The politicians

A panel of international experts was convened by President Thabo Mbeki’€™s AIDS panel in May last year to carry out further stringent studies on the reliability of the globally used Elisa HIV test. The group of four researchers appointed by Mbeki’s 44-member panel said the studies could help South Africa to improve its strategies for combating the spread of HIV which causes AIDS. The four researchers included AIDS dissident Peter Duesberg, Harvey Bialy of the Autonomous National University of Mexico, Helene Gayle of the Atlanta-based Centre for Disease Control and William Makgoba of the South African Medical Research Council. “In order to address this question in South Africa a series of HIV tests of the utmost stringency would be applied to see whether a degree of concordance exists between the tests,” Bialy said at the time.

Sibani Mngadi, spokesperson for Health Minister Dr Manto Tshabalala-Msimang said the minister would be discussing the report’€™s findings with minister in the office of the president Essop Pahad and Arts, Culture, Science and Technology Minister Ben Ngubane.

“They will then report to Cabinet, but I cannot say when this will happen,” Mngadi said.

Mbeki set up the panel in May after questioning assumptions that HIV was the sole cause of Aids. He suggested that other factors such as poverty, malnutrition and inadequate sanitation could also play a role. The panel comprised orthodox scientists as well as so-called dissidents who agreed that assumptions about HIV and Aids were open to question.

The insurance industry

Liberty Life’€™s Lorna Dee said they used the Elisa test “because it’€™s (at least) 99% correct”.

She said they realised that there were exceptions, because factors such as antibiotics could throw off the accuracy of a test.

“It is automatic for anyone positive to have a second test. The second test ‘€“ the Western Blott test ‘€“ is a more thorough test, more expensive and shows the exact nature of virus and what stage the person is in,” Dee said.

Old Mutual said they conducted a “normal” HIV test, the Elisa test.

“We usually do three Elisa tests – if the first Elisa positive but the second and third Elisa are negative then it’€™s classified a “doubtful case” and the company will decide whether to pay for further tests or not” said a woman at Old Mutual, who opted to remain anonymous.

She said that if the test was positive, Old Mutual wrote to the client’€™s doctor to inform the doctor of the result, saying it was only a screening test.

“Then the patient can go for a subsequent test if they wish to do so at their own expense,” she said.

Gerhard Jourbert executive director of the Life Offices’€™ Association (LOA), an umbrella body for life insurance companies said they believed that the three Elisa tests were reliable.

“We do not pay for Western Blot, but we also do not see it as the ultimate test.”

He stressed their procedure was for screening only, “we don’€™t do diagnostic testing”.

The LOA website, [ ‘€“ see section under “medical info’€™] clearly states HIV protocol used by members for testing.

The researchers

Elisa tests

The public and private sectors currently both use the Elisa test to screen for HIV.

Elisa stands for Enzyme-linked Immunosorbent Assay which tests for the presence of antibodies (an immune system response) to the HIV virus which may take up to three months from the date of infection to develop (The so-called “window period”).

Two Elisa tests are always done on the same blood before an HIV positive test result is given. The second test confirms the finding of the first; that HIV antibodies are present. In other words; all positive tests are routinely repeated and confirmed by another HIV antibody test.

The Elisa test has a 99,9% accuracy rate.

This rate of accuracy is not based on the say-so of the manufacturer but on a general average from a variety of laboratories that report the accuracy of their test results.

A 99,9% accuracy rate may seem fool-proof but it means that about one in every thousand results is wrong. When a lab is doing over 4000 tests per month, this means they could get about four false test results every month.

Note: 99.99% reliable = one in ten thousand; 99.95% is 1 in every 2000; 99.9% = 1 in a 1000) HIV tests are still streets ahead ‘€“ much more reliable and accurate ‘€“ than tests for most other infectious diseases like Tuberculosis, Malaria and Measles. (1 in 50 test results for measles is inaccuate)

When an HIV test comes up positive and the person is not in fact HIV infected this is because antibodies to other diseases sometimes react to the Elisa test.

Low positives are hard to interpret because they could indicate low levels of HIV infection, but they could also be non-specific reactivity caused by other antibodies. Low positives should always be interpreted carefully and re-confirmed at a later date.

Low positive results are sometimes interpreted as positive results without proper investigation and interpretation. Ideally, low positives should be followed up by an additional PCR (Polymerase Chain Reaction) test.

Low positive results used to occur when the Elisa test cross-reacted with other infectious diseases but this is no longer a problem with the newest version of the test, which has improved considerably over time.

However, what the latest tests do still cross-react with are glandular fever and systemic lupus. This is why a lab diagnosis is never cut and dry and must be interpreted carefully by the practitioner who gives the result.

The Polymerase Chain Reaction (PCR) tests

The PCR tests is also used in special circumstances.

The PCR test (Polymerase Chain Reaction) detects the genetic material of the HI virus from as early as 10 days after infection and will pick up the virus consistently from then on. It is therefore the best available test for HIV.

However, the PCR test costs R320 as opposed to R30 for the Elisa test in the public sector.

The private sector charges more for both tests, about R400 for a PCR and R100 for an Elisa.

Western blot test

The Western blot test is being re-evaluated for a number of reasons ‘€“ it is technically demanding on inexperienced laboratories, is difficult to standardise, different laboratories have different ways of determning results and it is expensive (About R400 in the private sector).

The new generation of Elisa and other enzyme immunoassays are at least as sensitive as the Western blot, much cheaper, and not as subject to observer bias, and would therefore be more economical and beneficial for HIV diagnostic purposes in developing countries.

Western blot’s advantages have been eroded by the accuracy of other HIV assays and the high sensitivity of the Elisa.


  • Health-e News

    Health-e News is South Africa's dedicated health news service and home to OurHealth citizen journalism. Follow us on Twitter @HealtheNews