Re: Is there proof of a correlation between the "HIV" antibody and PCR tests and what may underlie such a correlation? 15 December 2003
Previous Rapid Response Next Rapid Response Top
Christopher J Noble,
postdoctoral fellow
Bern Switzerland

Send response to journal:
Re: Re: Is there proof of a correlation between the "HIV" antibody and PCR tests and what may underlie such a correlation?

The Perth Group are once again attempting to deceive the readers of BMJ.

They write: "Most importantly, although these five studies may have been blind they may not have had controls as Owens et al point out. Thus it is not possible to state how many if any of the studies in the meta-analysis were indeed blind controlled studies."

The meta-study clearly states that all studies that were awarded a rating of 4 or 3 used both negative and positive controls (see table 1). These same studies were also blinded, thus these 11 were blind controlled studies.

These 11 studies showed "sensitivity ranged from 83% to 100%, and specificity ranged from 95% to 100%".

The Perth Group also attempt to argue that because the meta-study concluded that "Clearly, the performance of PCR is not adequate to justify its use as a clinical screening test" that this means that "neither the specificity of the PCR nor its correlation with the antibody tests has been proven". This is a non-sequitur. The study clearly shows that PCR results and antibody results are correlated. At least for the present, the specificity and sensitivity of PCR tests are lower than those for antibody tests. It is for this reason that antibody tests are recommended, where possible, rather than PCR for the diagnosis of HIV infection. It is quite simple to see that when there are two tests the one with the highest sensitivity and specificity should be used. It does not mean that the second is completely non-specific as the Perth Group would have us believe.

So, I return to my original question. How can the theories of the perth Group account for the correlation between the results of PCR and antibody tests for HIV infection? The observation of this correlation does not depend on the assumption that HIV exists. I am interested in hearing how their oxidative stress theory predicts this observation. I am still waiting for the Perth Group to respond without evasion or bold lies.

(1) Owens DK, Holodniy M, Garber AM, Scott J, Sonnad S, Moses L, et al. Polymerase chain reaction for the diagnosis of HIV infection in adults. A meta-analysis with recommendations for clinical practice and study design. Ann Int Med 1996;124(9):803-15

Competing interests: None declared