Accuracy

Although these tests have been used for the diagnosis of primary HIV infection, especially in the United States, they are not designed or approved by regulatory authorities for that purpose. False positive results are fairly common. A low level HIV RNA level (less than 5000 copies/ml) may be a false positive test result, because RNA levels during primary infection would normally be extremely high (above 100,000 copies/ml).

A CDC study assessed three NAAT tests against over 1100 samples of people who were either uninfected or had chronic infection. The sensitivities were somewhat low (indicating a reduced ability to correctly identify all HIV-positive samples) and varied according to the test used. The Roche Cobas Ampliscreen had a sensitivity of 92.6%, an in-house assay for RNA had a sensitivity of 95.8% and the Aptima qualitative assay had a sensitivity of 97.4%. Specificity was better, with scores of 96.9%, 99.4% and 99.6% respectively.¹

Most tests only detect HIV-1 – very few are capable of detecting HIV-2.

When these tests are run on pooled samples, there have been case reports of false negative results, possibly due to a sample with a very low viral load being harder to detect when mixed with other, uninfected, samples.²