tetrahedron.org MANUFACTURING CERTAINTY (in Pseudo Science)

June 23, 2003

MANUFACTURING CERTAINTY (in Pseudo Science)

By David Crowe

“Tests indicate that you have a 90% chance of being infected with a deadly virus. There is a 50% probability that it will cause disease within the next 10 years, and a possibility that it never will. If you take the drugs that I offer, there is a significant risk that you will experience a great decline in your quality of life, and a possibility that the drugs will kill you.”

Although it might be the truth, you are unlikely to hear a doctor saying this, because neither the doctor nor the patient can deal with the uncertainty that it admits.

Technology is the practical application of science, and one of the major distinctions is its need for certainty. Studying semiconductor physics can be a beautiful thing, but it remains pure science until a discovery results in products that can be reliably manufactured and used. Biological systems, especially human beings, are far more complex and less predictable than inorganic systems. Medicine, being the practical application (technology) of human biologic science, requires a high degree of certainty before new discoveries can be applied.

Unfortunately, a feeling of certainty can be manufactured, and there are many motivations to do so.

On October 12, 2001, a CDC scientist phoned then mayor of New York City, Rudolph Giuliani, to tell him that, “with a high degree of probability”, a sample of skin from an NBC employee in Manhattan was positive for cutaneous anthrax. The CDC scientist had this confidence, because he had confidence in a test that a colleague had previously developed. But this was not good enough for Giuliani. “Don�t give me that stuff. Is it anthrax or not?” An unqualified “Yes” from the CDC scientist kicked off the anthrax crisis in New York City. [Altman, 2001]

A “No,” under the circumstances, would have been almost impossible. The consequences for the CDC and Giuliani, if others had later confirmed anthrax, would have been devastating to their careers. While reporters might have questioned the accuracy of a “No,” there was not a whisper of dissent on the “Yes.”

Medical tests are a common way to manufacture certainty. A test usually measures a �surrogate marker� for a condition, something that is otherwise invisible, or at least much more difficult, expensive and time consuming to find directly. A nicely packaged test can instill confidence and, in a sense, create a disease when a positive test result is accepted without any symptoms being present.

An HIV test is perhaps the best example. A positive test is devastating to most people, particularly those who are outside the traditional risk groups and completely unprepared. Feelings of doom come, not surprisingly, even to those who are perfectly healthy at the time of the test [Gala, 1992].

Desperate feelings lead to desperate actions, and, for HIV, the desperate action is to take AIDS medications. Antiviral drugs have fatal side effects, and even those who avoid that are likely to experience a destruction of their quality of life, even if they were completely healthy at the time of the test [Goodman, 2002].

Obviously, the doctor and patient must feel certain that tests are accurate. If the patient was told that there was only a 90% certainty that the test was accurate they might be much less likely to take medications carrying such risks.

The almost universal impression among scientists, the media, governments and the general public that HIV tests are accurate enough to stake your life on is, strangely enough, so strong because there is no absolute measure against which the tests can be validated. Instead of accepting this as uncertainty over whether the tests are meaningful, it is accepted as lack of proof that they are not highly accurate.

All that Robert Gallo�s and Luc Montagnier�s research teams found was a high correlation between their antibody tests and AIDS. People with AIDS had a high probability (88% in the case of Gallo [Sarngadharan, 1984]) of testing positive, and people without AIDS had a very low probability of testing positive. A huge conceptual leap over a chasm of uncertainty was to conclude from this evidence that a positive test in a healthy person proved they had a condition that would inevitably kill them.

The science of HIV testing has progressed since then, but only in technological ways (such as the use of monoclonal antibodies); the original logical uncertainties still exist. Almost every scientific paper concerning HIV tests still uses antibody tests as the “gold standard.” This is unusual because antibody tests, even if one ignores the possibility of cross reactions, can only prove past exposure to a virus, not current infection.

HIV antigen tests, which are more direct, are only positive in about half the people who are HIV-antibody positive [McKinney, 1991; Semple, 1991]. This finding is explained away through an immune reaction which masks the antigen. But, this implies that the HIV infection is conquered, which is not compatible with the notion that HIV infection is incurable. Virus cultivation, often erroneously called ‘isolation’ is an even older method than antibody testing for HIV, but apart from being time consuming, expensive and difficult to perform, it also is negative quite frequently, and a positive antibody test usually trumps a negative culture [Layon, 1986] (and vice-versa [Eur Coll, 1991; Imagawa, 1989]).

The major new test since the early days of AIDS is the Polymerase Chain Reaction, often called ‘viral load’ when used for HIV tests. This also takes a back seat to antibody tests [Roche, 1996], likely because it is so ultra-sensitive that the risk of a false positive is high. Furthermore, detecting a snippet of genetic material (RNA or DNA, depending on the type of test) does not prove that the entire genome is present, and obviously does not prove that infectious virus particles are present. This test is particularly uncertain because the genetic material does not come from purified virus. Even accepting the test�s ability to specifically detect HIV DNA or RNA, one research team estimated that only one infectious virus particle was present for every 60,000 measured by viral load! [Piatak, 1993; Roche, 1996]

All HIV tests are indirect, even virus ‘isolation’ by culturing. Consequently, some ‘gold standard’ is necessary to validate them [Cleary, 1987; Abbott, 1997; Meyer, 1987; Daar, 2001; Papadopulos, 2003]. The only standard that is reasonable for a virus is actual purification direct from body fluids of people who are HIV infected and the inability to purify from people who are not. Virus purification would allow the proper characterization of the virus, so that antigens, antibodies, DNA and RNA that are generally believed to be from HIV could be proven to be from HIV (or not).

