Rapid tests

Rapid tests are often referred to as point-of-care tests because rather than sending a blood sample to a laboratory, the test can be conducted and the result read in a doctor’s office or a community setting, without specialised laboratory equipment. The sample taken is usually either finger-prick blood or saliva.

For people taking a test, the defining feature is that the result can usually be given within 20 or 30 minutes, which is often more convenient and can mean that many more people actually receive their test results.

However some laboratory professionals have viewed these tests with scepticism, noting inferior performance to fourth-generation tests and raising concerns about how quality control can be maintained away from the lab.

Rapid tests perform poorly in detecting very recent infection. Although a fourth-generation rapid test is now available (Determine HIV-1/2 Ag/Ab Combo) it does not match the performance of fourth-generation laboratory tests.

All HIV tests need to have reactive (positive) results confirmed with confirmatory tests. A particular challenge with rapid tests is how to communicate a reactive result to the client (who may be present while the result is being read) and explain that supplementary tests are needed.

Although the evaluation studies of rapid tests that are marketed in the UK have generally reported that the test are very specific (have a low rate of false positives), a number of testing sites in other countries have experienced unexplained rises in the numbers of false positives (particularly when using OraQuick on saliva samples).

Tests always produce a small number of false positive results, but the problems this will cause will depend on how many true positive results the test gives. This depends on the prevalence in the local population, and is expressed as the positive predictive value (PPV). In a low prevalence population of 0.1% (typical background UK HIV prevalence) a specificity of 99.4% leads to a positive predictive value of just 14%. That is, six of every seven positive tests are false positives. PPV improves with rising prevalence, being 97% with a prevalence of 15% (typical HIV prevalence in London gay men).¹

For that reason, less specific tests may appropriately be limited to use with high prevalence communities.

One option for confirming reactive results is to take a second rapid test, from a different manufacturer, immediately. If the second test is not reactive, it can be presumed that the person is not infected (as the tests are extremely sensitive). If the second test is reactive, it is likely that the person does have HIV, but a third test could still be performed. ^{2 3}

Nonetheless, the most common practice for most providers of rapid tests is to suggest that a negative result is definitive, but that a reactive result simply indicates the need for further laboratory testing.

Rapid tests are frequently available at outreach clinics run in the premises of voluntary organisations, community centres, commercial venues etc. An increasing number of NHS sexual health clinics also provide them, though they may be offered only to people with a higher risk of infection, to people with an urgent need to receive results, and to people who are very anxious.

Rapid tests were first developed in the early 1990s for use in developing countries (where specialised laboratory facilities may not be available), and their uptake has been very different in different countries. In the United States, the CDC has recommended their use with high-risk populations since 1998, and they have been an integral part of the testing strategy since 2003.

The UK testing guidelines are considerably more cautious: their use should be limited to clinical settings where a rapid turnaround of testing results is desirable, community testing sites, circumstances when venepuncture is refused, and for urgent source testing in cases of exposure incidents (e.g. before PEP).

Dozens of rapid tests have been developed, but to be suitable a test needs to be:

• Highly sensitive and specific (>99%).
• Easy for the operator to learn and use.
• Easy to read the result.
• As non-invasive as possible.
• Without complex storage requirements.