Re: Diagnosing “HIV” infection in neonates 1 October 2004
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: Re: Diagnosing “HIV” infection in neonates

The Perth Group do an admirable job of summarising the reasons why Serology is a superior diagnostic tool in ADULTS, but of course we were talking about neonates. As such, the hand of clinicians is forced in that serology is near useless except at later timepoints.

The results of the European Collaborative Study are at odds with the results here at SUNY Upstate, where serology does revert before 9 months in the vast majority of cases (at least when I made that specific query yesterday that was the answer I was given!) I question why the Perth Group assume that seroreversion between 9 and 22 months "cannot be due to loss of maternal antibodies" when in fact that is the most obvious explanation?

The tests cannot obviously replace clinical suspicion, and a negative result will get re-investigated if the child becomes inappropriately ill, or conversely a positive result will have further staging tests (CD4 count, viral load).

Since the Perth Group obviously agree with the fact that neonatal serology is useless, what else would they suggest? They present data that implies that nucleotide-based tests are less than ideal - no surprises there - but do not take the next logical step and think "what do we do then, in that situation?"

The answer is, we make do. It is not an argument for throwing the baby out with the bathwater, and not performing any kind of diagnostic test at all.

The Perth Group, and others writing here, are perfectly capable of linking together disparate facts to form an argument when it suits them, but are making critical errors in fact and logic. Regardless of the fact that nucleotide tests are inferior to serology, the logical step of therefore considering them useless cannot be made. Herpes encephalitis on the other hand is routinely diagnosed by PCR, because serology would be useless. Horses for courses.

They are also keen to quote from sources that support their theories, but fail to consider actual clinical practise. If they argue that "Researchers have shown that PCR is useless" and yet the standard of care in neonatal diagnostics here is DNA-PCR, surely that means that someone is out of the loop? I respectfully suggest it is the Perth Group. I do not intend that comment to imply that medicine ALWAYS gets it right (I can hear the cries of thalidomide already) but I do intend to imply that arguing a case when, based on the same evidence, standard practise goes contrary, suggests that the argument is biased. Such changes in thought must be evaluated carefully, and as Tony Floyd and others have shown, mere superficial analysis of the Perth Group's theories reveals gaping holes.

Competing interests: None declared