells us nothing about the significance of positive test results in persons in the general population without clinical symptoms of AIDS, and who are not routine blood donors.

By way of analogy, a simple test of having persons walk the distance of a balance beam may perform remarkably at distinguishing between persons who are intoxicated, and say, young athletic adults who have not been drinking. However, even if a preliminary evaluation of this test revealed that 99 out of a hundred intoxicated persons fell off the beam (99% sensitivity), and only one out of a hundred young athletes fell off the beam (99% specificity), one cannot then simply apply this test to persons in the general population and conclude that 99% of all persons who fall off the beam must be drunk.  While it may be the case that 99% of all intoxicated persons in the general population who are subjected to this test will fall off the beam; many others may fall off for other reasons:  Perhaps they are clumsy, or had a recent ankle injury or knee surgery, or maybe they are blind.  It is for this reason that we cannot say with any degree of certainty how likely it is that any given individual who falls off the beam is drunk.  Nevertheless, this would still represent an example of an excellent “screening assay” in that 99% of intoxicated persons who participate in the test most likely would fall off.  However, in the absence of a follow-up “confirmatory assay” — in this case measuring the actual blood alcohol content in all positives — the majority of persons who fall from the beam (i.e., positive) may actually be sober.

Unfortunately, in the case of HIV, there is no confirmatory laboratory test for either antibodies to HIV, or the virus itself.  Although – at the recommendation of the CDC – the WB test is utilized as a confirmatory test for HIV, it is unequivocally not.  All of the WB tests on the market are approved only as additional, more specific tests for antibodies to HIV.  And even when positive, the best one can do is “presume” the sample is positive for antibodies.  And even if that presumption were correct, the significance of those antibodies in persons without symptoms is “unknown.”  In other words, WB cannot even confirm the presence of antibodies, let alone the virus itself.  It is for this reason that the manufacturers of these tests warn that all persons testing positive for antibodies should be referred to a physician, who can, based on risks, symptoms, and other laboratory data, “decide whether a diagnosis of HIV infection is accurate.”  The true value of the WB test is that, since it is more specific, it can be used to rule out infection, when negative, in persons testing positive on ELISA screening tests.

Although CDC researchers acknowledged by 1986 that “no established standard exists for identifying [HIV] infection in asymptomatic people” (Ward, 1986), they announced that same year that, “For public health purposes,” all persons testing positive on ELISA and WB “should be considered both infected and infective.”  (Emphasis added.)  Of the three references put forth to support this conclusion, one states: “… the frequency of virus in antibody-positive persons is yet to be determined;”  another makes no mention of whether or not antibodies can be used to infer infection,  and the last one makes reference to yet another CDC publication that states, “the proportion of these seropositive [i.e., antibody-positive] donors who have been infected with [HIV] is not known.”

Even if it were the case that only half, or even a quarter, of persons testing positive on both ELISA and WB were actually infected with the virus, if physicians follow the manufactures instructions and simply tell these persons you “may be infected,” then those who may be truly infected and are still without symptoms might continue to engage in, for example, unprotected sexual activities thereby contributing to the further spread of the virus into the general population.  So if we want to guarantee that none of these persons contribute to the further spread of HIV into the general population, it is necessary to ignore the manufacturers’ instructions and the scientific facts and tell all of these persons that they are infected regardless of whether there is any medical or scientific basis to support such a conclusion.

By way of analogy, in the interest of public health, it would be best to tell all persons who fell off the balance beam in the above example that they cannot drive.  After all, this would guarantee that 99% of the intoxicated persons in the group tested would not be on the road.  Many other sober persons who fell off for other reasons would likewise not be on the road.  However, in the absence of a confirmatory test, this would be the sacrifice that would have to be made in order to protect the public from intoxicated drivers. Nevertheless, to turn this scenario on its head and declare all persons who fell of the beam to be unequivocally intoxicated would be absurd and without any scientific merit whatsoever.

Regardless, shortly after FDA approval of the first WB test just over a year later in 1987, the CDC published another report that stripped out any reference to public health, and simply declared, “The presence of antibody [to HIV] indicates current infection [with HIV];” this time without making reference to any scientific study or internal document to substantiate this statement.

