In his illuminating book, Three Letter Plague, South African writer Jonny Steinberg describes the complex reasons why people at high risk of HIV do not take up the offer of testing. Steinberg follows a young man, Sizwe Magadla, who lives in a small, poor village. Magadla is due to get married, and his fiancée is pregnant. He owns a shop, which lends him enormous social standing. But should he test positive, he fears his marriage plans would be cancelled and his shop’s custom ruined. Witchcraft and curses are used to explain illness and medical problems; the cultural fabric wreathed around HIV is made of shame and fear.

When testing comes to the village, rapid test kits are used, which check for antibodies to HIV in a similar way to blood tests. Sexual health clinics often have trouble getting people to return for test results, so doing everything in “one stop” is highly attractive. Many clinics in the developed world find the rapid tests popular, since oral swabs mean no needles.

However, there is a risk of false positives. This occurs when the test confuses the antibodies made in the presence of HIV with other similar antibodies. The test “misreads” them, and reports the presence of HIV when there is none. But the number of false positives depends, at least in part, on the prevalence of the disease.

In a population with a low prevalence, the risk of false positives is high. An Oregon doctor, in a letter to the Journal of the American Medical Association, put this in context using her state’s background rate of HIV infection – 0.4 per 1,000. Even an HIV test which is 99.9 per cent accurate would, she estimated, engender a false positive rate 2.5 times that of the true positive rate. In other words, if you had one rapid HIV test that proved positive, you are still more likely not to have HIV than to have it.

The US government’s Centers for Disease Control and Prevention reported clusters of false positive results from rapid HIV tests in New York, performed in sexual health clinics between 2005 and 2008. They noted that HIV testing was increasing, which they attributed to the availability of rapid tests. But this came at a cost. The clinics performed 138,581 rapid HIV tests in 30 months. Of these, 1,720 tested positive. These tests were usually repeated using standard means. In the retested group, 1,296 out of 1,664 tested positive, and 368 tested negative. Twenty-two per cent of the initial positive results were false positives.

The manufacturers do make it clear that positive rapid HIV tests need to be redone – not just “confirmed”. For Magadla, these differences are almost irrelevant. By the end of the book, after hours of discussion, he is no nearer to having the test. No matter how accurate or how quickly the results are available, it is only useful if it is acceptable to the people who are likely to benefit the most.

Margaret McCartney is a GP in Glasgow.
margaret.mccartney@ft.com

For lively discussion of the latest medical issues go to the FT’s Healthblog

Margaret McCartney is joint winner of the European School of Oncology’s Best Cancer Reporter for 2009. www.cancerworld.org