Re: Re: Re: Request for Peter Flegg and Outstanding 'Perth Group' Questions 23 November 2004
Previous Rapid Response Next Rapid Response Top
Peter J Flegg,
Blackpool, UK. FY3 8NR

Send response to journal:
Re: Re: Re: Re: Request for Peter Flegg and Outstanding 'Perth Group' Questions

Christopher Tyler has posed several questions regarding the use of tests for HIV in recent responses. This ground has been covered in the past, so I know he is well aware of the orthodox view on these tests. Since I am sure he does not expect to hear anything now to change his well -entrenched position, his questions are largely rhetorical, but deserve a response nevertheless.

Firstly a general point about the diagnostic process. A positive result for any test seldom confirms “proof” of the disease, as any clinician would tell him, be it blood sugar levels for diabetes or blood pressure for hypertension. The result need to be repeated, confirmed and placed into clinical context. Ultimately it is the clinician that makes the diagnosis, and not the laboratory.

A single positive HIV antibody ELISA or similar screening test is proof of nothing except that the source has a reactive assay. The result is merely supportive of the clinical diagnosis of possible HIV infection. First generation ELISA assays, although quite specific, were prone to false positive antibody reactions in patients with HLA class II antigens and therefore later 3rd generation assays using only purified or recombinant antigens have been developed which have largely overcome this problem, so Tyler’s reference to the numerous causes of “false positive” HIV tests is outdated.

If someone presented to me claiming they tested HIV positive on a “Home access express HIV-1 kit”, the first thing I would do would be to repeat the assay using a standard approved ELISA. I would then request confirmation utilising assays that use different antigen targets. The scenario Tyler describes would not arise in practice.

Further potential for confusion may result if the confirmatory assay is a Western Blot, and the result is “indeterminate”. HIV dissidents have long sought to exploit any apparent confusion over this issue to try and claim that HIV diagnostics is entirely bogus (and therefore as an argument that HIV does not exist). This is analogous to denying myocardial infarction exists because one laboratory uses troponin T as an index for myocardial damage while another laboratory uses troponin I, or different labs have different cut-off levels for a positive troponin. Negative HIV western blots have no reactive bands. Different countries may use slightly different interpretations of how many or which bands constitute a clear positive WB, or are “indeterminate”. In practice the issue is irrelevant, since with established HIV infection one usually sees more bands than you find at a LiveAid concert, and indeterminate reactions with only two or three bands are unusual. These may be a consequence of early HIV infection or represent a true false positive reaction. All indeterminate results will of course be repeated, and repeated again until it becomes clear what is going on. It is theoretically possible to have someone who is persistently “indeterminate” on WB (usually reactive to p18 and p24 or p55) but who tests persistently negative on other assays and for HIV-RNA. This person is very unlikely to be HIV infected.

The clinician ultimately has the responsibility for “making” the diagnosis of HIV infection or AIDS. Laboratory tests do not magically incorporate clinical information or sexual history, but the diagnosing clinician has to employ all this relevant information when interpreting the result. The positive predictive value (and therefore the clinician’s interpretation) of an “indeterminate” western blot or a weak positive ELISA is likely to be very different if the source comes from a high risk or low risk population. Therefore, if I saw a highly promiscuous gay man who recently had unprotected sexual intercourse I would carefully explain the situation and stress the need for repeated tests to clarify what might well represent a case of newly acquired HIV infection. I would also do an HIV-RNA level. If I were faced with a 50-year-old female with autoimmune disease who had tested weakly positive on ELISA at blood donation I would take a somewhat different approach, but the tests still need to be repeated and confirmed.

I could never provide the documented guarantee that Mr Tyler seeks that an isolated, unrepeated and uncoroborrated test is proof of any individual medical condition. I do not do this for patients with chest pain who have positive cardiac troponins, nor do I for patients with positive HIV assays. However, after further tests and clinical assessment, I hope to be in a position to tell the patient what is wrong with them, and to do so with a high degree of accuracy.

Competing interests: None declared