Fourteen years later (2001), the CDC published revised guidelines stating that persons testing positive for antibodies on both ELISA and WB “are considered HIV-positive and indicative of HIV infection.” Although the authors of this document emphasize that “approximately 5,000 abstracts were screened and approximately 600 relevant publications were reviewed,” along with “approximately 20 previously published CDC guidelines related to HIV,” they do not provide any references to substantiate this statement (i.e., that antibodies equal infection).  They also note that in cases where evidence to support their recommendations is lacking, “opinion of ‘best practices’ by specialists in the field have been used.”  In summary, the position taken by the CDC and the medical community that all persons testing positive for antibodies to HIV are unequivocally infected with HIV is not a position based on scientific evidence; but rather, one taken in the interest of “public health” based on the “opinion” of specialists in the field.

In addition to the concerns detailed above, the scientific literature reveals several studies demonstrating false positive HIV-antibody reactions on ELISA and WB as the result of other conditions completely unrelated to HIV.  These include conditions such as infection with other viruses and bacteria, autoimmune conditions, lymphoma, dermatologic disorders, vaccination, hepatitis, herpes, alcoholic liver disease, arthritis, multiple pregnancies.   (Midthun 1990, Celum 1994, Guan 2007)

Another class of FDA-approved “HIV tests,” which detect a molecular fragment of HIV (not the virus) are the so-called Viral Load (VL) tests.  Like the antibody tests, VL tests are not intended for use in diagnosing infection with HIV; rather, they are approved for either assisting a physician in offering a prognosis of disease progression, or to assist in the clinical management of medications in persons already deemed to be HIV positive by other means.  In fact, the manufacturers of these tests – i.e. COBAS AmpliPrep/TaqMan HIV-1 Test – explicitly warn that their tests are “not intended for use as a screening test for the presence of HIV-1 in blood or blood products or as a diagnostic test to confirm the presence of HIV-1 infection.”  The manufacturer of another VL test also emphasizes “The AMPLICOR HIV-l , MONITOR Test is not intended to be used as a screening test for HIV or as a diagnostic test to confirm the presence of HIV infection.” (Amplicore) The reason VL tests cannot be used to diagnose infection with HIV is because, as highlighted in the literature, “viral load tests for HIV-1 were neither developed nor evaluated for the diagnosis of HIV infection.”  (Rich 1999)  This is because, as mentioned above, there is no method to establish with certainty which samples are, or are not, truly HIV positive or negative, which could be used in order to establish how well the VL test results might compare.

The ultimate standard for proving the existence of a germ is to obtain it in sufficient quantities and purity so that its chemical, morphological, and biological properties can be determined.  Given that most germs are present in the body at insufficient levels to achieve this directly, they first have to be “grown” in the laboratory using a process known as “culture.” This involves taking a sample from a patient and placing it in an appropriate laboratory environment where the germ can replicate in vitro (outside of the body) to levels where it can be purified for characterization. This purification step is necessary in order to ensure that any chemical or biological phenomena observed during characterization is indeed due to the germ, rather than something else, either alone or in combination, that might be in the mixture.

Following such characterization of the purified germ, it may no longer be necessary for the scientist to purify the germ from every culture; rather, they can simply look for “phenomena,” which have been demonstrated to be unique to the newly characterized germ.  For example, the germ may have a peculiar shape that is so unique that the scientist can simply look at some fluid from the culture or directly from the patient under the microscope in order to declare the presence or absence of the germ.  Alternatively, the germ may have a unique chemical building block or unique biological activity that the scientist could use as proxy to determine if the germ is present or absent in the culture.

Unfortunately, some germs, such as HIV, are so fragile that they have never been obtained in purified form for direct chemical and biological characterization in the first place.  As a result, all chemical and biological phenomena said to be due to HIV are necessarily inferred through indirect techniques.  As such, while these phenomena are entirely consistent with the presence of HIV, it is impossible to prove that they are due to HIV, or even that they are unique to HIV.  Since the current consensus among virology experts is that a positive culture for HIV is synonymous with infection with HIV, it is reasonable to use HIV culture in combination with other factors to determine if a patient is infected with HIV.

For these reasons, the only valid way to determine that someone is infected with HIV requires the following minimum determination:

1. A positive result from ELISA

2. A positive result from WB

3. A positive culture of the virus

4. Risk factors consistent with the possibility of infection; AND

5. Clinical symptoms characteristic of AIDS that cannot be accounted for by other factors.

Anything short of meeting these criteria to determine the HIV status of a patient is speculative and wholly unreliable; but extremely profitable to the clinics that rely on HIV tests to convince healthy patients that they are infected with HIV.  Extricating oneself from an incompetent diagnosis based upon these tests is difficult, but not impossible.

For more information about HIV tests, visit DavidRasnick.com, Reappraising AIDS or VirusMyth.com.

 

Pages: 1 2

Tags: , , , , , ,

You must be logged in to comment

Log in