Without a �gold standard� for HIV infection the only way to validate the test is by repeating the test or by comparing it against different (also unvalidated) tests. This can establish the reproducibility of the test, but not its specificity (ability to react with the target and therefore avoid false positives) or sensitivity (ability to react to cases of infection and therefore avoid false negatives).

US army researchers claimed that the specificity of HIV antibody tests was only 1 false positive out of 135,187 tests [Burke, 1988]. However, although they claimed to have established a high specificity for antibody tests, they were actually verifying only reproducibility, and the researchers did not actually prove that the 15 people from this low risk population who were deemed to have had true positive tests actually had the virus in them.

Modern diseases that are blamed on a virus are often little more than the test because the disease can exist without clinical symptoms. There is an average of 10 years between becoming HIV positive and the first signs of AIDS in both rich countries [Munoz, 1995] and poor [Morgan, 2002]. In that time the HIV test is the only sign that anything is wrong. Worse yet, a low CD4 cell count test can result, in the United States, in a diagnosis of AIDS (not just HIV infection), again without any clinical symptoms. But even without symptoms a diagnosis of HIV infection or AIDS will still often result in treatment because of everyone�s confidence in the tests.

Other viral diseases might not have a long incubation period, but the test still plays the prime role in defining the condition. West Nile disease, for example, is associated with no illness in the majority of people who test positive, and serious illness in only about 1 out of 150 [Petersen, 2002]. The symptoms, when they do occur, are indistinguishable from many other viral diseases [CDC, 2002]. This has not resulted in a call to question the accuracy of the tests. Instead, the certainty that any symptoms found along with a positive test are due to the virus is so great that when the symptoms are uncharacteristic scientists want to add them to the definition, rather than to ask whether the tests are accurate and whether presence of a virus is proof of pathogenicity [Glass, 2002; Leis, 2002]

One of the strange phenomena with HIV and AIDS science was overwhelming feeling of certainty that crept over scientists in the mid-1980�s. Only 3.4% of papers in 1984 associated a reference to Gallo�s original 1984 papers on HIV (HTLV-III) with “explicit and unqualified” assertions that HIV caused AIDS but this increased to 25% in 1985 and 62% in 1986, even when these papers were referenced alone. [Epstein, 1996]

Kary Mullis, who received the 1993 Nobel for Chemistry (ironically because of his invention of the Polymerase Chain Reaction) has asked many scientists for a set of references that constitute proof that HIV causes AIDS [Duesberg, 1996] and has not yet received them. Yet, even without this proof being written down in a scientific paper, certainty still reigns.

SARS illustrates how quickly researchers can manufacture certainty today. The mainstream media (which claim to be “responsible”) have ensured us that everyone knows SARS is caused by a Coronavirus. Reports from Dr. Frank Plummer, one of Canada�s top virologist, that a diminishing percentage of patients (30% by mid-April) are testing positive do not dissuade them from this belief [Altman, 2003]. Everyone knows that there is no possible explanation for all the patients having some connection with the original cases other than an infectious agent, even though for some outbreaks there was no solid connection, and tautologically, the epidemiologic connection is supposed to be present before diagnosing SARS (as opposed to some other disease with similar symptoms). And, everyone also knows that there is no other explanation for the severity of the disease, certainly not the new phenomenon of aggressive prescription of steroids and the antiviral ribavirin that occurred as the fear of the outbreak spread [Koren, 2003].

What HIV/AIDS science took two years to do, SARS science took only two months to accomplish. I predict that a Coronavirus test will soon become part of the SARS case definition, which will immediately create a 100% correlation between the Coronavirus and SARS symptoms. Just as with AIDS, the same symptoms without a positive test will be another disease, and not taken nearly as seriously.

People demand simple answers to complex problems and modern medical science delivers. We are told that tests are highly accurate, that drugs will cure conditions or, if that is not possible, that they are the best bet. We are told that environmental conditions play little role in modern, emerging diseases. Alternative therapy is scoffed at because it has not been �proven� effective through randomized, placebo-controlled clinical trials.

The fundamental reason why this confidence game continues to be played is because of human laziness. It is much easier to learn about science by rote than by examining evidence and making up one�s own mind. Obviously, not every pronouncement on science can be taken seriously, so the status of a person or publisher becomes the way to distinguish between “good science” and “junk science.” Many people do not believe that they have the ability to understand scientific papers. The media, even most science reporters, are much more productive if they also adopt this attitude. Among scientists, there is a hierarchy which is constructed from the anonymous peer review system for publication and grant support. This allows longer-serving officers of science to anonymously subvert the attempts of younger scientists (and outsiders) to reappraise current dogmas, by denying them the ability to publish and obtain research funding.

E-Mail:
Firstname:
Lastname:
Action:

MANUFACTURING CERTAINTY (in Pseudo Science)

Leave a Comment! Cancel reply