HIV tests 26 January 2005
Previous Rapid Response Next Rapid Response Top
Nicholas Bennett,
Infectious Disease Postdoc/Clinician
Department of Pediatrics, University Hospital, Syracuse NY

Send response to journal:
Re: HIV tests

Rob McGregor's points don't actually detract from the HIV testing methodlogy.

The "arbitrary" testing level I presume he refers to is the ELISA colourmetric absorption level. This is, of course, an issue with all ELISAs in diagnostic and research use. By this argument we should throw out the tests for TSH, LH, bHCG, GH, Rheumatoid factor...

Any diagnostic test depends on three things - reagents, equipment and samples. Prior to use in a diagnostic facility, all tests are required to be (or should be!) validated against a standardised panel. Such results are usually available in the literature of validation/development papers or in the product literature itself. The apparently "arbitrary" cutoffs are merely the same local guidelines that vary from hospital to hospital as regards more mundane tests like haemoglobin, WCC, LFTs etc.

The differences in Western Blot criteria again boil down to differences in the test kits used. Each kit is developed differently - some using recombinant proteins, some using virus preps, some using peptides. It makes sense that depending on the specific antigen being presented and the mode of presentation, sensitivity and specificity for individual bands will vary. This is why each kit includes appropriate interpretation instructions, again based upon comparisons with standardised panels.

The fact that HIV seropositivity correlates with biological markers of HIV replication, and is predictive of AIDS in untreated individuals, strongly supports the notion that the tests are far from meaningless.

I can find no literature suggesting that MS in any way causes false- positive HIV test results.

Nick Bennett njb35@cantab.net

Competing interests: None declared