Incredible statistics
From Toby Gettins Sir: I congratulate The Spectator on breaking the story on the literally incredible Aids statistics in sub-Saharan Africa ('Africa isn't dying of Aids', 13/20 December). The antenatal tests that form the basis of the epidemiological models used by the WHO rely on Elisa testing. This test has never been calibrated against HIV, and so the false-positive rate for the test is completely unknown. It may be 100 per cent. It is certainly the case that the test has a high falsepositive rate when used on multiparous pregnant women and persons living in endemic disease conditions. Moreover the WHO epi-models then multiply the figures by a finger-in-the-air non-reportance factor (which historically has been as high as 49). This is ludicrous spreadsheet epidemiology.
The much-quoted epidemic in Africa would appear on current evidence to be completely specious: an artefact of driveby statistics of the most egregious — and possibly cynical — kind. We must use viral tests calibrated against the HIV virus and stop believing in fairy stories.
Toby Gettins
Birmingham
Previous page
